论文部分内容阅读
研究了药房智能存取系统拣选路径的动态规划问题,提出了该问题的数学模型,并设计了一种新的自适应混合粒子群遗传算法(Adaptive hybrid particle swarm algorithm)。该算法在粒子群遗传混合算法的基础上引入了动态调整和自适应进化的策略。在算法前期粒子群搜索阶段,建立了惯性权重系数、认知系数与收缩因子之间的联动关系,随着惯性权重的动态变化,认知系数与收缩因子也适时进行调整,提高了搜索效率和搜索精度。在算法的后期,采用了遗传算法的自适应交叉和变异的进化过程,对陷入局部最优的粒子群进行打散,使得每次迭代中都能最大限度的获取路径信息,使种群的搜索朝向解空间的不同区域发展。经过对某大型医院智能存取系统的路径规划仿真实验,验证了提出的算法相对于其他算法在求解速度和求解精度上都有较大的提高。
The problem of dynamic programming of the intelligent access system of pharmacy is studied. The mathematical model of the problem is put forward and a new adaptive hybrid particle swarm algorithm is designed. This algorithm introduces the strategy of dynamic adjustment and adaptive evolution on the basis of hybrid PSO algorithm. In the early stage of the algorithm, the correlation between inertia weight coefficient, cognitive coefficient and shrinkage factor was established. With the dynamic change of inertia weight, cognitive coefficient and shrinkage factor were also adjusted in time, which improved the search efficiency and Search accuracy. In the latter part of the algorithm, the evolutionary process of adaptive crossover and mutation of genetic algorithm is adopted to break up the particle swarm that falls into the local optimum so that the path information can be obtained to the maximum extent in each iteration, so that the search of the swarm Solutions to different areas of space development. After the path planning of a large hospital intelligent access system simulation experiments, the proposed algorithm compared to other algorithms in the solution speed and solution accuracy are greatly improved.