机构地区: 江南大学物联网工程学院
出 处: 《控制与决策》 2012年第5期719-723,730,共6页
摘 要: 为了进一步提高量子行为粒子群优化(QPSO)算法的全局收敛性能,有效改善算法中存在的粒子早熟问题提出一种基于完全学习策略的改进QPSO算法(CLQPSO).该学习策略改变了QPSO中局部吸引子的更新方式,充分利用了种群的社会信息.采用8个测试函数对算法性能进行比较分析.实验结果表明,所提出的改进算法不仅收敛速度快,而且全局收敛能力好,收敛精度优于PSO算法和QPSO算法. A quantum-behaved particle swarm optimization(CLQPSO) algorithm based on comprehensive learning strategy is presented,which helps prevent the original quantum-behaved particle swarm optimization(QPSO) algorithm's tendency to be easily trapped into local optima as a result of the rapid decline in diversity.The learning strategy changes the updating method of local attractor in QPSO,which makes fully use of the social information of the swarm.The 8 benchmark functions are used to test the performance of CLQPSO.The experiments results show that the proposed algorithm can find better solutions than the original QPSO algorithm and the PSO algorithm with higher efficiency.
关 键 词: 粒子群优化 量子行为粒子群优化 完全学习策略 局部吸引子
领 域: [自动化与计算机技术] [自动化与计算机技术]