Review

Concurrent Engineering: Research

and Applications
Research and development techniques 2018, Vol. 26(3) 215–230
Ó The Author(s) 2018
for early-warning satellite systems Article reuse guidelines:
sagepub.com/journals-permissions

using concurrent engineering DOI: 10.1177/1063293X18768668

journals.sagepub.com/home/cer

Fuyu Sun, Hua Wang and Jianping Zhou

Abstract
An early-warning satellite system is a complex project that requires the participation of many aerospace academies and
scientific institutions. In terms of software programming, this study proposes a new simulation integrated management
platform for the analysis of parallel and distributed systems. The platform facilitates the design and testing of both appli-
cations and architectures. To improve the efficiency of project development, new early-warning satellite systems are
designed based on the simulation integrated management platform. In terms of project management, this study applies
concurrent engineering theory to aerospace engineering and presents a method of collaborative project management.
Finally, through a series of experiments, this study validates the simulation integrated management platform, models, and
project management method. Furthermore, the causes of deviation and prevention methods are explained in detail. The
proposed simulation platform, models, and project management method provide a foundation for further validations of
autonomous technology in space attack–defense architecture research.

Keywords
early-warning satellite system, simulation integrated management platform, modeling method, collaborative project man-
agement, concurrent engineering

Introduction these advantages, space-based early-warning systems

are attracting significant attention.
In modern warfare technology, ballistic missiles are Contemporary research on early-warning satellite
important weapons that are being researched and systems has focused on the following aspects: (1) the
developed by several countries (Simon and Dan, 2012). tracking algorithm, which is mainly related to single/
In addition, ballistic missile defense system (BMDS), multiple target tracking and to the enhancement of the
which are used to detect, intercept, and destroy enemy tracking accuracy (Buontempo, 2015; Li and Jilkov,
ballistic missile warheads with loss of offensive capabil- 2010); (2) the optimal design of the constellation con-
ity, is becoming more advanced, and their weaponry figuration, mainly to improve the ground or space cov-
and control capabilities are increasingly powerful erage (Li et al., 2016; Meziane et al., 2016); and (3) the
(Handberg, 2015). satellite scheduling method, mainly in view of routine
Space-based early-warning systems play an impor-
missions and unexpected missions, and the planning
tant role in BMDS and present several advantages
and scheduling of satellite-borne sensors (Luo, 2015;
compared with ground-, sea-, and air-based early-warn-
ing systems. For instance, they are not restricted by
national boundaries and geographical conditions and College of Aerospace Science and Engineering, National University of
have a wider detection range. Moreover, they have the Defense Technology, Changsha, Hunan, China
ability to perform all-weather monitoring at key
Corresponding author:
regions, with more powerful detection and coverage Fuyu Sun, College of Aerospace Science and Engineering, National
(Garrett et al., 2011). It is also difficult to interfere with University of Defense Technology, Changsha 410073, Hunan, China.
or attack systems that are operating in space. Due to Email: sunfuyu89@126.com
216 Concurrent Engineering: Research and Applications 26(3)

Xhafa et al., 2012). Most contemporary research is lim- analyze the research and development techniques of
ited to specific parts, and the literature lacks studies on early-warning satellite systems. In terms of software
the simulation of entire systems and the relations programming, we propose a new simulation integrated
among subsystems. To fill the research gaps in this management (SIM) platform, which is oriented toward
domain, we study the characteristics of an early- analyzing and studying parallel applications on distrib-
warning satellite system by employing concurrent engi- uted systems in the aerospace domain (in the present
neering for software programming and project manage- study, application to an early-warning satellite system).
ment. In this study, we lay the foundations for a real SIM has been designed to provide flexibility, accuracy,
system and follow-up development. performance, and scalability. These features make it a
The simulation of early-warning satellite systems is a powerful simulation platform for designing, testing,
complicated task. It is necessary to verify the reliability and analyzing both actual and potential architectures.
and security of the mission through a distributed simu- Simulation systems range from a single computing
lation (Cai et al., 2007). Several countries or space node to a complete high-performance distributed sys-
agencies have invested time and resources in building tem. Regarding project management, this study ana-
suitable simulation platforms in related domains lyzes the project characteristics of an early-warning
(Garrett et al., 2011). These platforms play a highly satellite system and presents a method of collaborative
important role in the ground tests. project management that combines organization, inte-
Due to the large number of domains in the field of gration, time, communication, and risk. This simula-
distributed systems, developing a universal simulator is tion platform and project management method have
impractical and unfeasible. Naturally, all researchers been applied to numerical simulations by the China
set their own objectives and requirements; therefore, Manned Space Engineering Office (Zhou, 2011).
each simulator is developed for a specific purpose The remainder of this study is structured as follows.
(Chaudron et al., 2014). As existing simulators gener- In section ‘‘Requirements,’’ specific requirements
ally do not fulfill the requirements of a new project, regarding the application, simulator, and project man-
researchers typically modify an existing simulator or agement are presented. In section ‘‘The SIM platform,’’
code a new one. However, coding a simulator anew is a the SIM platform is introduced, and its features, archi-
very complex task. tecture, operation process, and modeling specifications
Moreover, a project such as an early-warning satel- are discussed. In section ‘‘Modeling method,’’ the trans-
lite system requires the participation of many aerospace formation method of an early-warning satellite
academies and scientific institutions. It is important to system is described, from the mathematical model to
attain efficient management for such complex projects. the SIM-compatible simulation module. In section
To date, many researchers have accomplished consider- ‘‘Experimental design,’’ an experimental design to test
able work on the key technologies in the field of con- the performance of the proposed system is presented.
current engineering, and most of them have focused on Section ‘‘Collaborative project management’’ intro-
collaboration and team work (Aleisa et al., 2011; duces the collaborative project management approach
Cosxkun et al., 2016; Prasad, 1999). Some researchers in terms of organization, integration, time, communica-
have applied concurrent engineering technology to tion, and risk. In section ‘‘Main results and analysis,’’
aerospace engineering. For example, Ahmed et al. the experimental results and analysis are presented.
(2013) presented a transformation process toward lean Finally, in section ‘‘Conclusion and future work,’’ the
product development in the aerospace industry, and conclusions to this study and ideas for future work are
Gabriel et al. (2012) examined particular issues associ- discussed.
ated with the application of contingencies and margins
in the concurrent engineering environment with consid-
eration of resource management techniques and stan-
Requirements
dards in the aerospace industry. Imbert and Li (2010) We analyze the requirements of early-warning satellite
proposed a new approach for product development in systems in terms of the simulation platform, modeling,
the spacecraft production industry using aerospace and project management. This facilitates the identifica-
concurrent engineering, with the primary focus on the tion of specific follow-up tasks.
technology itself, its optimal design, and environment
impact. These research results provide important theo-
retical support for the development, management, and Simulation platform requirements
production of complex engineering systems, particu- Considering the common problems of simulation plat-
larly for aerospace engineering projects. forms for different types of aerospace missions, it is
In this study, concurrent engineering is combined necessary to understand the structure of a new platform
with modern advanced simulation technology to and the relationship among function layers prior to the
Sun et al. 217

design process. It is imperative that we ensure sufficient Project management requirements

versatility, standardization, and extendibility of the Aerospace engineering is recognized as a multidisciplin-
platform (Sun et al., 2017). Special attention should be ary science. Moreover, an early-warning satellite system
paid to the following key points: involves the participation of many scientific institu-
tions. Functional division and organization manage-
1. The missile attack–defense architecture includes ment that is clear, rational, and concise will be
ground-based, air-based, and space-based early- advantageous in the initial project stages.
warning systems. Furthermore, space-based early- Integration is an important segment of system simu-
warning systems usually refer to low earth orbit lations. In macro terms, integration management
(LEO), geostationary orbit (GEO), and high earth includes the project quality, scope, cost, time, and other
orbit (HEO) early-warning constellations. In total, elements; from a micro perspective, it also includes the
several dozen satellites are usually involved. We model assembly, experimental design, system debug-
need to develop models for each satellite, as well as ging, and other project-specific jobs. The purpose of
for satellite-borne sensors, constellation schedul- integration management is to manage all aspects of the
ing, communication links, and other aspects. project through composition, coordination, and inte-
Understandably, the entire missile attack–defense gration, thus facilitating the development of the overall
architecture is extremely complex, requiring the program and the implementation plan of the specific
specialization and cooperation of multiple agents. stage (Demoly et al., 2013). This will ensure the overall
A simulation platform with strong model adapt- success of the project.
ability can simplify the development of the model- The constraints of aircraft launch times and launch
ing process and could ensure the smooth windows must be considered in aerospace missions.
implementation of system integration. Therefore, time management in the development of the
2. In space-based early-warning system simulation project is very stringent. The institutions involved in the
missions, different types of constellation configura- project development stage may be located in different
tions, satellite-borne sensors, and ballistic missiles cities; the ground stations, marine surveillance ships,
must be tested (Hu et al., 2017). The experimental and command posts in the launch and in-orbit opera-
results generated by a combination of different tion stages may be distributed across the country and
types of objects must be analyzed. If the models even around the world. Strong communication manage-
cannot be separated from the platform, we would ment is required to achieve effective communication
need to modify the internal code of the platform; and ensure correct and credible information (Maier
this would lead to an unacceptable workload. et al., 2008).
3. The absence of a good interface complicates the Aerospace projects usually have the high-risk char-
experimental design and may increase the occur- acteristics of high costs, technical complexity, and
rence of mistakes. Moreover, regarding complex numerous uncertainties (Watson et al., 2006). One of
aerospace engineering system simulations, particu- the top priorities of risk management is to take full
larly ballistic missile attack–defense system simula- account of various emergencies and make contingencies
tions including hundreds of models, if there is no to minimize the risks.
friendly user interface, the task of experimental
design will become very difficult and repetitive,
and it would be extremely easy to make mistakes. The SIM platform
Basic architecture
A simulation platform is the key to an overall simula-
Model requirements
tion system. To integrate different types of models, a
The realization of an early-warning satellite system sets simulation platform should provide support for differ-
higher requirements for the project system, such as ent simulation applications. The universalization, stan-
higher precision of satellite orbital predictions, more dardization, and scalability of the simulation platform
powerful satellite-borne sensor detections, shorter will reduce the difficulties of developing the system and
response periods for orbital determination, and higher the time frame required for the establishment of the
precision key-point estimations. Aspects of the entire model. Therefore, the quality of a platform is the eva-
project that need to be verified should be presented, luation standard for a simulation system. The universa-
such as full mission profile verification, mission soft- lization of a simulation system refers to its ability to
ware verification, deviation analysis strategy, and the adapt to several types of applications. In this study,
visualization display for mission scenarios. various ground tests were considered for early-warning
218 Concurrent Engineering: Research and Applications 26(3)

Figure 1. Global architecture of the SIM platform.

satellite systems. Standardization emphasizes the speci- configuration of connections among models, and deter-
ficity among functional layers and is the criterion for mination of model attributes. This module establishes a
platform–platform and platform–model connections. communication bridge between the user and the simu-
Scalability requires the SIM platform to cooperate with lation system. A good user interface creates a pleasant
other simulation tools by performing different roles. user experience and simplifies the entire process of
There are two main scenarios: integrating an external simulation experiments. Note that a layered design the-
simulator within SIM and integrating the SIM frame- ory is also used for the experiment management, with
work within another simulator. Figure 1 shows the glo- models and model connections separated to increase
bal architecture of SIM. the flexibility of the test design and the reusability of
The SIM platform integrates the management of the the models.
models, experiments, distribution, operations, data, The idea of ‘‘distributed calculation, concentrated
and nodes. Model management can unify different management’’ has been adopted for the distribution
types of models by conducting operations using entire management. Thus, each model and its parameters are
models or by amending the information of a single distributed to each node based on the simulation sce-
model. The design of module is based on a layered narios. The user can either operate the models on the
design theory that separates the platform from the node or operate the models remotely through the com-
models, model description, and model realization; this puter itself. Related information is presented on the test
is convenient for modeling, experiment designs, and the interface.
subsequent construction of a simulation system. The operation management considers the responsi-
Experiment management is mainly applied for the bility of driving different types of models and distribu-
assembly of models using visualization software, ted scheduling. Moreover, operation management can
Sun et al. 219

Figure 2. Organizational structure relationships.

be employed to monitor the state of a node and save Simulation. Users operate the distribution management
data at a breakpoint. tool to obtain the scenario file from the simulation sce-
Data management is deployed on the management nario library. The scenario file contains model informa-
node and is responsible for collecting simulation data tion, distribution information, and so on. The node
from the computing nodes and displaying them in the management tool participates in the distribution pro-
form of a diagram or report. Data management cess. The operation management tool is applied to
requires a real-time data display and centralized data drive the model on each node, with the cooperation of
processing. the node management tool. The distribution manage-
Node management is applied on each computing ment tool and the operation management tool are
node to guard the node, conduct the model scheduling, assisted by the distribution middleware.
and report the node’s state. Such functions are con-
ducted in cooperation with the management node. Evaluation. The data management tool obtains the
model output from each node and can save data at
breakpoints during model operation. In addition, it
Operation process can collect results from all nodes at the end of the
The SIM platform combines the functions of modeling, experiment. All data are saved to the experiment
simulation, and evaluation. The application of the SIM database.
platform is now described according to the aforemen-
tioned functions. The relationships within the organiza-
Modeling specifications
tional structure are shown in Figure 2.
Currently, the SIM platform can drive three types of
models, hereafter referred to as Models A, B, and C.
Modeling. Users apply the model development tool to Model A is a type of simulation subsystem software
create a model program framework. The core algo- that can independently operate in a Windows environ-
rithm of the model design can be found in the model ment. It exists in the form of an executable program
algorithm library, and a debugging environment is pro- (.exe). The connections of Model A include basic model
vided alongside a model debugging tool. The model is information, initial parameters, inputs, outputs, and
then stored in the library by means of the model man- specific files related to the executable program. SIM
agement tool. drives Model A by means of memory mapping. The file
According to the simulation requirements, specific mapping of the memory, which is assigned based on the
models can be selected from the aerospace model model name, responds to control commands to com-
library. After the visual assembly and configuration of plete the model initialization, operation, stopping, time-
the parameters according to the aerospace parameter hopping, and data recycling, among other functions.
library, an experimental scenario file is output and Model B exists in the form of a dynamic link library
saved to the scenario simulation library. (.dll), which is mainly composed of three special
220 Concurrent Engineering: Research and Applications 26(3)

functions (see Appendix 1). The data interface of

Model B must be standardized for user interaction.
The forms of the model ports are pParam, pInput, and
pOutput. The pParam port is the pointer for the initia-
lization parameters, representing specific constant
parameters that describe the model characteristics. In
addition, pInput is used to reserve the parameters
whenever any changes are imported into the model.
Similarly, pOutput is used to reserve the parameters
exported from the model. Model C denotes models
constructed in MATLAB (.m). These have no fixed
program template, but do require fixed files.
In this study, early-warning satellite system models
were developed in the form of Model B. Thus, we focus Figure 3. Information flowchart of modules.
on the modeling and driving method of Model B. The
simulation engine deals with different types of models SensorDetect module is partially obtained from the
in various approaches. For Model B, the most impor- SatDynamics module and target orbital information,
tant part of the engine is the callback function. This is which is loaded in the form of a configuration file. The
the bridge between the platform and the models. At the outputs of the module are observational data regarding
beginning of each simulation step, the simulation the target. The input of the MisTrajDetermine module
engine can run the callback function to achieve the pur- is partially obtained from the SatDynamics module and
pose of driving the models. SensorDetect module. The outputs of this module are
The callback function is mainly composed of four the target orbital estimation parameters after the filter-
branch statements. It is only necessary to embed the ing process. The input of the MisKeyPointEst module
model functions of the Model B template (Listings 1 to is partially obtained from the MisTrajDetermine mod-
3) into the corresponding branch. The callback func- ule and SensorDetect module. Information on the
tion contains platform instructions, and the simulation launch and landing points of the target and other key
engine drives the models according to these instruc- points is calculated based on the target orbital estima-
tions. The callback function is given in Listing 4. tion parameters.
In the process of modeling, we used a model tem-
plate to restrict the module ports. The aforementioned
Modeling method modules were developed in the form of Model B. Each
In the process of modeling an early-warning satellite module included the initial function MdlInit, step func-
system, it is very important to consider the complicated tion MdlSim, and reprocessing function MdlEnd. The
satellite dynamics and sensor detection conditions, the three functions have a unified form, which makes mod-
high precision of the orbital determination, and the ule development more convenient.
estimation of key points. These factors require the cor- Figure 4 illustrates the class diagram of the early-
relation modules to be sufficiently accurate. Regarding warning satellite system. From the diagram, we can
the modeling, the developed program must conform to observe that the entire simulation system consists of
the standard of each module type, particularly that of the SIM platform and the SatDynamics, SensorDetect,
the module ports. According to the requirements of the MisTrajDetermine, and MisKeyPointEst modules.
system simulation, we built a SatDynamics module, The SatDynamics module includes the CSat Dynamics_
SensorDetect module, MisTrajDetermine module, and init, CSatDynamics_sim, and CSatDynamics_ end func-
MisKeyPointEst module according to the relevant tions. CSatDynamics_sim includes several functions that
mathematical models. refer to basic satellite dynamics, such as the coordinate sys-
Figure 3 shows the flow of information among the tem transformation, orbital/attitude parameter transforma-
modules. The SatDynamics module can be initialized tion, and satellite orbital element prediction. This module
through the configuration file of the constellation, is the foundation and the key point for establishing the sys-
which contains the orbit and attitude parameters of tem, and most of its outputs are used by other modules.
each satellite at the initial time. In addition, the outputs The SensorDetect module includes the CSensor
of the module are the orbit and attitude parameters of Detect_init, CSensorDetect_sim, and CSensorDetect_
each satellite, except for the state calculated through end functions. This module simulates the satellite-borne
extrapolation at each simulation time. The input of the sensor operation principle for the calculation of satellite
Sun et al. 221

Figure 4. Class diagram of the early-warning satellite system.

observations using target orbital information. To simu- can draw specific significant conclusions that could be
late an actual disturbance, a random error is added to useful in future modeling studies. Moreover, the experi-
the observational parameters. mental results provide baselines for estimating the
The MisTrajDetermine module aims to mitigate validity of the platform and the models.
the effects of the observational random error in We conducted a series of tests on the early-warning
order to determine the target. It consists of the satellite system. These tests focused on the satellite con-
CMisTraj Determine_init, CMisTrajDetermine_sim, stellation module, sensor detection module, orbital
and CMis TrajDetermine_end functions. The CMisTraj determination module, and key-point estimation mod-
Determine_sim function is encapsulated by two orbital ule. The orbital data of the target were calculated via
determination methods for single/dual satellite detec- the STK software. The results were then fed into the
tion; these include specific filtering algorithms. sensor detection module as input from a data file.
The MisKeyPointEst module includes the CMisKey After adding the models into the model base as
PointEst_init, CMisKeyPointEst_sim, and CMisKey planned, we designed experiments using the experiment
PointEst_end functions. This module is mainly used to management tool of the SIM platform. Note that the
calculate the target launch/landing location using the interface between the modules must be matched cor-
estimated target orbital parameters. Thus, the SIM rectly. In the event of a data type mismatch, SIM will
platform executes model management, experiment issue a warning regarding the nature of the error.
management, distribution management, operation Before the final experiment, we needed to test each
management, data management, and node manage- module independently to ensure the validity of the mod-
ment functions. els. The SIM debugging tool helped us to complete this
task, shortening the model development period, and
making the process more pleasant. Figure 5 shows the
Experimental design interface of the experimental design tool.
Testing is highly important in the verification of simula- The main aim of the testing is to solve the problem
tion platforms. By analyzing the results of experiments, of top-level and subsystem design based on the project
the advantages and disadvantages of the platform and requirements. The purpose of the testing is divided into
the models can be identified. Regarding the models, we four levels: (1) help experts work out the preliminary
222 Concurrent Engineering: Research and Applications 26(3)

Figure 5. Interface of the experimental design tool.

overall plan and technical indicators; (2) provide sup- suitability of the technical requirements and preci-
port for the joint simulation, and verify the interface sion index.
correctness and compatibility of the subsystems; (3) lay
the foundation for BMDS and subsequent operational
missions; and (4) make preparations for technical inno-
Collaborative project management
vation and mission evaluation. An early-warning satellite system requires the participa-
The test procedure includes a monitoring strategy tion of many aerospace academies and scientific institu-
analysis, normal operational process simulation, emer- tions. It is important to achieve reasonable and efficient
gency simulation, and other thematic simulations: management for such a complex project. This section
mainly discusses the management of the organization,
1. Monitoring strategy analysis. Analyze the structure integration, time, communication, and risk.
and deployment of the early-warning satellite con-
stellation, analyze the space coverage ability of the
constellation, and analyze the observation and
Organization management
scheduling strategy of the constellation with respect As shown in Figure 6, China Aerospace Science and
to the multiple targets. Industry Corp. (CASIC) plays the role of project mas-
2. Normal operational process simulation. Verify the ter for the overall design. The expert team is mainly in
monitoring strategy, correctness of the interface charge of the design, decision, and evaluation.
between subsystems, and convergence of each flight Moreover, the project team concentrates on the practi-
phase and control command. The test period cov- cal implementation. The early-warning satellite system
ers the entire mission process. can be divided into five subsystems: HEO satellite sys-
3. Emergency simulation. Research the influence of tem, LEO satellite system, space-borne sensor system,
satellite voidness and measurements regarding the telemetry track and command system, and ground con-
sudden emergence of new targets. trol system. The relationship between CASIC and the
4. Other thematic simulations. Improve the precision other teams relies on information feedback and mission
of short-arc orbital determination based on assignment. Moreover, the relationship between the
the data fusion of ground-based, sea-based, and expert team and project team relies on information
air-based systems. Verify the correction and feedback and project review.
Sun et al. 223

plan is the prerequisite and foundation for the project

implementation. The lifecycle, contributory factors,
and entire system must be considered in accordance
with the project management system. The process of
project programming is shown in Figure 7.

Time management
Figure 8 shows the execution process of the plan. The
dynamic management and project assessment should
be strengthened in terms of the overall plan, stage
division, technical program, technical economics,
development process, major experiment items, and
design. Potential risks and technical difficulties within
the project organization and implementation must be
considered before the formation of the project plan.
Figure 6. Organizational structure of project The project must proceed according to the plan after
clearing the main mission nodes, risk, quality control
level, and other detailed requirements. Furthermore,
Integration management suggestions from the research and development units
The main processes in this project are the execution and are also important and must be taken into full
control of the simulation. The integration management consideration.

Figure 7. Process of project programming.

224 Concurrent Engineering: Research and Applications 26(3)

Figure 8. Execution process of the plan.

Communication management
To ensure that the schedule is implemented at all levels,
a detailed reporting and meeting system, such as weekly
scheduling, thematic scheduling, quality analysis, and
reviews, is established according to the responsibilities
and working characteristics of each project team. Each
week, the expert team should make a detailed assess-
ment on the progress of the entire project and solve
existing problems. The process of communication man-
agement is shown in Figure 9.

Risk management
The process of risk management is shown in Figure 10.
The system consists of three submodules, namely risk
identification, risk analysis, and risk treatment
modules.
The risk identification module mainly identifies the Figure 9. Process of communication management.
project risk by analyzing the project characteristics and
the risk case statistics of the early-warning satellite sys- In the third part, the subject of deviation analysis is dis-
tem. Finally, the risk classification and index system of cussed. This section presents several meaningful conclu-
the entire project are established. sions on the early-warning satellite system.
The risk analysis module mainly evaluates the over-
all project risk. This module provides a comprehensive
risk evaluation in terms of cost, schedule, and Monitoring of real-time operating results using the
performance. SIM platform
The risk treatment module mainly compares the A number of important model outputs were monitored
result of the risk analysis module with the preset risk and controlled. Figures 11 and 12 present the real-time
level. If the preset risk level has not been exceeded, the estimations of the launch point and landing point,
project will continue. Otherwise, effective measures which were obtained from the key-point estimation
must be taken to reduce the risk, and it will be neces- model. The launch point estimation maintains a high
sary to return to the second module to re-evaluate the precision at the start of the simulation. The estima-
risk until the end of the project. tion results vary within a small interval around the
true value (latitude/longitude is –39.696°/100.266°).
However, the estimation of the landing point displays
Main results and analysis
the opposite trend. As the number of sampling points
This section consists of three parts. The first deals with increases, the landing point estimation becomes more
the monitoring of the runtime results using the accurate. The results then converge and approach the
SIM.Data software as a data management tool within actual value (latitude/longitude is –56.464°/–13.346°).
the SIM platform. The second part concerns the post- The final estimation of the launch/landing points has
processing of the experimental data within MATLAB. a small deviation from the actual values.
Sun et al. 225

Figure 10. Process of risk management.

Figure 11. Launch point estimation (from the key-point estimation model).
226 Concurrent Engineering: Research and Applications 26(3)

Figure 12. Landing point estimation (from the key-point estimation model).

quickly converges to the actual value after being

Table 1. Launch/landing point estimated at the boost phase detected. However, the landing point estimation is
terminus (in ECF coordinates).
a process of iterative corrections, with hysteretic
Actual Estimation Deviation convergence to the actual value.
(°) (°) (°)

Launch Latitude –39.696 –39.431 –0.27

point Deviation analysis
Longitude 100.266 100.547 –0.28
Landing Latitude –56.464 –56.155 –0.31 Launch point estimation error.
point The precise modeling of the target motion during the
Longitude –13.346 –13.001 –0.35
boost phase is difficult because of the complexity and
diversity of physical models. Hence, there are likely to
be errors in the estimation of the launch point. The
Post-processing of experimental results in MATLAB boost phase presents different characteristics from the
rising, turning, and aiming phases. Moreover, multi-
We now describe the post-processing and analysis of
stage targets exist. It is important to select a simple and
the launch/landing points. Figure 13 shows the fitting
practical model to describe the actual motion features
process of the launch/landing point estimation. We can
of the target. The method of polynomial fitting can
draw the following meaningful conclusions:
solve this problem, along with orbit reconstruction and
launch point estimation. However, some deviation
1. At the boost phase terminus, the actual launch between the reconstructed orbit and the actual orbit
point of the target is –39.696° (latitude), 100.266° will remain.
(longitude) and the landing point is –56.464° (lati- Early-warning satellite systems typically have a detec-
tude), –13.346° (longitude) (Table 1). The estima- tion height of 30 km or more. Therefore, it is necessary
tion results from the key-point model are –39.431°, to estimate the launch point through orbit extrapola-
100.547° for the launch point and –56.155°, – tion. The difference between the extrapolated orbit and
13.001° for the landing point. The deviation the real trajectory will result in estimation errors.
between the actual and estimated data is less than
0.5°. Converted into distance, the deviation of the
launch point is 38.34 km and that of the landing Landing point estimation error.
point is 40.55 km. This deviation is only acceptable Target state error at burnout point. The estimation of
for high-orbit early-warning satellite systems with the landing point is calculated by orbit extrapolation
observational information of the target at the from the moment at which the burnout point is
boost phase. reached. The estimation relies on the target state para-
2. Because the two-body gravitation model was used meters of the burnout point. Hence, the estimation
for the orbital model of the target, the launch point accuracy of the burnout point directly affects the
Sun et al. 227

Figure 13. Launch/landing point fitting.

landing point estimation. The target state error at the In addition, SIM provides several modules for the simu-
burnout point is affected by the determination of the lation of different components and modeling strategies.
burnout point and the state estimation itself. Furthermore, the system uses a standardized interface
Model error. For the extra-atmospheric flight phase, that allows the implementation of new modules.
the model achieves high precision. However, for the re- A series of models for early-warning satellite systems
entry phase, the uncertainty in the aerodynamic coeffi- was developed. Relative mathematical models were
cient of the missile and inaccuracy of the atmospheric designed and simulation modules were configured and
model makes it difficult to improve the accuracy of the executed in SIM. Our platform produced very satisfac-
landing point estimation. tory results in terms of accuracy and performance.
Height error of landing point. Assuming that that the Finally, the experimental results were post-processed in
earth is a standard ellipsoid, we can obtain the landing MATLAB. Through a systematic analysis of the experi-
point at which the target orbit intersects with the earth. mental results, we comprehensively investigated the
The difference between the actual terrain elevation and deviations in the data and proposed reasonable model
the earth ellipsoid model will result in an error in the improvements.
landing point estimation. Due to the steepness and In terms of project management, this study analyzed
transience of the re-entry phase, this error can be the project characteristics of early-warning satellite sys-
ignored. Moreover, we can utilize global elevation tems and presented a method of collaborative
information to satisfy the accuracy requirements and project management covering organization, integration,
improve the precision of the landing point estimation. time, communication, and risk. The method provides a
solid guarantee of successful project completion.
Furthermore, it forms a management basis for other
Conclusion and future work complex aerospace projects.
In this study, concurrent engineering has been com- Based on the results of this study, we plan to
bined with modern advanced simulation technology to upgrade the SIM platform from the engineering level
analyze the research and development techniques used to the engagement level; SIM will become an excellent
in early-warning satellite systems. In terms of software simulation platform, with sufficient versatility and flex-
programming, this study has presented a modeling and ibility for aerospace systems engineering. From the per-
simulation platform for investigating the complicated spective of modeling development, a method for the
applications of aerospace engineering systems. The pro- precise modeling of early-warning satellite systems is
posed SIM platform facilitates the design and testing greatly required. Modeling of space attack–defense
process of both applications and architectures. architecture will be promoted as the foundation of
SIM has a modular design in which the main compo- national security strategies. Project management will be
nents are the basic systems of distributed architecture, appended by investigating a reward and punishment
such as computing, memory, storage, and network mechanism. Ideas such as the dedication of team mem-
components. This design follows a hierarchical philoso- bers and load balancing across teams are other poten-
phy, with basic modules grouped into larger modules. tial research pathways.
228 Concurrent Engineering: Research and Applications 26(3)

Declaration of conflicting interests Li YJ, Zhao SH and Wu JL (2016) A general evaluation cri-
The author(s) declared no potential conflicts of interest with terion for the coverage performance of LEO constellations.
respect to the research, authorship, and/or publication of this Aerospace Science and Technology 48: 94–101.
article. Luo KP (2015) Space-based infrared sensor scheduling with
high uncertainty: issues and challenges. Systems Engineer-
ing 18(1): 102–113.
Funding Maier AM, Kreimeyer M, Hepperle C, et al. (2008) Explora-
The author(s) received no financial support for the research, tion of correlations between factors influencing communi-
authorship, and/or publication of this article. cation in complex product development. Concurrent
Engineering: Research and Applications 16(1): 37–59.
Meziane TI, Métris G, Lion G, et al. (2016) Optimization of
References small satellite constellation design for continuous mutual
Ahmed AA, Golob M and Attia UM (2013) The transforma- regional coverage with multi-objective genetic algorithm.
tion of product development process into lean environment International Journal of Computational Intelligence Systems
using set-based concurrent engineering: a case study from 9(4): 627–637.
an aerospace industry. Concurrent Engineering: Research Prasad B (1999) System integration techniques of sharing and
and Applications 21(4): 268–285. collaboration among work-groups, computers and pro-
Aleisa E, Suresh NC and Li L (2011) Team formation in con- cesses. Journal of Systems Integration 9(2): 115–139.
current engineering using group technology (GT) concepts. Simon AM and Dan R (2012) Ballistic missiles and interna-
Concurrent Engineering: Research and Applications 19(3): tional conflict. Journal of Conflict Resolution 57(5): 854–880.
213–224. Sun FY, Wang H and Zhou JP (2017) Simulation integrated
Buontempo JT (2015) A trajectory for homeland ballistic mis- management: a new type of simulation platform application
sile defense. Defense & Security Analysis 31(2): 99–109. for aerospace systems engineering. Simulation 93(2): 149–171.
Cai ZH, Xiao Z and Yao F (2007) Modeling/simulation of Watson P, Curran R, Murphy A, et al. (2006) Cost estima-
space early warning system based on HLA and simulation tion of machined parts within an aerospace supply chain.
grid. Modern Defense Technology 35: 11–15. Concurrent Engineering: Research and Applications 14(1):
Chaudron JB, Saussie D and Siron P (2014) Real-time distrib- 17–26.
uted simulations in an HLA framework: application to air- Xhafa F, Sun JZ, Barolli A, et al. (2012) Genetic algorithms
craft simulation. Simulation 90(6): 627–643. for satellite scheduling problems. Mobile Information Sys-
Cosxkun H, Erdisx E and Genc x O (2016) Improving the perfor- tems 8(4): 351–377.
mance of construction projects by employing concurrent Zhou JP (2011) Rendezvous and docking technology of
engineering. Journal of the Faculty of Engineering and manned space flight. Manned Spaceflight 2: 1–8.
Architecture 31(2): 47–58.
Demoly F, Dutartre O, Yan XT, et al. (2013) Product rela-
tionships management enabler for concurrent engineering Appendix 1
and product lifecycle management. Computers in Industry
64(7): 833–848. Model template
Gabriel K, Tupper H, Hume P, et al. (2012) Resource man-
To clarify the functionality of the proposed SIM plat-
agement and contingencies in aerospace concurrent engi-
form, we present some supplementary explanation.
neering. In: AIAA SPACE 2012 conference & exposition,
Pasadena, CA, 11–13 September 2012, pp. 1–15.
Appendix 1 introduces the model template as a supple-
Garrett RK, Anderson S, Baron NT, et al. (2011) Managing ment to section ‘‘Modeling specifications.’’ Listings 1 to
the interstitials: a system of systems framework suited for 3 present the code framework of Model B. Listing 4
the ballistic missile defense system. Systems Engineering presents the architecture of the callback function.
14(1): 87–109. Currently, the SIM platform can drive three types of
Handberg R (2015) The symbolic politics of ballistic missile models: an executable program (Model A), dynamic
defense: seeking the perfect defense in an imperfect world. link library (Model B), and MATLAB (Model C). The
Defense & Security Analysis 31(1): 44–57. following two points need to be noted:
Hu YP, Huang JY and Chen L (2017) Space-based visible
observation strategy for beyond-LEO objects based on an
1. The SIM platform drives Model A without the
equatorial LEO satellite with multi-sensors. Advances in
Space Research 59(7): 1751–1762.
assistance of model-driven software, whereas
Imbert M and Li X (2010) Aerospace concurrent engineering: Models B/C require the assistance of model-driven
a modern global approach. International Conference on software. The model-driven software is SIM.Dll
Space Information Technology 7651(1): 1–13. Drive/SIM.MatlabDrive, respectively. We have
Li XR and Jilkov VP (2010) Survey of maneuvering target independently developed these two model-driven
tracking. Part II: motion models of ballistic and space tar- software programs.
gets. IEEE Transactions on Aerospace and Electronic Sys- 2. In the programming phase, Models A and B
tems 46(1): 96–119. require fixed program templates. Users need only
Sun et al. 229

Listing 1. Initialization function.

Listing 2. Advanced function.

Listing 3. Stop function. Listing 4. Callback function.

fill in the corresponding position in the template Therefore, Listings 1 to 3 show the template for
with their own core code. Model C is a MATLAB Model B. We packaged a dynamic link library with a
script with no fixed C++ program template. We fixed program template to restrain the initialization,
need to solve the problem of C++ /MATLAB running, and post-processing functions. Listing 4 is the
mixed programming to drive Model C. template for Model A and represents a core function of
230 Concurrent Engineering: Research and Applications 26(3)

model-driven software, herein referred to as SIM. from the SIM platform with the callback function, and
DllDrive/SIM.MatlabDrive. The main design idea of realize model-driven functionality. The model-driven
SIM.DllDrive/SIM.MatlabDrive is to parse the model software is like a bridge connecting the model with the
program information of Models B/C, execute orders SIM platform.

Author biographies
Fuyu Sun is studying for his PhD in the College of Aerospace Science and Engineering at the
National University of Defense Technology. His current research focuses on aerospace orbital
dynamics and control, as well as attack and defense system simulation.

Hua Wang is an associate professor in the College of Aerospace Science and Engineering at the
National University of Defense Technology. His current research focuses on aerospace systems
simulation.

Jianping Zhou is an academician at the Chinese Academy of Engineering. He has been serving as
chief commander and chief designer of China’s manned space flight projects. He is also the director
of the Chinese Society of Astronautics.