Quantum particle swarm optimization algorithm with the truncated mean stabilization strategy
Nan-Run Zhou,Shu-Hua Xia,Yan Ma,Ye Zhang
DOI: https://doi.org/10.1007/s11128-021-03380-x
IF: 1.965
2022-01-19
Quantum Information Processing
Abstract:Many feasible and promising swarm intelligence algorithms have been proposed to solve the optimization problems, especially the global numerical optimization problems. It is easier to be in not a global state but a local optimal state, as the particle swarm optimization algorithm cannot update the population samples during the iterative process. And the search performance of the particle swarm optimization algorithm is poor in complex optimization problems. To solve these problems, a quantum particle swarm optimization algorithm based on a truncated mean stabilization strategy is designed. In the proposed algorithm, the processed particles are firstly intercepted with equal proportions, and then the worst particles are replaced by new particles during the iterative process, finally the global optimal value is located with the quantum wave function. The proposed strategy can improve population diversification and increase convergence efficiency. The performance of the quantum-behaved particle swarm optimization algorithm with truncated mean stabilization strategy is researched on a set of unimodal and multimodal benchmark function optimization problems, and is compared with several popular population-based algorithms. Experimental results show that the search accuracy and the convergence of the quantum particle swarm optimization algorithm with truncated mean stabilization strategy are better than other typical particle swarm optimization variants.
physics, multidisciplinary,quantum science & technology, mathematical