Abstract
Combat system effectiveness simulation (CoSES) needs to model both the physical aspect (i.e. physics modelling) and intelligent aspect (i.e. decision modelling) of combat systems. Combat platform decision-making has several characteristics such as cognition, diversity, agility, uncertainty and higher abstraction level, which bring tough challenges for decision model design, implementation and optimization. In this paper, we propose a domain-specific modelling approach which develops friendly modelling environments for model design, we design code generation mechanisms to transform domain-specific decision models to Python code which is supported by a Python script framework to implement decision models and we present a Bayesian network-based statistical analysis method on simulation output data to optimize the decision model. The case study shows that the proposed modelling and optimization approach effectively supports CoSES with decision models of higher efficiency and increased effectiveness.
Acknowledgements
We thank the anonymous reviewers from both Journal of Statistical Computation and Simulation and AsiaSim 2012 conference for their valuable suggestions. We are grateful to the discussions with Dr. Lei Wang, Professor Qi Liu and Professor Jing Feng on Bayesian decision analysis. The work presented in this paper is partly supported by the National Natural Science Foundation of China (Nos. 61273198, 91024015, 61074107, 60974073, 60974074 and 71031007).