Search Results (190)

The cooperative optimization and dispatch operation of the integrated energy system (IES) depends on accurate load forecasts. A multivariate load, joint prediction model, based on the combination of multi-task learning (MTL) and dynamic time warping (DTW), is proposed to address the issue of the prediction model’s limited accuracy caused by the fragmentation of the multivariate load coupling relationship and the absence of future time series information. Firstly, the MTL model, based on the bidirectional long short-term memory (BiLSTM) neural network, extracts the coupling information among the multivariate loads and performs the preliminary prediction; secondly, the DTW algorithm clusters and splices the load data that are similar to the target value as the input features of the model; finally, the BiLSTM-attention model is used for secondary prediction, and the improved Bayesian optimization algorithm is applied for adaptive selection of optimal hyperparameters. Based on the game-theoretic view of Shapley’s additive interpretation (SHAP), a model interpretation technique is introduced to determine the validity of the liquidity indicator and the asynchronous relationship between the significance of the indicator and its actual contribution. The prediction results show that the joint prediction model proposed in this paper has higher training speed and prediction accuracy than the traditional single-load prediction model. Full article

The convergence of data mining and deep learning has become an invaluable tool for gaining insights into evolving events and trends. However, a persistent challenge in utilizing these techniques for forecasting lies in the limited access to comprehensive, error-free data. This challenge is particularly pronounced in financial time series datasets, which are known for their volatility. To address this issue, a novel approach to data augmentation has been introduced, specifically tailored for financial time series forecasting. This approach leverages the power of Generative Adversarial Networks to generate synthetic data that replicate the distribution of authentic data. By integrating synthetic data with real data, the proposed approach significantly improves forecasting accuracy. Tests with real datasets have proven that this method offers a marked improvement over models that rely only on real data. Full article

Chlorophyll prediction facilitates the comprehension of red tide characteristics and enables early warning. In practice, it is formulated as a multivariate time series forecasting problem aimed at forecasting future chlorophyll concentrations by considering both exogenous factors and chlorophyll. However, the multi-step prediction of chlorophyll concentration poses a formidable challenge due to the intricate interaction between factors and the long temporal dependence between input sequences. In this work, we propose a Multi-attention Recurrent Neural Network (MaRNN) for the multi-step prediction of chlorophyll concentration. The MaRNN comprises an encoder incorporating two-stage spatial attention and a decoder employing temporal attention. The encoder first learns the significance of exogenous factors for prediction in the first phase, and subsequently captures the spatial correlation between the exogenous sequence and chlorophyll sequence in the second phase. The decoder further excavates input sequences that exhibit a strong correlation with the task through temporal attention module, thereby enhancing the prediction accuracy of the model. Experiments conducted on two real-world datasets reveal that MaRNN not only surpasses state-of-the-art methods in performance, but also offers interpretability for chlorophyll prediction. Full article

Accurate prediction of electricity consumption is crucial for energy management and allocation. This study introduces a novel approach, named Attention-based Multivariate Temporal Convolutional Network (AMTCN), for electricity consumption forecasting by integrating attention mechanisms with multivariate temporal convolutional networks. The method involves feature extraction from diverse time series of different feature variables using dilated convolutional networks. Subsequently, attention mechanisms are employed to capture the correlation and contextually important information among various features, thereby enhancing the model’s predictive accuracy. The AMTCN method exhibits universality, making it applicable to various prediction tasks in different scenarios. Experimental evaluations are conducted on four distinct datasets, encompassing electricity consumption and weather temperature aspects. Comparative experiments with LSTM, ConvLSTM, GRU, and TCN—widely-used deep learning methods—demonstrate that our AMTCN model achieves significant improvements of 57% in MSE, 37% in MAE, 35% in RRSE, and 12% in CORR metrics, respectively. This research contributes a promising approach to accurate electricity consumption prediction, leveraging the synergy of attention mechanisms and multivariate temporal convolutional networks, with broad applicability in diverse forecasting scenarios. Full article

This study uses real-world illustrations to explore the application of deep learning approaches to predict economic information. In this, we investigate the effect of deep learning model architecture and time-series data properties on prediction accuracy. We aim to evaluate the predictive power of several neural network models using a financial time-series dataset. These models include Convolutional RNNs, Convolutional LSTMs, Convolutional GRUs, Convolutional Bi-directional RNNs, Convolutional Bi-directional LSTMs, and Convolutional Bi-directional GRUs. Our main objective is to utilize deep learning techniques for simultaneous predictions on multivariable time-series datasets. We utilize the daily fluctuations of six Asian stock market indices from 1 April 2020 to 31 March 2024. This study’s overarching goal is to evaluate deep learning models constructed using training data gathered during the early stages of the COVID-19 pandemic when the economy was hit hard. We find that the limitations prove that no single deep learning algorithm can reliably forecast financial data for every state. In addition, predictions obtained from solitary deep learning models are more precise when dealing with consistent time-series data. Nevertheless, the hybrid model performs better when analyzing time-series data with significant chaos. Full article

This study explores innovative methodologies for estimating the energy baseline (EnBL) of a university classroom building, emphasizing the critical roles of data quality and model selection in achieving accurate energy efficiency assessments. We compare time series models that are suitable for buildings with limited consumption data with univariate and multivariate regression models that incorporate additional variables, such as weather and occupancy. Furthermore, we investigate the advantages of dynamic simulation using the EnergyPlus engine (V5, USDOE United States) and Design Builder software v7, enabling scenario analysis for various operational conditions. Through a comprehensive case study at the UAO University Campus, we validate our models using daily monitoring data and statistical analysis in RStudio. Our findings reveal that model choice significantly influences energy consumption forecasts, leading to potential overestimations or underestimations of savings. By rigorously assessing statistical validation and error analysis results, we highlight the implications for decarbonization strategies in building design and operation. This research provides a valuable framework for selecting appropriate methodologies for energy baseline estimation, enhancing transparency and reliability in energy performance assessments. These contributions are particularly relevant for optimizing energy use and aligning with regulatory requirements in the pursuit of sustainable building practices. Full article

Demographic aging results in a growing number of older people in need of care in many regions all over the world. Germany has witnessed steady population aging for decades, prompting policymakers and other stakeholders to discuss how to fulfill the rapidly growing demand for care workers and finance the rising costs of long-term care. Informed decisions on this matter to ensure the sustainability of the statutory long-term care insurance system require reliable knowledge of the associated future costs. These need to be simulated based on well-designed forecast models that holistically include the complexity of the forecast problem, namely the demographic transition, epidemiological trends, concrete demand for and supply of specific care services, and the respective costs. Care risks heavily depend on demographics, both in absolute terms and according to severity. The number of persons in need of care, disaggregated by severity of disability, in turn, is the main driver of the remuneration that is paid by long-term care insurance. Therefore, detailed forecasts of the population and care rates are important ingredients for forecasts of long-term care insurance expenditures. We present a novel approach based on a stochastic demographic cohort-component approach that includes trends in age- and sex-specific care rates and the demand for specific care services, given changing preferences over the life course. The model is executed for Germany until the year 2050 as a case study. Full article

In the era of big data, prediction has become a fundamental capability. Current prediction methods primarily focus on sequence elements; however, in multivariate time series forecasting, time is a critical factor that must not be overlooked. While some methods consider time, they often neglect the temporal distance between sequence elements and the predicted target time, a relationship essential for identifying patterns such as periodicity, trends, and other temporal dynamics. Moreover, the extraction of temporal features is often inadequate, and discussions on how to comprehensively leverage temporal data are limited. As a result, model performance can suffer, particularly in prediction tasks with specific time requirements. To address these challenges, we propose a new model, TE-LSTM, based on LSTM, which employs a temporal encoding method to fully extract temporal features. A temporal weighting strategy is also used to optimize the integration of temporal information, capturing the temporal relationship of each element relative to the target element, and integrating it into the LSTM. Additionally, this study examines the impact of different time granularities on the model. Using the Beijing International Airport station as the study area, we applied our method to temperature prediction. Compared to the baseline model, our model showed an improvement of 0.7552% without time granularity, 1.2047% with a time granularity of 3, and 0.0953% when addressing prediction tasks with specific time requirements. The final results demonstrate the superiority of the proposed method and highlight its effectiveness in overcoming the limitations of existing approaches. Full article

The development of accurate models to forecast load demand across different time horizons is challenging due to demand patterns and endogenous variables that affect short-term and long-term demand. This paper presents two contributions. First, it addresses the problem of the accuracy of the probabilistic forecasting model for short-term time series where endogenous variables interfere by emphasizing a low computational cost and efficient approach such as Granular Weighted Multivariate Fuzzy Time Series (GranularWMFTS) based on the fuzzy information granules method and a univariate form named Probabilistic Fuzzy Time Series. Secondly, it compares time series forecasting models based on algorithms such as Holt-Winters, Auto-Regressive Integrated Moving Average, High Order Fuzzy Time Series, Weighted High Order Fuzzy Time Series, and Multivariate Fuzzy Time Series (MVFTS) where this paper is based on Root Mean Squared Error, Symmetric Mean Absolute Percentage Error, and Theil’s U Statistic criteria relying on 5% error criteria. Finally, it presents the concept and nuances of the forecasting approaches evaluated, highlighting the differences between fuzzy algorithms in terms of fuzzy logical relationship, fuzzy logical relationship group, and fuzzification in the training phase. Overall, the GranularWMVFTS and weighted MVFTS outperformed other evaluated forecasting approaches regarding the performance criteria adopted with a low computational cost. Full article

With the global objectives of achieving a “carbon peak” and “carbon neutrality” along with the implementation of carbon reduction policies, China’s industrial structure has undergone significant adjustments, resulting in constraints on high-energy consumption and high-emission industries while promoting the rapid growth of green industries. Consequently, these changes have led to an increasingly complex power system structure and presented new challenges for electricity demand forecasting. To address this issue, this study proposes a 24-step multivariate time series short-term load forecasting algorithm model based on KNN data imputation and BiTCN bidirectional temporal convolutional networks combined with BiGRU bidirectional gated recurrent units and attention mechanism. The Kepler adaptive optimization algorithm (KOA) is employed for hyperparameter optimization to effectively enhance prediction accuracy. Furthermore, using real load data from a wind farm in Xinjiang as an example, this paper predicts the electricity load from 1 January to 30 December in 2019. Experimental results demonstrate that our comprehensive short-term load forecasting model exhibits lower prediction errors and superior performance compared to traditional methods, thus holding great value for practical applications. Full article

Solar flares are significant occurrences in solar physics, impacting space weather and terrestrial technologies. Accurate classification of solar flares is essential for predicting space weather and minimizing potential disruptions to communication, navigation, and power systems. This study addresses the challenge of selecting the most relevant features from multivariate time-series data, specifically focusing on solar flares. We employ methods such as Mutual Information (MI), Minimum Redundancy Maximum Relevance (mRMR), and Euclidean Distance to identify key features for classification. Recognizing the performance variability of different feature selection techniques, we introduce an ensemble approach to compute feature weights. By combining outputs from multiple methods, our ensemble method provides a more comprehensive understanding of the importance of features. Our results show that the ensemble approach significantly improves classification performance, achieving values 0.15 higher in True Skill Statistic (TSS) values compared to individual feature selection methods. Additionally, our method offers valuable insights into the underlying physical processes of solar flares, leading to more effective space weather forecasting and enhanced mitigation strategies for communication, navigation, and power system disruptions. Full article

Accurate forecasting of high-resolution particulate matter 2.5 (PM_2.5) levels is essential for the development of public health policy. However, datasets used for this purpose often contain missing observations. This study presents a two-stage approach to handle this problem. The first stage is a multivariate spatial time series (MSTS) model, used to generate forecasts for the sampled spatial units and to impute missing observations. The MSTS model utilizes the similarities between the temporal patterns of the time series of the spatial units to impute the missing data across space. The second stage is the high-resolution prediction model, which generates predictions that cover the entire study domain. The second stage faces the big N problem giving rise to complex memory and computational problems. As a solution to the big N problem, we propose a Gaussian Markov random field (GMRF) for innovations with the Matérn covariance matrix obtained from the corresponding Gaussian field (GF) matrix by means of the stochastic partial differential equation (SPDE) method and the finite element method (FEM). For inference, we propose Bayesian statistics and integrated nested Laplace approximation (INLA) in the R-INLA package. The above approach is demonstrated using daily data collected from 13 PM_2.5 monitoring stations in Jakarta Province, Indonesia, for 1 January–31 December 2022. The first stage of the model generates PM_2.5 forecasts for the 13 monitoring stations for the period 1–31 January 2023, imputing missing data by means of the MSTS model. To capture temporal trends in the PM_2.5 concentrations, the model applies a first-order autoregressive process and a seasonal process. The second stage involves creating a high-resolution map for the period 1–31 January 2023, for sampled and non-sampled spatiotemporal units. It uses the MSTS-generated PM_2.5 predictions for the sampled spatiotemporal units and observations of the covariate’s altitude, population density, and rainfall for sampled and non-samples spatiotemporal units. For the spatially correlated random effects, we apply a first-order random walk process. The validation of out-of-sample forecasts indicates a strong model fit with low mean squared error (0.001), mean absolute error (0.037), and mean absolute percentage error (0.041), and a high R² value (0.855). The analysis reveals that altitude and precipitation negatively impact PM_2.5 concentrations, while population density has a positive effect. Specifically, a one-meter increase in altitude is linked to a 7.8% decrease in PM_2.5, while a one-person increase in population density leads to a 7.0% rise in PM_2.5. Additionally, a one-millimeter increase in rainfall corresponds to a 3.9% decrease in PM_2.5. The paper makes a valuable contribution to the field of forecasting high-resolution PM_2.5 levels, which is essential for providing detailed, accurate information for public health policy. The approach presents a new and innovative method for addressing the problem of missing data and high-resolution forecasting. Full article

Short-term load forecasting is a challenging research problem and has a tremendous impact on electricity generation, transmission, and distribution. A robust forecasting algorithm can help power system operators to better tackle the ever-changing electric power demand. This paper presents a novel deep neural network for short-term electric load forecasting for the St. John’s campus of Memorial University of Newfoundland (MUN). The electric load data are obtained from the Memorial University of Newfoundland and combined with metrological data from St. John’s. This dataset is used to formulate a multivariate time-series forecasting problem. A novel deep learning algorithm is presented, consisting of a 1D Convolutional Neural Network, which is followed by an encoder–decoder-based network with attention. The input used for this model is the electric load consumption and metrological data, while the output is the hourly prediction of the next day. The model is compared with Gated Recurrent Unit (GRU) and Long Short Term Memory (LSTM)-based Recurrent Neural Network. A CNN-based encoder–decoder model without attention is also tested. The proposed model shows a lower mean absolute error (MAE), mean squared error (MSE), mean absolute percentage error (MAPE), and higher R² score. These evaluation metrics show an improved performance compared to GRU and LSTM-based RNNs as well as the CNN encoder–decoder model without attention. Full article

This paper proposes a three (3)-step methodology to forecast the future water demands of a water distribution network (WDN) composed of ten (10) district metered areas (DMAs). First, pre-processing of the time-series data was performed through outlier elimination, imputation by K-Nearest Neighbors (KNN), and statistical data scaling. Second, the model hyperparameters were calibrated using Bayesian optimization. Third, Long Short-Term Memory (LSTM) coded as a Multi-Step Multivariate Time-Series forecasting model was implemented. Our results indicate that the proposed model produces accurate future water demands, suggesting that feasible short-term water demand forecasting models require combining engineering judgment and computational tools to achieve reliability. Full article

In the scope of energy management systems (EMSs) for microgrids, the forecasting module stands out as an essential element, significantly influencing the efficacy of optimal solution policies. Forecasts for consumption, generation, and market prices play a crucial role in both day-ahead and real-time decision-making processes within EMSs. This paper aims to develop a machine learning-based multivariate forecasting methodology to account for the intricate interplay pertaining to these variables from the perspective of day-ahead energy management. Specifically, our approach delves into the dynamic relationship between load demand variations and electricity price fluctuations within the microgrid EMSs. The investigation involves a comparative analysis and evaluation of recurrent neural networks’ performance to recognize the most effective technique for the forecasting module of microgrid EMSs. This study includes approaches based on Long Short-Term Memory Neural Networks (LSTMs), with architectures ranging from Vanilla LSTM, Stacked LSTM, Bi-directional LSTM, and Convolution LSTM to attention-based models. The empirical study involves analyzing real-world time-series data sourced from the Australian Energy Market (AEM), specifically focusing on historical data from the NSW state. The findings indicate that while the Triple-Stacked LSTM demonstrates superior performance for this application, it does not necessarily lead to more optimal operational costs, with forecast inaccuracies potentially causing deviations of up to forty percent from the optimal cost. Full article