Hidden markov model control of inertia weight adaptation for Particle swarm optimization

Particle swarm optimization (PSO) is a stochastic algorithm based population that integrates social interactions of animals in nature. One of the main challenges within PSO is to balance between global and local search throughout the course of a run. To achieve this trade-off, various adaptive PSOs have been proposed in order to control the values of its parameters. The present paper makes an attempt to determine a generalized adaptive framework for the setting of the inertia weight parameter which is named HMM-wPSO. That is, a control mechanism of the inertia weight is proposed based on the estimation of states using hidden Markov model (HMM). We performed evaluations on ten benchmark functions to test the HMM control of inertia weight parameter for the PSO. Experimental results show that our proposed scheme outperforms other compared PSO variants in major cases in terms of solution accuracy and convergence speed. © 2017