变异粒子群算法在装配线平衡问题中的应用

Mutation Particle Swarm Optimization Algorithm in the Application of the Assembly Line Balancing Problem

为找到更加符合实际的解,建立装配线平衡问题模型时,考虑在最大化生产线效率的基础上,增加了平滑指数这一目标函数。应用粒子群算法进行求解时,为避免常规算法易过早陷入局部最优这一不足,提出了一种变异粒子群算法。该算法对设定步长内位置没有更新的个体采用多点变异的方法增加种群多样性,从而达到改变个体极值与全局极值的目的。通过横向搜索、纵向进化的机制,可有效提高种群的搜索能力。最后,通过对实例库中例子的求解,验证了算法的可行性。

To search for more suitable feasible solution,the smooth index of the objective function was added in this paper when establishing assembly line balancing problem model of considering based on maximizing the efficiency of the production line. This paper proposed an improved algorithm for avoiding the original algorithm trapping in local optimal value early,when apply particle swarm optimization algorithm to solve the problem. In this paper,the proposed algorithm adopted multipoint mutation methods increasing population diversity to achieve the purpose of changing individual optimal value and global optimal value,when individuals are unchanged in the setting step location. The search ability of the population is increased by the mechanism of transverse search and longitudinal evolution. Finally,the proposed algorithm is proved feasible by solving the instance.