Apply Modified Discrete Particle Swarm Optimization Algorithm to Solve Simulation Task Scheduling Problem

Aiming at the problems that the algorithm was greatly influenced by initial population and the result was not stable when solving the simulation task scheduling in distributed radar simulation system, due to the defects that the discrete particle swarm optimization (DPSO) algorithm was easy to encounter local optimization, a modified DPSO algorithm based on the mutation strategy of pheromone was proposed. The mutation strategy of pheromone was first introduced in this paper after analysis the cause of high probability to encounter local optimization. Then the optimization empirical information of all particles in swarm was fully used to accumulate pheromone. Finally, mutation operation was conducted on particles obtained from iteration of DPSO according to the distribution of pheromone and efficiency matrix. Simulation results show the effectiveness to avoid the problem of local optimization and the better stability of results than DPSO. Meanwhile, the scheduling span and load balance are both improved greatly compared with DPSO, ACO, Max-Min and Min-Min.