摘要
通过对自动化药房快速发药系统的储位优化所受约束条件和目标函数的分析,基于静态存储理论,建立了该问题的数学模型,提出了利用GA-PSO混合粒子群算法来解决此问题.该算法引入了遗传算法的搜索机制形成初始粒子群,并引入"交叉"和"变异"的概念形成离散粒子群算法.在优化的过程中,采用精英策略和进化逆转操作提高了搜索能力和寻优速度.对陷入局部最优的粒子群进行变异,使粒子群在新的引导下改变方向,继续寻找问题最优解,从而避免了粒子重复收敛于一点的现象.仿真试验结果表明,该策略是有效的,在求解速度和求解质量上得到了很大提高.
After analyzing the constraints and the objective functions of the rapid drug delivery system of automatic pharmacy, the mathematical models were established based on static storage theory, and a GA-PS0 Hybrid Particle Swarm Algorithm was proposed to solve this problem. This algorithm introduced the search mechanism of Genetic Algorithm to form the initial particle swarm and the concept of crossover and mutation to form Discrete Particle Swarm Algorithm. During the process of optimization, searching capabilities and optimizing speed were increased by introducing the elitist strategies and evolutionary reversal operations. Besides, it mutated the particle swarms which were trapped into local optimum, and leaded the particle swarm to change the direction along new guidance, and continued to search for the optimal solution of the problem, thus the particles were avoided the repeated convergent phenomenon. Simulation experimental results show this algorithm is effective, and the solution speed and quality are greatly improved.
出处
《华北水利水电大学学报（自然科学版）》
2014年第6期84-88,共5页
Journal of North China University of Water Resources and Electric Power：Natural Science Edition
基金
国家自然科学基金项目(U1204605)
关键词
自动化药房
储位优化
混合算法
交叉
变异
automatic pharmacy
optimization of storage location
hybrid algorithm
crossover
mutation
