面向柔性作业车间调度问题的混沌编码量子粒子群优化算法

Chaotic-encode Quantum PSO Algorithm for Flexible Job-shop Scheduling Problem

为解决柔性作业车间调度问题,提出一种混沌编码量子粒子群优化算法。针对标准量子粒子群优化算法中粒子过早收敛于局部最优值的缺点,提出具有扰动行为的自适应收缩-扩张系数和关联粒子适应度值的计算方法,改善算法的全局搜索能力;通过引入混沌边界变异策略,减少粒子大量聚集在边界的概率,增加种群的多样性来提高搜索最优解的能力;针对量子粒子群优化算法的迭代特性,设计一种适用的混沌编码策略。将提出的改进量子粒子群优化算法应用于柔性作业车间调度问题,并通过多种基准算例与标准量子粒子群优化算法、粒子群优化算法和混合遗传算法进行对比,验证所提算法的性能。实验结果表明:混沌编码量子粒子群优化算法具有更好的稳定性和更强的寻优能力。

To solve the flexible job-shop scheduling problem(FJSP),a chaotic-encode quantum PSO(CQPSO)algorithm is proposed.Aiming at the premature convergence of particles to local optimum in standard QPSO,the methods for computing the adaptive contraction-expansion coefficient and mean best position using fitness values of associated particles are proposed to improve the global search ability of QPSO.Through chaotic boundary variation strategy,the probability of a large number of particles gathering at the boundary is reduced and the population diversity is increased to enhance the ability of searching the optimal solution.According to the iterative property of QPSO,a chaotic-encode strategy is designed.The proposed CQPSO is applied to solve FJSP and the result is compared with QPSO,PSO,and hybrid genetic algorithm result on several benchmarks to confirm the performance.The experimental results show that CQPSO has the better stability and the stronger optimization ability.