Improved fireworks algorithm and its application in PID parameters tuning

Abstract

In this paper, an improved fireworks algorithm is proposed in PID control system performance optimization. This new algorithm called fireworks algorithm with genetic mutation (FWAGM) is based on adaptive fireworks algorithm (AFW), which is inspired on fireworks explosion process and has adaptive local search ability. However, the Gaussian mutation in AFW is not efficient to avoid stuck in local optimum. In FWAGM, the genetic mutation operation using the better individuals in all the generations to generate diverse mutation sparks is introduced. The FWAGM is applied in the PID control performance optimization in the MATLAB software. The result shows that FWAGM acquires better control performance than AFW and genetic algorithm (GA). The new algorithm improves heuristic search efficiency, and it's proved to have obvious advantage for both control performance optimization and other optimization problems.