An improved quantum particle swarm optimization algorithm

Abstract

Inspired by the classical PSO method and quantum mechanics theories, this essay presents a new Quantum-behaved PSO (QPSO) algorithm using hyper-chaotic discrete system equation, as hc-QPSO. Our key novelty is to confine all the particles in the identical particle system and use 2-dimensional hyper-chaotic sequence theory and average out the search length to research and consider seasonal fluctuation's interference on particle's search route. First, the identical particle system can remove the seasonal fluctuation's interference on the particle search route in initializing the swarm and achieve better updating of the particle's position. Second, 2-dimensional hyperchaotic sequence theory can conduct chaotic search for every particle. Third, average out the search length is introduced to improve the swarm's full search capacity in the search process, and thereby prevent the particle from falling into local minimum prematurely, fasten the convergence velocity and enhance the accuracy of the algorithm. The simulation results of the classical function have demonstrated that the hc-QPSO algorithm is superior to the classical PSO algorithm and the quantum PSO algorithm in its performance.