A Novel Quantum-behaved Particle Swarm Optimization Algorithm and Its Application to Parameter Optimization of Fuzzy Neural Networks
Jing Zhao, Xinyi Sheng, Jun Jie Sun, Wenbo Xu · INTERNATIONAL JOURNAL ON Advances in Information Sciences and Service Sciences · 2012
A novel Quantum-behaved Particle Swarm Optimization Algorithm with comprehensive learning and cooperative learning approach (CCQPSO) was introduced to improve the global convergence property of QPSO. In the proposed algorithm, the updating method of local attractor, particle’s previous best position and swarm’s global best position were performed in each dimension of the solution vector to avoid loss some components that had moved closer to the global optimal solution in the vector. Then the novel algorithm was applied to parameter optimization of fuzzy neural networks. The introduction of cooperative learning strategy is the originality in the proposed method. The results of prediction of chaotic time series experiment show that the proposed technique can converge more rapidly than other evolutionary computation methods, and the novel method is effective and efficient.