A Dynamic Chaotic Mutation Based Particle Swarm Optimization for Dynamic Optimization of Biochemical Process

It is difficult to optimize complicated biochemical process. Particle swarm optimization algorithm is suitable for solving this kind of problem with non-linear, high dimension, non-differentiable. However, PSO has not a good trade-off between global exploration and local exploitation, which limits its applications. In order to improve search performance, a dynamic chaotic mutation based particle swarm optimization(dcmPSO) is proposed in this work. The operation of dynamic chaotic mutation leads searching in a wide region at the beginning of iteration and in a small one at the end of evolution. The proposed dcmPSO is applied to a dynamic optimization problem of a Park-Ramirez bioreactor and the computational results validate the effectiveness.