Dynamic Battle Management Using Meta-Heuristic Algorithms, Fuzzy Inference Systems and the Decision Tree

Nowadays, the threats caused by aerial targets are among the most critical challenges. Two main components in threat management are visualization and decision-making based on compiled pictures. In this process, the decisions are made based on threat evaluation and resource allocation modelling. To do so, the sequential nature of resource allocation is considered to provide a more accurately visualized situation. Here, a multi-stage multi-objective model is constructed based on the fuzzy inference system and the decision tree. Also, NSGA-II and SPEA-II are used to find Pareto solutions. Both Algorithms have been compared based on generational distance as a convergence measure, spread as a diversity measure, and the actual computational time. Ultimately, the TOPSIS method is used to make the final decision while the results are reported using a simulated scenario. The simulation results show that the SPEA-II has better convergence and spread, while NSGA-II is faster and has less standard deviation in the execution time. We believe that the fuzzy inference system is more suitable than the decision tree in practical applications. Nevertheless, when facing data shortage or incompatibility, the decision tree would be the preferred choice.