摘要
针对当前车间调度多目标优化研究存在收敛速度慢、精度低的问题,提出了混沌多目标粒子群优化算法。在算法中,设计了一种新的叠加Logistic扰动的Tent混沌映射算子,通过该算子周期性地更新种群以保证种群的多样性;对收缩粒子群算法进行了扩展使其能够快速收敛到Pareto前沿。通过标准测试问题与实际应用对所提方法进行了验证,实验结果显示混沌多目标粒子群优化算法无论在收敛速度还是在优化精度上都优于其它典型多目标进化算法。
Since the current job shop scheduling multi-objective optimization has the drawbacks of slow conver-gence speed and low accuracy, it proposes a chaotic multi-objective particle swarm optimization algorithm. In the algo-rithm, designed the Tent chaotic mapping a new stack Logistic disturbance, the operator periodically update population in order to ensure the diversity of population;on the contraction of particle swarm algorithm is extended so that it can rapidly converge to the Pareto front. The standard test problems and practical application to verify the proposed meth-od, experimental results show that the chaotic multi-objective particle swarm optimization algorithm both in conver-gence speed and optimization accuracy is better than other typical multi-objective evolutionary algorithm.
出处
《激光杂志》
CAS
北大核心
2015年第1期122-127,共6页
Laser Journal
基金
宁波市自然科学基金2012A610071
关键词
车间调度
混沌算子
种群多样性
多目标优化
粒子群算法
scheduling
chaos operator
population diversity
multi-objective optimization
particle swarm algorithm
