Particle Back Propagation Neural Network Optimization Algorithm for CMTA

摘要

Cooperative multiple target attack(CMTA) is one of the most important part in air combat, many optimization methods such as heuristic algorithm, ant colony algorithm have been used to solve the problem. This study proposes a novel algorithm that is Particle Back Propagation neural network Optimization (PBPO) for solving decision making problem of air combat CMTA. PBPO algorithm took the basic idea of Particle Swarm Optimization (PSO) algorithm, such as particles, global best position, past best position and moving particles' position to get best solution. Instead of using conventional PSO formula to compute particle position, BPNN (Back Propagation Neural Network) was constructed and trained to get a new particle position. Using PBPO to solve the decision making problem in CMTA, simulation result shows that the novel algorithm has a good performance.