applied

sciences
Article
Switched Auto-Regressive Neural Control (S-ANC) for Energy
Management of Hybrid Microgrids
Muhammed Cavus 1,2, * , Yusuf Furkan Ugurluoglu 1,3 , Huseyin Ayan 1,4 , Adib Allahham 5 , Kabita Adhikari 1
and Damian Giaouris 1

1 School of Engineering, Newcastle University, Newcastle upon Tyne NE1 7RU, UK;
y.f.ugurluoglu2@newcastle.ac.uk (Y.F.U.); h.ayan2@newcastle.ac.uk (H.A.);
kabita.adhikari@newcastle.ac.uk (K.A.); damian.giaouris@newcastle.ac.uk (D.G.)
2 The School of Engineering, Iskenderun Technical University, Iskenderun 31200, Turkey
3 Department of Mechanical Engineering, Necmettin Erbakan University, Konya 42080, Turkey
4 The School of Engineering, Istanbul University-Cerrahpasa, Istanbul 34320, Turkey
5 Faculty of Engineering and Environment, Northumbria University, Newcastle NE1 8ST, UK;
adib.allahham@northumbria.ac.uk
* Correspondence: m.cavus2@newcastle.ac.uk

Abstract: Switched model predictive control (S-MPC) and recurrent neural networks with long short-
term memory (RNN-LSTM) are powerful control methods that have been extensively studied for the
energy management of microgrids (MGs). These methods ease constraint satisfaction, computational
demands, adaptability, and comprehensibility, but typically one method is chosen over the other.
The S-MPC method dynamically selects optimal models and control strategies based on the system’s
operating mode and performance objectives. On the other hand, integration of auto-regressive (AR)
control with these powerful control methods improves the prediction accuracy and the adaptability
of the system conditions. This paper compares the two control approaches and proposes a novel
algorithm called switched auto-regressive neural control (S-ANC) that combines their respective
strengths. Using a control formulation equivalent to S-MPC and the same controller model for
learning, the results indicate that pure RNN-LSTM cannot provide constraint satisfaction. The
novel S-ANC algorithm can satisfy constraints and deliver comparable performance to MPC, while
Citation: Cavus, M.; Ugurluoglu, Y.F.; enabling continuous learning. The results indicate that S-MPC optimization increases power flows
Ayan, H.; Allahham, A.; Adhikari, K.;
within the MG, resulting in efficient utilization of energy resources. By merging the AR and LSTM,
Giaouris, D. Switched
the model’s computational time decreased by nearly 47.2%. In addition, this study evaluated our
Auto-Regressive Neural Control
predictive model’s accuracy: (i) the R-squared error was 0.951, indicating a strong predictive ability,
(S-ANC) for Energy Management of
and (ii) mean absolute error (MAE) and mean square error (MSE) values of 0.571 indicate accurate
Hybrid Microgrids. Appl. Sci. 2023,
13, 11744. https://doi.org/10.3390/
predictions, with minimal deviations from the actual values.
app132111744
Keywords: auto-regressive; control and optimization; energy management; recurrent neural network;
Academic Editor: Frede Blaabjerg
long short-term memory; microgrid; switched model predictive control
Received: 12 September 2023
Revised: 19 October 2023
Accepted: 23 October 2023
Published: 26 October 2023 1. Introduction
Model predictive control (MPC) is a control approach that is widely utilized in many
industries, including chemical, electrical, and mechanical engineering. It is well-suited
Copyright: © 2023 by the authors.
to microgrids (MGs) because it deals with restrictions and optimizes performance over
Licensee MDPI, Basel, Switzerland.
time [1–5]. MPC entails formulating and solving an optimization problem at each time step,
This article is an open access article
to determine the optimal control inputs for the next step. A MPC was described in [6–9] for
distributed under the terms and effective MG optimization, and mixed integer linear programming (MILP) was employed
conditions of the Creative Commons to solve the problem posed. An MPC-inspired energy management (EM) system employing
Attribution (CC BY) license (https:// a neuro-fuzzy method, which accounted for renewable energy sources (RES’s) intermittent
creativecommons.org/licenses/by/ nature in grid-connected MG with loads and photovoltaic (PV) sources, was reported
4.0/). in [10]. Ref. [11] presented scenario-based stochastic programming with a rolling horizon

Appl. Sci. 2023, 13, 11744. https://doi.org/10.3390/app132111744 https://www.mdpi.com/journal/applsci

Appl. Sci. 2023, 13, 11744 2 of 23

strategy for minimizing the operating expenses of MGs when the wind speed is unknown.
Rolling horizon or MPC techniques are reactive-based methodologies that modify or
update data using deterministic approaches. A scenario-based MPC was developed in [12]
to reduce operating expenses and overall emissions. To achieve inexpensive and flexible
operation, Ref. [13] provided an MPC-based optimum management for renewable energy
MGs with hybrid energy storage systems (ESSs), such as hydrogen, batteries, and capacitors.
A hierarchical MPC-based technique for islanded AC MGs addressed power-quality and
unbalanced power-sharing difficulties [14]. Despite this, the traditional MPC cannot control
a MG in various operational modes.
In contrast, switched model predictive control (S-MPC) is a variant of MPC that
employs multiple models, each representing a unique mode of operation or scenario of
the system. S-MPC selects the optimal model and associated control strategy based on
the current system state and the desired performance goals. This makes it possible for
S-MPC to handle systems with mode-dependent dynamics. MPC is distinguished from
S-MPC in using a single model to predict the system’s future behavior [15]. S-MPC employs
multiple models and switches between them, based on the system’s current state. S-MPC
can provide better performance and robustness than MPC, especially for complex systems
with multiple modes or operating conditions [16,17]. Another novel study presented a
hybrid MG model that incorporated two switched receding horizon control laws. This
strategy reduces the overall energy expenses and maximizes the efficient utilization of
RESs for expansive business establishments, while accommodating fluctuations in grid
connectivity [18]. Moreover, Ref. [19] outlined the process of designing and applying a
S-MPC to wind turbine systems, intending to manage the intricate nature and nonlinearity
inherent in wind turbine systems. The system employs qpOASES as an integrated solver for
optimum online control. It incorporates a cyber-physical real-time emulator for utility-scale
wind turbines, with variable-speed and variable-pitch capabilities. The study showcased
the viability and efficacy of S-MPC in attaining control objectives for wind turbine systems
in real-time, utilizing brief control periods. In addition, the study in [20] presented a
novel technique for enhancing wind turbine control by introducing a S-MPC framework.
The proposed approach aimed to solve the limitations of the conventional continuous
control-based MPC algorithm. The results of the comparative analysis indicated that the
proposed algorithm exhibited superior performance compared to the existing MPC in
various aspects, including computational efficiency, load mitigation, and dynamic response.
Ref. [21] presented a novel S-MPC method specifically tailored to discrete-time nonlinear
systems. The simulation outcomes emphasized its superiority over a conventional MPC
technique regarding computational efficacy and control effectiveness. Another study
presented a novel S-MPC methodology for power converters. During transient periods, the
system utilized horizon-one nonlinear finite control set MPC to steer the system towards
the intended reference [22].
On the other hand, S-MPC performance is highly vulnerable to model mismatch. In
other words, it must select a suitable system model. Furthermore, the increased complexity
of S-MPC impacts the stability and maintainability of MG control [23,24]. These challenges
lead to accuracy issues with S-MPC methods. In addition, the computational time of S-MPC
is much longer, because of the prediction horizon and various steps. Many authors have
studied machine learning (ML) techniques to increase the accuracy of MG systems.
When improving the scheduling effectiveness in networked microgrids (NMGs), the
main goal is minimizing the effects of electricity outages.This paper presents a framework
consisting of three stages, to evaluate power transactions, manage renewable energy and
market price risks, and tackle uncertainties. This framework is formulated as a mixed-
integer linear programming problem [25–28]. On the other hand, ref. [29] introduced a
novel approach utilizing the Internet of things (IoT) to dynamically optimize and regulate
power loads in citizen energy communities. This technique is compared to the conventional
direct load control (DLC) method. This technique aims to enhance power use efficiency
using programmable appliances and dynamic demand response. In order to model the
Appl. Sci. 2023, 13, 11744 3 of 23

behavior of RESs, such as wind and solar, auto-regressive moving-average (ARMA)-based

scenario generation was implemented. Large industries could receive direct assistance from
storage and demand-side management systems, to reduce energy costs [30]. The other work
employed an ARMA model to forecast solar PV, wind power generation, and electricity
demand. Second, an optimal generation scheduling procedure is intended to reduce system
operating expenses. The simulation results indicated that the optimal generation scheduling
could minimize operating expenses under the worst-case scenario [31]. In [32], combining
two models, the ARMA and nonlinear auto regressive with exogenous input (NARX),
a novel method was presented for predicting solar radiation. The decision was made
to utilize the benefits of both models, to produce more accurate prediction results. The
simulation results validated this hybrid model’s ability to predict weekly solar radiation
averages. Although the previous solar radiation forecasting techniques, particularly ARMA
models, are effective for particular uses, they are unsuitable for other uses requiring a high
forecasting precision. Several researchers have proposed hybrid models to improve the
precision of solar radiation forecasting. Moreover, there is still not a proper plant model or
prediction horizon, so the computational time of the model is still to high [32].
There have been numerous studies on using ML methods in place of AR models. For
instance, ref. [33] thoroughly investigated the prediction performance of several recur-
rent neural network (RNNs) designs, such as a long short-term memory (LSTM), gated
recurrent unit (GRU), and bidirectional LSTM. Using local weather forecasts and historical
weather data, ref. [34] proposed a LSTM-based next-day forecasting model of hourly global
horizontal irradiance (GHI). Refs. [35,36] suggested LSTM-based models using only the
next day’s weather forecast as input. The studies by [37,38] used similar LSTM-based
techniques. Refs. [39,40] validated the performance of hybrid deep learning models built
on convolutional neural networks (CNNs) and LSTM for day-ahead GHI forecasting. In
addition to RNN-based approaches, there have been studies evaluating the performance
of other statistical and ML models for solar irradiance forecasting, such as coupled AR
and a dynamic system by [41], a Markov switch model [42], and a support vector machine
(SVM) by [43]. Ref. [44] reported an LSTM-based model for hour-ahead solar irradiance
forecasting. The inputs, which included historical GHI and meteorological data from the
preceding 24 h, were utilized to forecast the GHI for the next hour [44]. The results revealed
that the LSTM-based model outperformed the other models, such as auto-regressive inte-
grated moving average (ARIMA) and CNN [44]. Ref. [45] investigated the performance
of LSTM and GRU. Refs. [46,47] published hybrid CNN-LSTM models for hour-ahead
GHI forecasting. Their study showed that incorporating external weather information
considerably increased the prediction accuracy. Unlike day-ahead irradiance forecasting
methods, hour-ahead forecasting algorithms create projections for the following hour only
using historical data.
On the other hand, RNNs are a form of ML technology widely employed for time
series prediction and the modeling of dynamic systems [48,49]. RNNs are artificial neural
networks (ANN) that are particularly useful for modeling time-series data and may be used
to anticipate future MG behavior [50,51]. RNNs may learn and adapt to system dynamics
by learning the temporal dependencies in data. RNNs have been used to solve various MG
control challenges, including load forecasting, renewable energy integration, and demand
response management [52–54]. RNNs have been applied to various systems, including
power systems [55,56], with promising prediction accuracy and flexibility results.
In summary, both control families have benefits and drawbacks, and their comple-
mentarity is evident. On the one hand, S-MPC struggles with system complexity and
long-term prediction horizons, whereas the combination of AR and LSTM (AR-LSTM) can
deal with complex systems and infinite prediction horizons. AR-LSTM, conversely, has
difficulty in satisfying constraints and lacks interpretability, whereas S-MPC can provide
safety guarantees and understandability.
Although there is a clear potential for a synergy between the two families of methods,
there have been few attempts to combine their relative advantages. This research deficiency
Appl. Sci. 2023, 13, 11744 4 of 23

is not limited to applying EM for MG. The control and ML communities evolve indepen-
dently, adopting radically different notations to formulate the same problem. In spite of the
parallel developments, several authors [57–59] have suggested that a collaboration between
the two groups could result in potential advantages. Combining these methodologies
represents a powerful method for integrating robust control theory methods with ML
approaches, to exploit additional information from real-time data [60,61].
As shown in Table 1, each control method has its strengths and limitations. MPC
and S-MPC offer robust optimality and constraint handling but may have computational
challenges. AR and RNN-LSTM are efficient in computation but may not manage complex
constraints effectively. S-ANC combines AR models with neural control, balancing opti-
mality and computational efficiency. The choice of control method depends on the specific
application and trade-offs between these criteria.

Table 1. Comparison of Control Methods.

Computational
Control Method Optimality Multiple Models Adaptability Constraints
Time [s]
MPC [62,63] X >104 [High] × Good X
S-MPC [15,16,20,64] X ≈104 [Moderate] X Outstanding X
DLC [29] X ≈104 [Moderate] × Good X
AR [32] × ≈104 [Moderate] × Poor ×
CNN [65,66] × <104 [Low] × Poor ×
RNN-LSTM [67,68] X <104 [Low] × Poor ×
S-ANC X < 104 [Low] X Outstanding X

Contributions and Research Questions

This paper was motivated by how AR-LSTM and S-MPC could collaborate in applying
the EM of MGs. While there is a consensus that combining the two algorithms may yield
benefits, little has been done to develop methods that involve the two algorithms working
together. In addition, works have investigated how these controllers can collaborate with
the algorithms working with different control designs and modes. No previous research
has compared and combined S-MPC and AR-LSTM for the same optimal control problem
formulation in EM for MGs.
The second objective of this paper was to propose a novel method known as switched
auto-regressive neural control (S-ANC), which merges S-MPC and AR-LSTM synergistically.
The development and formulation of this new S-ANC algorithm were motivated by the
conceptual and practical comparison of S-MPC and AR-LSTM. In contrast to comparable
approaches, our method combines the S-MPC objective function and constraints with the
AR-LSTM optimization and prediction function. This practice ensures interoperability
between the two methods and enables the truncation of the S-MPC optimization problem,
which can become highly complex, even for relatively simple MG structures. Finally, a
flexible hybrid MG case is used to describe and evaluate this new algorithm.
Consequently, the primary contribution of this paper is the introduction of S-ANC, a
control algorithm that combines techniques from the communities of control theory and
ML. This algorithm is evaluated, and a new standard framework is generated for EM of
hybrid MGs. In addition, the proposed S-ANC algorithm applies to various applications
and domains, such as complex industrial processes and energy markets. This study also
combines control theory and ML by comparing and disentangling the key distinctions
between S-MPC and AR-LSTM.
Appl. Sci. 2023, 13, 11744 5 of 23

2. Identifying the Distinctions between S-MPC and AR-RNN-LSTM

Optimal control determines the actions that optimize a performance objective by
solving a sequential decision-making problem. The preceding section highlighted the
need to compare S-MPC and AR-LSTM, the two primary approaches for optimal control
applied to EM for hybrid MG control, and the possibility of combining them. Both methods
utilize certain components, while others are more controller-specific. These formulation
differences make it difficult to compare and combine the two approaches, necessitating a
conceptual analysis. This assists in identifying the primary methods of optimal control and
establishes a common ground for a comprehensive classification. The sections that follow
detail the most important aspects of these control methods.

2.1. Strategy
There are typically two ways to approach an optimal control problem: by employing
the S-MPC-inherent receding horizon principle or formalizing the problem as an AR-LSTM.
S-MPC is a control strategy that involves using a mathematical model of the system
being controlled to predict the system’s future behavior and optimize a control signal
over a finite time horizon. At each time step, the control signal is updated based on the
current state of the system and the predictions made by the model. This is widely used in
industrial control applications, such as process control, automotive control, and robotics,
where it is important to consider the system dynamics being controlled and to optimize
performance over a prediction horizon. At each time step k in S-MPC, switching logic
controls multiple modes for the accumulators, which fully describe the controller model at
the current time.Then, the trajectories of the future state x and input u are optimized for a
prediction horizon NP based on the explicit representation of an objective function J and
a controller model F. J is the minimization of the imported energy and maximization of
the exported energy. The constraints H are also introduced explicitly in the optimization
problem. The objective function, model, and constraints may also depend on the model
outputs y and time-invariant parameters p. In addition, r (k) is the reference variable
representing the PV, load data, and zero along the prediction horizon NP . wx (k) and
wu (k ) are weighting coefficients reflecting the relative significance of x (k) and penalizing
relatively large variations in u(k), respectively. Only the initial control input from the
optimized trajectory [16] is Implemented. Figure 1a depicts the full S-MPC procedure.

NP NP NP
J (k) = ∑ wx (k)(r(k) + x(k))2 + ∑ wu (k)∆u(k)2 + ∑ wy (k)y(k)2 + p2 (1)
k =0 k =0 k =0

In the application of S-MPC to EM of MGs, the state vector x represents the state of
charge of the accumulators (SOAcc), such as the battery, fuel tank, and water tank, and the
model output y illustrates the imported and exported energy, such as a grid to the load
GR LD and PV to the grid, and the battery (PVGR + PVBAT ). Depending on whether or not
the controller model employs physical insights, the set of time-invariant parameters p may
or may not represent the physical properties of the MG.
In contrast to RNN-LSTM, AR models are not neural network architectures. On
the contrary, they are statistical models that identify dependencies and patterns within
a time series, based on its own lagged values. The AR model predicts the future values
of a variable based on its historical values and the estimated coefficients during model
training. In other words, AR models are a statistical modeling technique that assumes a
variable’s current value is a function of its previous values. They are frequently utilized
for time series analysis and forecasting. Therefore, AR models can be viewed as a linear
regression, in which the predictors are the values of the same variable at a prior time [32].
AR models can be used to model the system’s dynamics within the context of control
systems or reinforcement learning. A model can predict future states or observations by
estimating the AR coefficients. These predictions can then be fed into control algorithms
or reinforcement learning agents, in order to optimize control signals or decision-making.
Appl. Sci. 2023, 13, 11744 6 of 23

Unlike neural network architectures, AR models are not adaptive by nature. The estimation
of AR coefficients requires training on historical data, and their performance may degrade
if the underlying dynamics of the system change significantly over time.

Controller
Optimizer
𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁
2 2 2
𝐽𝐽 𝑘𝑘 = � 𝑤𝑤𝑥𝑥 𝑘𝑘 𝑟𝑟 𝑘𝑘 + 𝑥𝑥 𝑘𝑘 + � 𝑤𝑤𝑢𝑢 𝑘𝑘 ∆𝑢𝑢 𝑘𝑘 + � 𝑤𝑤𝑦𝑦 𝑘𝑘 𝑦𝑦 𝑘𝑘 + 𝑝𝑝2
𝑢𝑢(𝑘𝑘)
𝑘𝑘=0 𝑘𝑘=0 𝑘𝑘=0 Plant
0 = 𝐹𝐹(𝑥𝑥 𝑘𝑘 , 𝑟𝑟 𝑘𝑘 , 𝑢𝑢 𝑘𝑘 , 𝑦𝑦 𝑘𝑘 , 𝑝𝑝)

0 ≤ 𝐻𝐻(𝑥𝑥 𝑘𝑘 , 𝑟𝑟 𝑘𝑘 , 𝑢𝑢 𝑘𝑘 , 𝑦𝑦 𝑘𝑘 , 𝑝𝑝) 𝑦𝑦(𝑘𝑘)

k+1
𝑁𝑁𝑃𝑃 𝑁𝑁𝑃𝑃 𝑦𝑦(𝑘𝑘)
�

𝑦𝑦(𝑘𝑘 + 1)
Switching logic Reference Predictor

(a) S-MPC

Objective function
𝑇𝑇 θ 𝑘𝑘
Auto-regressive model
𝐽𝐽 𝜃𝜃 = � 𝐿𝐿(𝑦𝑦� 𝑘𝑘 , 𝑦𝑦 𝑘𝑘 )
𝑘𝑘=1

k+1
Predictor
𝑋𝑋 𝑘𝑘 + 1
𝑦𝑦(𝑘𝑘)
�
𝑹𝑹 𝑘𝑘 𝑋𝑋 𝑘𝑘
LSTM 𝑦𝑦� 𝑘𝑘

𝑦𝑦(𝑘𝑘 − 1) 𝑦𝑦(𝑘𝑘)
𝑦𝑦� 𝑘𝑘 − 1
k
FG IG
𝑥𝑥(𝑘𝑘 − 1) OG 𝑥𝑥(𝑘𝑘)

𝑢𝑢(𝑘𝑘)

(b) AR-LSTM
Figure 1. Block diagram of (a) S-MPC and (b) AR-LSTM.

The following equation can mathematically represent an AR model of order q [32]:

X ( k ) = c + ϕ (1) X ( k − 1) + ϕ (2) X ( k − 2) + · · · + ϕ ( q ) X ( k − q ) + ε ( k ) (2)

where X (k) represents the value of the time series at time k in this equation. c is a constant
term or an intercept. ϕ terms represent AR model coefficients. The coefficients or weights
associated with the previous values of the time series are denoted by 1, 2, . . . , q. X (k − 1),
and X (k − 2), . . . , X (k − q) represent the lagged values of the time series at time points k − 1,
k − 2, . . . , k − q, respectively. ε(k) is the error term or random noise at time k, representing
the data portion the model cannot explain.
RNN-LSTM is a neural network type ideally suited to processing sequential data.
Unlike feed-forward neural networks, it has loops that allow information to be passed from
one sequence step to the next. The approach for employing RNN-LSTM includes selecting
an appropriate network architecture, an optimization algorithm for training the network,
and an appropriate set of hyperparameters. RNN-LSTM is an extension of a feed-forward
neural network with internal memory. RNN-LSTM is recurrent because it performs the
same function for each data input, while the output of the current input is dependent on the
previous computation. After the output has been generated, it is duplicated and sent back
into the recurrent network [69]. For decision-making, it considers both the current input
and the output from the previous input it learned. As shown in Figure 1b, the input vector
of an LSTM network is u(k − 1) at time step k. y(k) represents the output vectors passed
through the network between time steps k and k + 1. Three gates update and control the
Appl. Sci. 2023, 13, 11744 7 of 23

cell states in an LSTM network: the forget gate, input gate, and output gate. The gates
are activated by hyperbolic tangent and sigmoid functions. Given new information that
has entered the network, the forget gate determines which cell state information to forget.
Given new input information, the input gate determines what new information will be
encoded into the cell state. Using the output vector y(k), the output gate controls what
information encoded in the cell state is sent to the network as input in the subsequent
time step.
In the mathematical modeling of RNN-LSTM, the current state can be expressed
mathematically as

x (k) = f ( x (k − 1), u(k )) (3)

where x (k) represents the current state, x (k − 1) represents the previous state, and u(k) is
the current input. Because the input neuron would have applied the transformations to
the previous input, we now have the state of the previous input rather than the input itself.
Each successive input is, therefore, referred to as a time step.
Considering the simplest form of RNN-LSTM, where the activation function is tanx,
the weight at the recurrent neuron is Wxx , and the weight at the input neuron is Wux , we
can write the equation for the state at time k as follows [69]:

x (k) = tanx (Wxx x (k − 1) + Wux u(k)) (4)

In this instance, the recurrent neuron only considers the previous state. The equation
may involve multiple such states for longer sequences. After calculating the final state, the
output can be generated. Once the current state has been computed, we can then calculate
the output state as follows [69]:

y(k) = Wxy x (k) (5)

where y(k) is the output state and Wxy is the weight at the output state. This process is
represented in Figure 2.

𝑦𝑦 𝑦𝑦(𝑘𝑘 − 1) 𝑦𝑦(𝑘𝑘) 𝑦𝑦(𝑘𝑘 + 1)

𝑊𝑊𝑥𝑥𝑥𝑥 𝑊𝑊𝑥𝑥𝑥𝑥 𝑊𝑊𝑥𝑥𝑥𝑥 𝑊𝑊𝑥𝑥𝑥𝑥

𝑥𝑥 𝑓𝑓 𝑓𝑓 𝑓𝑓 𝑓𝑓
𝑓𝑓 𝑥𝑥(𝑘𝑘 − 1) 𝑥𝑥(𝑘𝑘) 𝑥𝑥(𝑘𝑘 + 1)
Blossom
𝑊𝑊𝑢𝑢𝑢𝑢 𝑊𝑊𝑢𝑢𝑢𝑢 𝑊𝑊𝑢𝑢𝑢𝑢 𝑊𝑊𝑢𝑢𝑢𝑢
𝑢𝑢
𝑢𝑢(𝑘𝑘 − 1) 𝑢𝑢(𝑘𝑘) 𝑢𝑢(𝑘𝑘 + 1)

Figure 2. Structure of the RNN.

First, it extracts u(0) from the input sequence and then outputs y(0), which, along
with u(1), is the input for the subsequent step. Therefore, y(0) and u(1) are the inputs for
the subsequent step. Similarly, y(1) from the subsequent step is the input for u(2) for the
subsequent step, and so on. Consequently, it remembers the context throughout training.
A cost function quantifies “how well” a neural network performs with respect to
the training sample and the expected output. This may also depend on factors such as
weights and biases. This is a single value, not a vector, because it evaluates the overall
performance of the neural network. The objective of the cost function is to evaluate the
Appl. Sci. 2023, 13, 11744 8 of 23

network’s performance, to minimize its value during training. The cost function for a
typical RNN-LSTM is the sum of losses at each time step [70].

T
J (θ ) = ∑ L(ŷ(k), y(k)) (6)
k =1

where θ represents the parameters of the RNN, T represents the length of the input sequence,
ŷt represents the predicted output, and yt represents the actual output at time step k. L
is the loss function quantifying the difference between the predicted and actual output.
The RNN’s training parameters are adjusted to minimize the cost function using gradient
descent or a comparable optimization algorithm. The objective is to identify the parameters
that minimize the loss over all time steps, resulting in an RNN that can accurately predict
the output for a given input sequence.

2.2. Problem-Solving Method

By analyzing the control processes illustrated in Figure 1a,b, it is possible to identify a
number of expressions with total or partial equivalence between the two methods.
S-MPC can be solved implicitly by performing switching logic, forecasting, and resolv-
ing a dynamic optimization problem at each time step or explicitly by learning a control
policy from data generated by a S-MPC with any function approximation. Consequently,
S-MPC has a higher online computational cost, because every control step requires esti-
mation of the states and dynamic optimization. Typically, the optimization problem in
S-MPC is solved using numerical optimization techniques, such as nonlinear programming
or quadratic programming (QP) (in this paper, QP has been used), to solve the optimization
problem. The solution to the optimization problem over the prediction horizon provides
the optimal control signal. At each time step, the first component of the optimal control
signal is applied to the system, and the process is repeated with the updated state and
prediction horizon values. S-MPC requires the solution of an optimization problem at each
time step, which can be computationally expensive for large systems.
The training process for AR-LSTM involves back-propagation through time (BPTT),
a variation in the back-propagation algorithm that considers temporal dependencies in
the data. The RNN is unrolled throughout the training for a predetermined number of
time steps, and gradients are calculated at each step. The RNN’s weights are then updated
based on the gradients accumulated across all time steps. The most prevalent optimization
algorithm for training RNNs is gradient descent, which involves updating the weights
iteratively in the direction of the loss function’s negative gradient [69]. However, the
standard gradient descent algorithm is susceptible to issues such as vanishing gradients,
in which the gradients become extremely small and the weights do not update. Several
variants of gradient descent, such as the adaptive gradient descent algorithms AdaGrad,
RMSProp, and Adam, have been developed to address this issue [71].

2.3. Peak Performance

In S-MPC, the quality of the optimization solution depends on the controller model’s
precision, which is frequently simplified for computational purposes. Stability and practi-
cability are intrinsically ensure for S-MPC, whereas there is only an immature theory for
these issues in AR-LSTM [72]. The absence of safety guarantees in AR-LSTM results from
the constraints not being directly imposed in the formulation of the solution method. The
optimality of the S-MPC solution depends on the accuracy of the model used to predict
the system’s behavior and the optimization algorithm’s ability to find the optimization
problem’s global optimum. If the model is inaccurate or the optimization algorithm fails to
find the global optimum, the performance of the S-MPC controller may not be optimal.
The optimality of AR-LSTM relies on several factors, including the network’s architec-
ture, the training optimization algorithm, and the complexity of the task being performed.
AR-LSTM is capable of achieving high levels of performance on a wide variety of sequen-
Appl. Sci. 2023, 13, 11744 9 of 23

tial data processing tasks, such as language modeling, machine translation, and speech
recognition. AR-LSTM is able to model complex temporal dependencies in sequential
data, which is one of its main advantages. The ability of AR-LSTM to incorporate feedback
loops enables them to capture long-term dependencies that would be challenging to rep-
resent using other models, such as a gated recurrent unit (GRU). In addition, the ability
to incorporate memory into the network via mechanisms improves the performance of
RNNs in tasks requiring long-term memory. Nonetheless, several factors can restrict the
optimality of AR-LSTM. One difficulty is the issue of vanishing and exploding gradients,
which can hinder the network’s ability to discover long-term dependencies. This issue can
be mitigated by employing specialized units, such as LSTM and GRU, and optimization
algorithms designed to deal with these issues. Another issue is overfitting, which can
occur when the model becomes excessively complex and begins to fit the noise in the data
rather than the underlying patterns. This can be remedied by employing regularization
techniques such as early stopping and dropout [73].

2.4. Calculational Effort

S-MPC can require significant computational effort, especially for large-scale systems.
S-MPC necessitates the solution of an optimization problem in each time step, which can
be computationally costly. Moreover, a significant disadvantage of S-MPC is the need to
solve an optimization problem online, which can be complex and involve many optimiza-
tion variables. Consequently, controller models for S-MPC are commonly simplified at
the expense of optimality, and gains in optimization solver efficiency are highly desired.
Moreover, switching logic and prediction must be performed at each control step. Nonethe-
less, several techniques have been developed to reduce the computational effort required
for S-MPC, such as online optimization and ML techniques that update the optimization
problem as the system evolves.
The computational effort required for training and utilizing AR-LSTM can be substan-
tial, especially for large-scale problems with many time steps and/or parameters. BPTT is
the primary computational bottleneck, because it is required to compute the gradients of the
loss function with respect to the network parameters. Considering that the computational
complexity of BPTT scales linearly with the number of time steps, training AR-LSTM on
lengthy sequences can be computationally expensive. In addition, the number of network
parameters can contribute to the computational complexity, as larger networks require
more computation to update weights during training and to make predictions during
inference. Several techniques have been developed to ease these computational challenges,
including mini-batch training, which involves updating the weights based on a subset
of the training data at each iteration, and gradient clipping, which involves capping the
magnitude of gradients to prevent gradients from exploding during training [69].

3. Switched Auto-Regressive Neural Control (S-ANC)

This section introduces the specifics of the proposed novel S-ANC algorithm. The
objective is to learn from the architecture of RNN-LSTM, while satisfying constraints.
Switching logic, dynamic optimization, and learning are the elements from the control and
ML communities that are effectively combined to achieve this objective. First, Section 3.1
introduces the hybrid MG structure. Section 3.2 provides an overview of how S-MPC and
AR-LSTM are merged logically. Then, Section 3.3 formally describes the S-ANC algorithm.

3.1. Hybrid MG Description

This is a case study of a system constructed in Xanthi, Greece [74]. As depicted in
Figure 3, the hybrid MG is comprised of a 15 kW PV array; a battery (BAT), a water tank
(WT), and a fuel tank (FT) serving as energy storage systems (ESSs); an electrolyzer (EL);
and a fuel cell (FC), as well as the utility grid (GR). The PV can be utilized in the hybrid
MG as the primary energy source. If the PV cannot provide sufficient power, the BAT or
the FC will meet the load. The GR will provide energy if the battery is depleted and no
Appl. Sci. 2023, 13, 11744 10 of 23

hydrogen is available. Alternatively, when the BAT is full and there is an excess, the EL will
be utilized if there is space in the WT and the FT. The energy will then be sent to the GR.

AC Bus
PVLD = P2

Building
PPV PPV PVBAT = P4 PVBAT = P4
DC AC
AC DC
Converter
BATLD = P5 BATLD = P5
Bi-directional converter Battery
PV Panel
PVGR = P1 PVGR = P1
AC
FCBAT = P6 DC FCBAT = P6 DC
AC GRLD = P3 GRLD = P3
Bi-directional converter
Converter Utility grid
Fuel cell ELFT = P8 ELFT = P8
AC
DC
BATEL = P7 BATEL = P7 FTFC = P9 FTFC = P9
DC Fuel tank
AC Bi-directional converter

Converter
Electrolyzer FCWT = P10 FCWT = P10
AC
DC
WTEL = P11 WTEL = P11
Bi-directional converter Water tank

Figure 3. Hybrid MG Structure.

3.2. Simple Definition of the Proposed Method

To comprehend the concept underlying S-ANC, one must first comprehend the distinc-
tions between MPC and S-MPC when solving the QP and the two main learning methods
of RNN: AR and LSTM. The S-MPC controller must be capable of selecting the appropriate
model and control strategy based on the system’s current state, which necessitates addi-
tional computational resources and algorithmic complexity. In this paper, for instance, the
system dynamics change significantly as the state of each accumulator in the hybrid MG
changes; consequently, S-MPC can use different models for various states. This requires
creating and validating multiple models, and the S-MPC controller must be able to switch
between these models based on the current state.
The construction of S-MPC is challenging and intricate, particularly for a hybrid MG,
which must accommodate many operating modes and complex switching conditions. This
complexity is caused by a number of factors, including
• Model development: S-MPC requires the creation of multiple models that represent
the system’s behavior in different operating modes. This requires an efficient system
architecture and behavior;
• Mode detection: The S-MPC controller must be able to detect the current mode of
operation of the system, which can be difficult in certain circumstances;
• Switching logic: The S-MPC controller must select the appropriate model and control
strategy based on the current operating mode and desired performance objectives.
This necessitates the designing of switching logic that maps the system’s current state
to the appropriate model and control strategy (a mode’s objective function and an
operational mode’s objective function may differ).
The S-MPC solution method takes information from the hybrid MG, such as PV
and load data and accumulator parameters, including their charging and discharging
efficiencies. Then, the input u, state x, and output vectors y are defined. Based on the
controller model, the objective function J is inferred at each control step using this method.
Appl. Sci. 2023, 13, 11744 11 of 23

After that, the state vector is converted to an AR model X (k), in order to predict the value
at the subsequent time step. This is a straightforward concept that can produce accurate
forecasts for various time series problems. Nevertheless, the AR model needs a plant model
and a prediction horizon, so the computational time of the model is still high. Therefore,
the current state x, input u, and output vectors y are updated using the AR-LSTM method.
S-ANC employs time series value-based AR-LSTM to estimate the value of being in a
particular output vector ŷ, as determined by S-MPC, with a prediction horizon of only one
control step. By doing so, in S-ANC, the S-MPC method is truncated with the predicted
output vector ŷ and optimized by the hybrid MG system during k steps ahead through
employing the AR-LSTM method.Consequently, the principal components of the S-MPC,
namely the reference, predictor, and switching logic, remain active in S-ANC; however,
the time series value function is utilized to shorten the nonlinear program and enable
learning. The interaction of S-ANC’s primary components is depicted in a diagram in
Figure 4. The merging of S-MPC and AR-LSTM in the S-ANC algorithm is intuitively
depicted in Figure 4.

S-ANC

Auto-regressive
S-MPC + +
RNN-LSTM 𝑼𝑼(𝑘𝑘)
𝑿𝑿(𝑘𝑘)
Plant/Auto-
regressive model

𝑁𝑁𝑃𝑃 𝑁𝑁𝑃𝑃 𝑦𝑦(𝑘𝑘)

� 𝑹𝑹(𝑘𝑘) 𝑘𝑘 + 1

Switching 𝑦𝑦(𝑘𝑘)
Reference Predictor 𝑘𝑘
logic

Figure 4. Block diagram showing the introduction of S-ANC.

3.3. Formal Definition

Initially, the system state, control, and output vectors are defined for the hybrid MG
system in the S-MPC:
The system-state vector of the MG is as follows:

x (k ) = [SOAccl (k )] (7)

where l ∈ { BAT, FT, WT }. SOAcc BAT (k), SOAcc FT (k), and SOAccWT (k) are the state of
the accumulators for the battery, hydrogen tank, and water tank, respectively.
The system-control (input) vector of the MG is defined as follows:

u(k) = [ P3 (k); P4 (k ); . . . P11 (k)] (8)

The system-output vector of the MG is defined as follows:

y(k) = [ P1 (k); P2 (k)] (9)

Consider the discrete-time linear state-space system:

X ( k + 1) = A x x ( k ) + B u u ( k ) (10)

where k = 0, 1, 2, . . . , NP − 1 symbolizes the discrete-time instant.

Appl. Sci. 2023, 13, 11744 12 of 23

By defining the following matrices:

0 0 ··· 0
 
I
 
 A   B 0 ··· 0
.. 
 
A2
  
Ax =  Bu =  AB B ··· . (11)
 
 ..  
.. .. ..

.
 
. . . 0
  
A NP ; A NP −1 B A NP −2 B · · · B

where

 
0 0 0
 0 0 0 
 
 η
 ch 0 0 

−η 0 0 
1

0 0
  dis 
0 0
 
 η
A = 0 1 0 B =  ch
T
(12)

−ηdis 0 0

0 0 1

 0 ηch , H2 0
 

−ηdis , H2
 
 0 0 
 
 0 0 ηch , H2 O 
0 0 −ηdis , H2 O

The linear state-space equation can be stated depending on the battery, fuel tank, and
water tank equations as follows [75]:
j j
Pa→b (k) − Pb→ a (k)
SOAccl (k + 1) = SOAccl (k) + (13)
Cl

where j is the energy flows, so j ∈ { Power, Hydrogen, Water }. a → b represents the energy
flows between accumulators and converters; for example, P4 is the power from the PV to
the battery.
Define the constraints for the hybrid MG: Energy flows from the PV, GR, BAT, FT, EL,
FC, and WT are positive and subject to maximum values.

0 ≤ P1 (k) = PLD (k ) − y a (k) ≤ P1max

j jmax
(14)
0 ≤ Pm (k) ≤ Pm

jmax
where Pm (m = 1, 2, . . . , 11) imply the maximum values of energy/matter flows.
The sum of PV energy supplied directly for the load ( P2 (k)) and the battery for the
charging ( P4 (k)) should be smaller than the energy flow from the PV array, ( PPV (k)).

P2 (k ) + P4 (k) ≤ PPV (k) (15)

The SOAccl is restricted between the minimum and maximum values [16].
min max
SOAccl ≤ SOAccl ≤ SOAccl (16)

Define the reference matrix (R) for the hybrid MG system:

R(k) = [wxx ( PLD (k); PPV (k); 0); . . . ; PLD (k + Np − 1); PPV (k + Np − 1); 0)] (17)
Appl. Sci. 2023, 13, 11744 13 of 23

Design and control the multiple models (converting MPC to S-MPC) depending on
several parameters as follows:
j min max
if Pa→b < 0; SOAccl > SOAccl and SOAccl < SOAccl (18)

j max min
if Pa→b > 0; SOAccl > SOAccl and SOAccl < SOAccl (19)

where i = 1, 2, . . . , 11.
Regarding the AR-LSTM formulation, if Equations (2) and (4) are merged, the new
state vector will be as follows:

X(k ) = c + ϕ(1)tanx (Wxx x (k − 1) + Wux u(k )) + ϕ(2)tanx (Wxx x (k − 2)

(20)
+Wux u(k − 1)) + · · · + ϕ(q)tanx (Wxx x (k − NP ) + Wux u(k − NP + 1)) + ε(k)

The objective function of the hybrid MG system using S-ANC (the combination of
Equations (1) and (6)):

NP NP
J (k) = ∑ Wxx (k )(R(k) + X(k))2 + ∑ Wux (k)∆u(k)2
k =0 k =0
(21)
NP NP
+ ∑ Wyy (k)y(k)2 + p2 + ∑ ŷ(k), y(k)
k =0 k =0

0 = F (X(k), y(k), u(k), ŷ(k), R(k), p)

(22)
0 ≤ H (X(k), y(k), u(k), ŷ(k), R(k), p)

The main advantage of employing the formulation presented by Equations (21) and (22)
is that it imposes short-term safety constraints, while allowing for continuous empirical
experience-based learning. In addition, reducing the prediction horizon of the dynamic
optimization problem significantly simplifies the resulting nonlinear program. Notably,
both optimization functions from Equation (21) must be jointly merged, such that the
state X must be related to the expected optimization variables in k + 1. This results in
lower overheads than optimizing with longer prediction horizons that must be discretized
over time.
Notably, domain knowledge is encoded in the controller model F for optimization and
control vectors, providing the algorithm with understandability. Then, the constraints are
implied for the hybrid MG system. The next step is to automatically convert the traditional
MPC into S-MPC. The final steps in the S-MPC are to solve the cost function and obtain
“optimal decision variables”, as shown in Algorithm 1. After that, the hybrid AR-LSTM
method is initiated by configuring the controller model F. The current state X is found
using Equation (20), before training the “optimal control decisions”. Finally, the control
variable U and ŷ are solved by utilizing updated reference R and Equations (21) and (22).
To begin, design a model of the MG system. The system reads some MG specifications,
such as PV and load data, accumulator data, and the maximum values of power flows
among the components of the hybrid MG. Following that, the MPC controller is imple-
mented, which will state the optimization problem and solve it at each time step, to obtain
the optimal control inputs for the next time step. However, the MPC is converted into the
S-MPC before it is applied. The optimization problem should consider the objectives and
constraints given in this paper’s methodology section. Implement an AR-LSTM model and
train it on past data to increase the accuracy of the predictive model utilized by the S-MPC
controller. Based on the present and previous system conditions, the AR-LSTM should
be able to anticipate future MG behavior. The prediction should be input into the S-MPC
controller’s optimization problem. Finally, as indicated in the methodology section of this
Appl. Sci. 2023, 13, 11744 14 of 23

paper, the S-MPC and AR-LSTM controllers in a closed-loop control system are combined.
The proposed method can test the control strategy under various operating situations and
evaluate its performance using the provided performance criteria (cost functions).

Algorithm 1: Switched Auto-regressive Neural Control (S-ANC)

Identify: F ( x (k), r (k ), u(k), y(k), p) = 0
Imply: H ( x (k), r (k), u(k), y(k ), p) ≥ 0
Switching logic: Conversion MPC into S-MPC
Solve: Objective function for S-MPC using Equation (1)
Obtain: “Optimal decision variables”
Configure: F (X(k ), y(k), u(k), ŷ(k), R(k ), p) = 0
while true do
X(k) ← current state using Equation (13)—AR-LSTM
Pretrain “optimal decision variables” using F
U(k) ← control (input) variable
R(k) ← estimate from measurements
ŷ(k) ← Solve Equations (14) and (15)
R(k + 1), U(k + 1) ← Apply ŷ(k) to F
end

More specifically, to implement our proposed method in operation, initially, model

the MG system and the S-MPC and AR-LSTM controllers, and then combine these models
into a closed-loop control system. Here are the detailed steps that need to be followed, as
illustrated in Figure 5:
• Initiate the system specifications and operational conditions from the MG operator;
• Solve the systematic generation of the control problem, employing the MPC with the
QP;
• Using switching logic, automatically convert the MPC into the S-MPC;
• The optimal control decisions are obtained;
• The optimal control decisions are employed as input data for the AR method;
• The data preparation is initiated. This step has several parameters, such as data
cleaning, extracting features, and merging the input data and PV constraints;
• The AR model is implemented to increase the accuracy of our proposed method;
• After that, multivariate time series are employed;
• Then, the training and test data are selected and evaluated;
• To move the LSTM layer to after the RNN, a sequential network of an input LSTM
layer is produced;
• In this step (implementation of LSTM), several parameters are defined, including the
batch size, epoch number, and type of optimizer;
• Before moving the calculation to the model accuracy, the scaling for the forecast and
actual data are inverted;
• The model accuracy is calculated using various methods, along with the mean direc-
tional accuracy, R2 method, and so on;
• Integrate the S-MPC and AR-LSTM controllers into a closed-loop control system by
connecting the RNN output to the MPC controller’s input, and the MPC controller’s
output to the MG system’s input;
• Then, the optimal control decisions and references are updated. In other words, X, U,
and R are re-evaluated depending on the model accuracy;
• If this accuracy is unreasonable, the S-MPC is reapplied using the updated control
decisions.
Appl. Sci. 2023, 13, 11744 15 of 23

Read system specifications and operational conditions from MG operator

Some parameters are defined,

such as accumulators’ charging
and discharging efficiencies and
Implementation the traditional capacities, and power flows.
MPC
𝒋𝒋
𝑷𝑷𝒊𝒊 >0 The MPC solution method takes as input the union
No Yes
of the estimated states 𝑥𝑥(𝑘𝑘)
� and the disturbance
Deficit Excess forecast 𝑑𝑑.
power power Conversion of MPC into S-
MPC

Check the SOAcc for the accumulators

Systematic generation of the control problem using

S-MPC and solving the quadratic problem.
Charge the Discharge the
accumulators accumulators Define the cost function (objective
function) for the MG system

Quadratic programming approach

Data Preparation
Data
Optimal control decision
cleaning
Integrating optimal control decision into the RNN based-EMS
for the current time step.
Extracting
Apply RNN method features

Merging data
Implementation LSTM
Update the optimal control decisions

(control decision, 𝑃𝑃𝑃𝑃𝑃𝑃 ,

𝑃𝑃𝑃𝑃 𝑚𝑚𝑚𝑚𝑚𝑚 and 𝑃𝑃𝐿𝐿𝐿𝐿 data)

Use multivariate time series Auto-

Inverting the scaling for forecast regressive
and actual data model

Calculation the model accuracy Evaluate

Define
batch size Evaluate mean directional train and test
accuracy

Updated the optimal control decision

Specify Create a sequential
Identifying the input data to re- network of an input
Move to the next

epoch number
apply the S-MPC LSTM layer
time step

Re-implement S-MPC method

Select a proper
“optimizer”

Compare the results. Are the

No results more accurate? Yes

Figure 5. Flow chart of the proposed method.

4. Results and Discussion

4.1. Case 1: Implementation of S-MPC
The non-optimal and optimal control (S-MPC) were compared for 96 h (four days)
in this case. In other words, in case 1, the emphasis was on the optimization of S-MPC
and its effect on the power flows of the system (Figure 6a) and the SOC of the battery
(Figure 6b). Non-control methods employ simplistic control strategies or heuristic rules,
disregarding the system’s dynamic nature. In addition, they do not have any constraints,
so there are disadvantages associated with this strategy, such as poor SOC management,
Appl. Sci. 2023, 13, 11744 16 of 23

potential deviations from desired SOC levels, and an inability to adapt to changing system
conditions. As shown in Figure 6b, the SOC of the battery went below the critical value
(20%), since the non-optimal method had no constraints. In contrast, the S-MPC method is
an alternative to the non-optimal control method. S-MPC’s ability to dynamically select
the appropriate model and control strategy based on the system’s current state is one of
its key advantages. S-MPC optimizes control actions to achieve the desired performance
objectives, particularly in effectively managing the SOC of a battery.

(a) (b)
Figure 6. Comparison of (a) power flows and (b) SOC of the battery using the optimal method
(S-MPC) and non-optimal method.

The S-MPC controller was designed to select the optimal model and control strategy
based on the current operating mode and performance objectives. S-MPC enables enhanced
power flow control and EM by effectively adapting to the changing dynamics of the
hybrid MG, utilizing distinct models for each state. Case 1’s implementation of the S-MPC
controller successfully optimized the hybrid MG system’s power flows. It substantially
reduced energy imports and increased energy exports, resulting in a more efficient use of
resources and enhanced energy flow management. The improved control strategy enabled
the MG to operate closer to its optimal performance, enhancing its dependability and
reducing the operational costs. However, it is important to note that developing and
implementing the S-MPC controller for the hybrid MG system presented obstacles due to
the complex switching conditions and multiple operating modes. To guarantee the selection
of the optimal model and control strategy, the switching logic had to be meticulously
designed. The controller’s increased complexity required more computational resources
than traditional MPC methods. The model’s computational time was almost 405 s.

4.2. Case 2: Implementation of the Merged S-MPC and AR

Case 2 investigated the integration of the AR model with S-MPC. AR models accurately
predict future time steps by capturing the time series behavior of the system. Various
analyses, including variations, cross-validation (CV) iteration-time series behavior for
training and validation, CV iteration-training data for each CV iteration, and predictions
ordered by test prediction number were used to evaluate the performance of the AR models.
Figure 7a shows a visualization designed to provide insight into the behavior of the
lagged target feature over time. We can identify patterns, trends, and correlations within
the lagged target data by examining the plot. Understanding the characteristics of the
lagged target can aid in developing and optimizing an AR linear regression model that
uses this characteristic for prediction. By displaying the lagging feature of the target, we
can observe its values across multiple time steps. This lets us determine whether the lagged
target exhibits specific patterns, trends, or variations over time.
Appl. Sci. 2023, 13, 11744 17 of 23

(a)

(b)
Figure 7. Visualization of (a) the behavior of the lagged target feature over time, (b) the adaptability
of the AR models to various patterns and tendencies

As illustrated in Figure 7b, the variations in the AR models illustrate their capacity to
capture and model the system’s complex dynamics. By analyzing the CV iteration-time
series behavior, the adaptability of the AR models to various patterns and tendencies in the
training and validation datasets was assessed. This analysis shed light on how the models
learn and generalize from the available data, enabling accurate predictions for various time
series problems.
Case 2’s successful integration of AR models with S-MPC illustrated the importance
of incorporating time series behavior and forecasting capabilities into a control system.
Combining S-MPC and AR models permitted enhanced adaptation to system dynamics and
improved the prediction accuracy, thereby enhancing the MG’s overall control performance.

4.3. Case 3: Implementation of the S-ANC

Case 3 examined the combination of S-MPC and AR-LSTM models. Integrating these
advanced models aimed to enhance the predictive capabilities of the control system. The
S-ANC predicted the last month of the year using the first eleven months as training data.
Using metrics such as train–test (Figure 8a) and grid consumption prediction with AR
regression and S-ANC (Figure 8b), the performance of this merged approach was evaluated.
This integration (S-MPC and AR-LSTM) increased the precision of forecasting and the
precision of power flow optimization. The enhanced prediction capabilities of the AR-
LSTM models allowed the control system to anticipate future energy requirements and
adjust the operation of the MG accordingly. In Case 3, the S-MPC controller was improved
by combining it with AR and RNN-LSTM models. This integration aimed to improve the
precision of predictions and the overall performance of the control system. Two primary
figures were generated for analysis: a comparison between the training and the test, and a
prediction of grid consumption using AR regression and S-ANC.
Appl. Sci. 2023, 13, 11744 18 of 23

(a)

(b)
Figure 8. Illustration of (a) train–test data and (b) prediction of grid consumption using AR regression
and S-ANC.

The train–test comparison diagram visually represents the AR-LSTM models’ capacity
to generalize effectively for unobserved data. It compares the predicted grid consumption
values during the testing phase with the actual values, indicating the AR-LSTM models’
ability to capture the hybrid MG’s complex patterns and dynamics (Figure 8b). The
diagram depicts the performance of the combined S-MPC, AR, and RNN-LSTM models on
the training and testing datasets (Figure 8a). According to our simulation, the model could
generalize well for new data and the integration strategy’s effectiveness. Moreover, the
prediction of grid consumption using AR regression and S-ANC illustrated the ability of
the combined method to optimize power flows, while accurately predicting future power
demands. By leveraging the predictive capabilities of AR-LSTM models within the S-MPC
framework, the control system could more precisely estimate grid consumption, allowing
for more effective EM and enhancing the MG’s adaptability to load demands and renewable
energy generation fluctuations. The grid consumption forecast graph (Figure 8b) depicts
the projections generated by the AR and S-ANC models. This enables a comparison of
the two methods and highlights the advantages of the S-ANC method, which employs
the AR-LSTM model for accurate predictions, while reducing computational time. The
computational time of the model was reduced by nearly 214 s.

4.4. Calculation of Model Accuracy

To comprehensively evaluate the performance of the S-ANC prediction model, three
unified evaluation indices, including R-squared score, mean absolute error (MAE), and
mean square error (MSE) [76,77] were selected in this paper. The evaluation indexes have
the following mathematical definitions:
R-squared Score:
∑n (yi − ŷi )2
R2 = 1 − in=1 (23)
∑i=1 (yi − ȳ)2
Appl. Sci. 2023, 13, 11744 19 of 23

Mean Absolute Error:

∑in=1 |yi − ŷi |
MAE = (24)
n
Mean Squared Error:
∑in=1 (yi − ŷi )2
MSE = (25)
n
where yi is the output vector using the S-MPC; ŷi represents the predicted value of the
output vector by employing the S-ANC; ȳ represents the average value of output vector; n
represents the total number of samples.
In this study, we assessed the performance of our predictive model using various
accuracy metrics. The R-squared error, which measures the proportion of the variance in
the dependent variable that can be predicted from the independent variables, was 0.951.
This suggests that our model accounted for approximately 95.1% of the data variance,
indicating a strong predictive ability. In addition, we determined that the MAE was 0.571.
The MAE is the mean absolute difference between observed (from the MPC) and predicted
values. A smaller MAE indicates that the predicted and observed values correspond more
closely. In our case, the relatively low MAE indicates that our model’s predictions deviated
from the true values, on average, by approximately 0.571%. Likewise, we determined the
MSE to be 0.571. The MSE measures the average squared deviation between predicted
and observed values. Similarly to the MAE, a lower MSE indicates greater precision. The
MSE value of our model indicated that the squared differences between the predicted and
observed values were, on average, 0.571 units. Overall, the results show that our predictive
model was effective. The relatively low MAE and MSE values of 0.571 indicated precise
predictions, with minimal deviations from the actual values.

5. Conclusions
Our findings show the efficacy and advantages of the S-ANC method for the intelligent
control and management of hybrid MGs. The optimization of S-MPC improves energy
management and power flow control, resulting in more efficient use of resources. The
integration of AR and RNN-LSTM models improves the accuracy of predictions, allowing
the control system to adapt to dynamic system conditions and to optimize the operation of
the MG. The successful implementation of S-ANC significantly affects the dependability,
sustainability, and cost-effectiveness of hybrid MG systems. We can achieve efficient
control and management of complex energy systems by leveraging the capabilities of
advanced modeling techniques within the S-MPC framework. These findings support
the incorporation of hybrid MGs in future energy systems and contribute to developing
intelligent control strategies. By combining the AR-LSTM, the computational time of the
model was reduced by approximately 47.2%. In addition, this study assessed the accuracy
of our predictive model. The R-squared error, which quantifies the amount of variance in the
dependent variable that can be predicted from the independent variables, was 0.951. Our
model predicted 95.1% of the variance in the data, indicating a high level of predictive ability.
The MAE and MSE values of 0.571 indicated precise forecasts, with minimal deviations from
the actual values. The focus of future research and development should be on validating
larger-scale systems and incorporating additional advanced models. These developments
will enhance the performance and applicability of the S-ANC methodology and contribute
to the efficient operation and integration of hybrid MGs in future energy systems.

Author Contributions: Conceptualization, M.C. and Y.F.U.; methodology, M.C., A.A. and D.G.;
software, M.C. and A.A.; validation, M.C., H.A. and A.A.; investigation, M.C.; writing—original draft
preparation, M.C. and Y.F.U.; writing—review and editing, M.C., Y.F.U., H.A., A.A., K.A. and D.G.;
visualization, M.C., Y.F.U. and H.A.; supervision, A.A., K.A. and D.G. All authors have read and
agreed to the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: Not applicable.
Appl. Sci. 2023, 13, 11744 20 of 23

Informed Consent Statement: Not applicable.

Data Availability Statement: Not applicable.
Conflicts of Interest: The authors declare no conflict of interest.

Abbreviations
The following abbreviations are used in this manuscript:

ANN Artificial Neural Network

AR Auto-regressive
AR-LSTM Auto-regressive Long Short-Term Memory
ARIMA Auto-regressive Integrated Moving Average
ARMA Auto-regressive Moving Average
BAT Battery
BPTT Back-Propagation Through Time
CNN Convolutional Neural Network
CV Cross-validation
DLC Direct Load Control
EL Electrolyzer
EM Energy Management
ESS Energy Storage System
FC Fuel Cell
FT Fuel Tank
GHI Global Horizontal Irradiance
GR Grid
GRU Gated Recurrent Unit
IoT Internet of Things
LSTM Long Short-Term Memory
MAE Mean Absolute Error
MG Microgrid
MSE Mean Squared Error
MILP Mixed Integer Linear Programming
ML Machine Learning
MPC Model Predictive Control
NARX Nonlinear Auto-regressive with exogenous input
NMG Networked Microgrid
RNN Recurrent Neural Network
RES Renewable Energy Source
S-ANC Switched Auto-regressive Neural Control
S-MPC Switched Model Predictive Control
SVM Support Vector Machine
QP Quadratic Programming
WT Water Tank
l
ηch Charging efficiency of accumulator l
l
ηdis Discharging efficiency of accumulator l
F controller model
H constraint
J objective function
Pmmax Maximum values of power flows, 5 kW
PV Photovoltaic
PVGR or P1 Power flow from PV to grid
PVLD or P2 Power flow from PV to load
GR LD or P3 Power flow from grid to load
PVBAT or P4 Power flow from PV to battery
BATLD or P5 Power flow from battery to load
FCBAT or P6 Power flow from fuel cell to battery
BATEL or P7 Power flow from battery to electrolyzer
EL FT or P8 Hydrogen flow from electrolyzer to fuel tank
Appl. Sci. 2023, 13, 11744 21 of 23

FTFC or P9 Hydrogen flow from fuel tank to fuel cell

FCWT or P10 Water flow from fuel cell to water tank
WTEL or P11 Water flow from water tank to electrolyzer
j
Fa→b (k) Flow of j from node a to node b
Cl Capacities of accumulator l, [kWh]
j
Pa→b Power of j from node a to node b
ϕ Auto-regressive model coefficient
Np Prediction horizon, 24 h
SOAccl State of accumulator l
SOAcclmax Maximum value state of accumulator l
SOAcclmin Minimum value state of accumulator l
ε(k) error term or random noise at time k

References
1. Kumar, A.S.; Ahmad, Z. Model predictive control (MPC) and its current issues in chemical engineering. Chem. Eng. Commun.
2012, 199, 472–511.
2. Garcia, C.E.; Prett, D.M.; Morari, M. Model predictive control: Theory and practice—A survey. Automatica 1989, 25, 335–348.
3. Pamulapati, T.; Cavus, M.; Odigwe, I.; Allahham, A.; Walker, S.; Giaouris, D. A Review of Microgrid Energy Management
Strategies from the Energy Trilemma Perspective. Energies 2022, 16, 289.
4. Khawaja, Y.; Qiqieh, I.; Alzubi, J.; Alzubi, O.; Allahham, A.; Giaouris, D. Design of cost-based sizing and energy management
framework for standalone microgrid using reinforcement learning. Sol. Energy 2023, 251, 249–260.
5. Nikkhah, S.; Allahham, A.; Bialek, J.W.; Walker, S.L.; Giaouris, D.; Papadopoulou, S. Active participation of buildings in the
energy networks: Dynamic/operational models and control challenges. Energies 2021, 14, 7220. [CrossRef]
6. Parisio, A.; Rikos, E.; Glielmo, L. A model predictive control approach to microgrid operation optimization. IEEE Trans. Control.
Syst. Technol. 2014, 22, 1813–1827.
7. Nikkhah, S.; Allahham, A.; Royapoor, M.; Bialek, J.W.; Giaouris, D. A Community-Based Building-to-Building Strategy for
Multi-Objective Energy Management of Residential Microgrids. In Proceedings of the 2021 12th International Renewable
Engineering Conference (IREC), Amman, Jordan, 14–15 April 2021; pp. 1–6.
8. Allahham, A.; Greenwood, D.; Patsios, C.; Taylor, P. Adaptive receding horizon control for battery energy storage management
with age-and-operation-dependent efficiency and degradation. Electr. Power Syst. Res. 2022, 209, 107936.
9. Nikkhah, S.; Sarantakos, I.; Zografou-Barredo, N.M.; Rabiee, A.; Allahham, A.; Giaouris, D. A joint risk-and security-constrained
control framework for real-time energy scheduling of islanded microgrids. IEEE Trans. Smart Grid 2022, 13, 3354–3368.
10. Ulutas, A.; Altas, I.H.; Onen, A.; Ustun, T.S. Neuro-fuzzy-based model predictive energy management for grid connected
microgrids. Electronics 2020, 9, 900.
11. Silvente, J.; Kopanos, G.M.; Dua, V.; Papageorgiou, L.G. A rolling horizon approach for optimal management of microgrids under
stochastic uncertainty. Chem. Eng. Res. Des. 2018, 131, 293–317.
12. Parisio, A.; Rikos, E.; Glielmo, L. Stochastic model predictive control for economic/environmental operation management of
microgrids: An experimental case study. J. Process Control 2016, 43, 24–37. [CrossRef]
13. Garcia-Torres, F.; Bordons, C. Optimal economical schedule of hydrogen-based microgrids with hybrid storage using model
predictive control. IEEE Trans. Ind. Electron. 2015, 62, 5195–5207. [CrossRef]
14. Jayachandran, M.; Ravi, G. Decentralized model predictive hierarchical control strategy for islanded AC microgrids. Electr. Power
Syst. Res. 2019, 170, 92–100. [CrossRef]
15. Cavus, M.; Allahham, A.; Adhikari, K.; Zangiabadia, M.; Giaouris, D. Control of microgrids using an enhanced Model Predictive
Controller. In Proceedings of the 11th International Conference on Power Electronics, Machines and Drives (PEMD 2022),
Newcastle, UK, 21–23 June 2022.
16. Cavus, M.; Allahham, A.; Adhikari, K.; Zangiabadi, M.; Giaouris, D. Energy Management of Grid-Connected Microgrids using
an Optimal Systems Approach. IEEE Access 2023, 11, 9907–9919. [CrossRef]
17. Zhu, B.; Tazvinga, H.; Xia, X. Switched model predictive control for energy dispatching of a photovoltaic-diesel-battery hybrid
power system. IEEE Trans. Control. Syst. Technol. 2014, 23, 1229–1236.
18. Maślak, G.; Orłowski, P. Microgrid operation optimization using hybrid system modeling and switched model predictive control.
Energies 2022, 15, 833. [CrossRef]
19. Moness, M.; Moustafa, A.M. Real-time switched model predictive control for a cyber-physical wind turbine emulator. IEEE Trans.
Ind. Inform. 2019, 16, 3807–3817. [CrossRef]
20. Pervez, M.; Kamal, T.; Fernández-Ramírez, L.M. A novel switched model predictive control of wind turbines using artificial
neural network-Markov chains prediction with load mitigation. Ain Shams Eng. J. 2022, 13, 101577. [CrossRef]
21. Magni, L.; Scattolini, R.; Tanelli, M. Switched model predictive control for performance enhancement. Int. J. Control 2008,
81, 1859–1869. [CrossRef]
22. Aguilera, R.P.; Lezana, P.; Quevedo, D.E. Switched model predictive control for improved transient and steady-state performance.
IEEE Trans. Ind. Inform. 2015, 11, 968–977. [CrossRef]
Appl. Sci. 2023, 13, 11744 22 of 23

23. Kwadzogah, R.; Zhou, M.; Li, S. Model predictive control for HVAC systems—A review. In Proceedings of the 2013 IEEE
International Conference on Automation Science and Engineering (CASE), Madison, WI, USA, 17–20 August 2013; pp. 442–447.
24. Forbes, M.G.; Patwardhan, R.S.; Hamadah, H.; Gopaluni, R.B. Model predictive control in industry: Challenges and opportunities.
IFAC-PapersOnLine 2015, 48, 531–538. [CrossRef]
25. Teimourzadeh, S.; Tor, O.B.; Cebeci, M.E.; Bara, A.; Oprea, S.V. A three-stage approach for resilience-constrained scheduling of
networked microgrids. J. Mod. Power Syst. Clean Energy 2019, 7, 705–715. [CrossRef]
26. Giaouris, D.; Papadopoulos, A.I.; Seferlis, P.; Papadopoulou, S.; Voutetakis, S.; Stergiopoulos, F.; Elmasides, C. Optimum energy
management in smart grids based on power pinch analysis. Chem. Eng. 2014, 39, 55–60.
27. Giaouris, D.; Papadopoulos, A.I.; Voutetakis, S.; Papadopoulou, S.; Seferlis, P. A power grand composite curves approach for
analysis and adaptive operation of renewable energy smart grids. Clean Technol. Environ. Policy 2015, 17, 1171–1193. [CrossRef]
28. Khanna, S.; Becerra, V.; Allahham, A.; Giaouris, D.; Foster, J.M.; Roberts, K.; Hutchinson, D.; Fawcett, J. Demand response
model development for smart households using time of use tariffs and optimal control—The isle of wight energy autonomous
community case study. Energies 2020, 13, 541. [CrossRef]
29. Oprea, S.V.; Bâra, A. Edge and fog computing using IoT for direct load optimization and control with flexibility services for
citizen energy communities. Knowl.-Based Syst. 2021, 228, 107293. [CrossRef]
30. Bodong, S.; Wiseong, J.; Chengmeng, L.; Khakichi, A. Economic management and planning based on a probabilistic model in a
multi-energy market in the presence of renewable energy sources with a demand-side management program. Energy 2023, 269,
126549. [CrossRef]
31. Wynn, S.L.L.; Boonraksa, T.; Boonraksa, P.; Pinthurat, W.; Marungsri, B. Decentralized Energy Management System in Microgrid
Considering Uncertainty and Demand Response. Electronics 2023, 12, 237. [CrossRef]
32. Sansa, I.; Boussaada, Z.; Bellaaj, N.M. Solar Radiation Prediction Using a Novel Hybrid Model of ARMA and NARX. Energies
2021, 14, 6920. [CrossRef]
33. Brahma, B.; Wadhvani, R. Solar irradiance forecasting based on deep learning methodologies and multi-site data. Symmetry 2020,
12, 1830. [CrossRef]
34. Jeon, B.k.; Kim, E.J. Next-day prediction of hourly solar irradiance using local weather forecasts and LSTM trained with non-local
data. Energies 2020, 13, 5258. [CrossRef]
35. Husein, M.; Chung, I.Y. Day-ahead solar irradiance forecasting for microgrids using a long short-term memory recurrent neural
network: A deep learning approach. Energies 2019, 12, 1856. [CrossRef]
36. Zafar, R.; Vu, B.H.; Husein, M.; Chung, I.Y. Day-Ahead Solar Irradiance Forecasting Using Hybrid Recurrent Neural Network
with Weather Classification for Power System Scheduling. Appl. Sci. 2021, 11, 6738. [CrossRef]
37. Qing, X.; Niu, Y. Hourly day-ahead solar irradiance prediction using weather forecasts by LSTM. Energy 2018, 148, 461–468.
[CrossRef]
38. Srivastava, S.; Lessmann, S. A comparative study of LSTM neural networks in forecasting day-ahead global horizontal irradiance
with satellite data. Sol. Energy 2018, 162, 232–247. [CrossRef]
39. Wang, F.; Yu, Y.; Zhang, Z.; Li, J.; Zhen, Z.; Li, K. Wavelet decomposition and convolutional LSTM networks based improved
deep learning model for solar irradiance forecasting. Appl. Sci. 2018, 8, 1286. [CrossRef]
40. Ghimire, S.; Deo, R.C.; Raj, N.; Mi, J. Deep solar radiation forecasting with convolutional neural network and long short-term
memory network algorithms. Appl. Energy 2019, 253, 113541. [CrossRef]
41. Huang, J.; Korolkiewicz, M.; Agrawal, M.; Boland, J. Forecasting solar radiation on an hourly time scale using a Coupled
AutoRegressive and Dynamical System (CARDS) model. Sol. Energy 2013, 87, 136–149. [CrossRef]
42. Jiang, Y.; Long, H.; Zhang, Z.; Song, Z. Day-ahead prediction of bihourly solar radiance with a Markov switch approach. IEEE
Trans. Sustain. Energy 2017, 8, 1536–1547. [CrossRef]
43. Ekici, B.B. A least squares support vector machine model for prediction of the next day solar insolation for effective use of PV
systems. Measurement 2014, 50, 255–262. [CrossRef]
44. Yu, Y.; Cao, J.; Zhu, J. An LSTM short-term solar irradiance forecasting under complicated weather conditions. IEEE Access 2019,
7, 145651–145666. [CrossRef]
45. Wojtkiewicz, J.; Hosseini, M.; Gottumukkala, R.; Chambers, T.L. Hour-ahead solar irradiance forecasting using multivariate gated
recurrent units. Energies 2019, 12, 4055. [CrossRef]
46. Zang, H.; Liu, L.; Sun, L.; Cheng, L.; Wei, Z.; Sun, G. Short-term global horizontal irradiance forecasting based on a hybrid
CNN-LSTM model with spatiotemporal correlations. Renew. Energy 2020, 160, 26–41. [CrossRef]
47. Gao, B.; Huang, X.; Shi, J.; Tai, Y.; Zhang, J. Hourly forecasting of solar irradiance based on CEEMDAN and multi-strategy
CNN-LSTM neural networks. Renew. Energy 2020, 162, 1665–1683. [CrossRef]
48. Connor, J.; Atlas, L. Recurrent neural networks and time series prediction. In Proceedings of the IJCNN-91-Seattle international
joint conference on neural networks, Seattle, WA, USA, 8–12 July 1991; Volume 1, pp. 301–306.
49. Brownlee, J. Time series prediction with lstm recurrent neural networks in python with keras. Mach. Learn. Mastery 2016, 18.
50. Moreno, J.J.M. Artificial neural networks applied to forecasting time series. Psicothema 2011, 23, 322–329.
51. Shen, Z.; Zhang, Y.; Lu, J.; Xu, J.; Xiao, G. A novel time series forecasting model with deep learning. Neurocomputing 2020,
396, 302–313. [CrossRef]
Appl. Sci. 2023, 13, 11744 23 of 23

52. Bianchi, F.M.; Maiorino, E.; Kampffmeyer, M.C.; Rizzi, A.; Jenssen, R. Recurrent Neural Networks for Short-Term Load Forecasting:
An Overview and Comparative Analysis; Springer: Berlin/Heidelberg, Germany, 2017.
53. Kumar, D.; Mathur, H.; Bhanot, S.; Bansal, R.C. Forecasting of solar and wind power using LSTM RNN for load frequency control
in isolated microgrid. Int. J. Model. Simul. 2021, 41, 311–323. [CrossRef]
54. Li, D.; Tan, Y.; Zhang, Y.; Miao, S.; He, S. Probabilistic forecasting method for mid-term hourly load time series based on an
improved temporal fusion transformer model. Int. J. Electr. Power Energy Syst. 2023, 146, 108743. [CrossRef]
55. DiPietro, R.; Hager, G.D. Deep learning: RNNs and LSTM. In Handbook of Medical Image Computing and Computer Assisted
Intervention; Elsevier: Amsterdam, The Netherlands, 2020; pp. 503–519.
56. Gupta, A.; Gurrala, G.; Sastry, P.S. Instability Prediction in Power Systems using Recurrent Neural Networks. In Proceedings of
the IJCAI, Melbourne, Australia, 19–25 August 2017; pp. 1795–1801.
57. Huo, Y.; Chen, Z.; Bu, J.; Yin, M. Learning assisted column generation for model predictive control based energy management in
microgrids. Energy Rep. 2023, 9, 88–97. [CrossRef]
58. Cabrera-Tobar, A.; Massi Pavan, A.; Petrone, G.; Spagnuolo, G. A Review of the Optimization and Control Techniques in the
Presence of Uncertainties for the Energy Management of Microgrids. Energies 2022, 15, 9114. [CrossRef]
59. Zhou, Y. Advances of machine learning in multi-energy district communities–mechanisms, applications and perspectives. Energy
AI 2022, 10, 100187. [CrossRef]
60. Li, B.; Roche, R. Optimal scheduling of multiple multi-energy supply microgrids considering future prediction impacts based on
model predictive control. Energy 2020, 197, 117180. [CrossRef]
61. Nyong-Bassey, B.E.; Giaouris, D.; Patsios, C.; Papadopoulou, S.; Papadopoulos, A.I.; Walker, S.; Voutetakis, S.; Seferlis, P.; Gadoue,
S. Reinforcement learning based adaptive power pinch analysis for energy management of stand-alone hybrid energy storage
systems considering uncertainty. Energy 2020, 193, 116622. [CrossRef]
62. Tang, W.; Zhang, Y.J. A model predictive control approach for low-complexity electric vehicle charging scheduling: Optimality
and scalability. IEEE Trans. Power Syst. 2016, 32, 1050–1063. [CrossRef]
63. Karamanakos, P.; Geyer, T.; Kennel, R. A computationally efficient model predictive control strategy for linear systems with
integer inputs. IEEE Trans. Control Syst. Technol. 2015, 24, 1463–1471. [CrossRef]
64. Zhang, L.; Zhuang, S.; Braatz, R.D. Switched model predictive control of switched linear systems: Feasibility, stability and
robustness. Automatica 2016, 67, 8–21. [CrossRef]
65. Ayumi, V.; Rere, L.R.; Fanany, M.I.; Arymurthy, A.M. Optimization of convolutional neural network using microcanonical
annealing algorithm. In Proceedings of the 2016 International Conference on Advanced Computer Science and Information
Systems (ICACSIS), Malang, Indonesia, 15–16 October 2016; pp. 506–511.
66. Liu, C.T.; Wu, Y.H.; Lin, Y.S.; Chien, S.Y. Computation-performance optimization of convolutional neural networks with
redundant kernel removal. In Proceedings of the 2018 IEEE International Symposium on Circuits and Systems (ISCAS), Florence,
Italy, 27–30 May 2018; pp. 1–5.
67. Huang, R.; Wei, C.; Wang, B.; Yang, J.; Xu, X.; Wu, S.; Huang, S. Well performance prediction based on Long Short-Term Memory
(LSTM) neural network. J. Pet. Sci. Eng. 2022, 208, 109686. [CrossRef]
68. Zhang, Y.; Hao, X.; Liu, Y. Simplifying long short-term memory for fast training and time series prediction. J. Phys. Conf. Ser.
2019, 1213, 042039. [CrossRef]
69. Schmidt, R.M. Recurrent neural networks (rnns): A gentle introduction and overview. arXiv 2019, arXiv:1912.05911.
70. Amidi, A.; Amidi, S. Vip Cheatsheet: Recurrent Neural Networks; Stanford University: Stanford, CA, USA, 2018.
71. Ruder, S. An overview of gradient descent optimization algorithms. arXiv 2016, arXiv:1609.04747.
72. Görges, D. Relations between model predictive control and reinforcement learning. IFAC-PapersOnLine 2017, 50, 4920–4928.
[CrossRef]
73. Ying, X. An overview of overfitting and its solutions. J. Phys. Conf. Ser. 2019, 1168, 022022. [CrossRef]
74. Giaouris, D.; Papadopoulos, A.I.; Patsios, C.; Walker, S.; Ziogou, C.; Taylor, P.; Voutetakis, S.; Papadopoulou, S.; Seferlis, P. A
systems approach for management of microgrids considering multiple energy carriers, stochastic loads, forecasting and demand
side response. Appl. Energy 2018, 226, 546–559. [CrossRef]
75. Cavus, M.; Allahham, A.; Adhikari, K.; Giaouris, D. A Hybrid Method Based on Logic Control and Model Predictive Control for
Synthesizing Controller for Flexible Hybrid Microgrid with Plug-and-Play Capabilities. Available online: https://ssrn.com/
abstract=4473008 (accessed on 15 July 2023).
76. Chicco, D.; Warrens, M.J.; Jurman, G. The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE,
MSE and RMSE in regression analysis evaluation. PeerJ Comput. Sci. 2021, 7, e623. [CrossRef]
77. Kong, X.; Du, X.; Xu, Z.; Xue, G. Predicting solar radiation for space heating with thermal storage system based on temporal
convolutional network-attention model. Appl. Therm. Eng. 2023, 219, 119574. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.