Abstract: In the paper，a hybrid of genetic algorithm is proposed for the shelf-space allocation problem.Firstly，the genetic algorithm encodes space-allocations as matrixes whose element of ith row and kth column is the allocation amount of facing of product i on shelf k.Secondly，crossover operator and mutation operator are designed，and offspring generated by crossover and mutation is feasible.Thirdly，a threshold value ζ based on structure of phenotype and its fitness is defined to generate good enough initial population.Fourthly，simulated annealing algorithm and local adjusted method are parallelly executed to each individual in subpopulation generated by crossover and mutation operator separately to increase abilities of local search of the algorithm.Compared with other heuristic methods，the hybrid genetic algorithm is shown to be a very efficient algorithm which allocates shelf space at near optimal levels.

摘要： 针对货架分配问题提出了一个遗传算法与模拟退火算法及一个局部搜索算法混合的算法。首先，设计了一种比较直观的编码方法，用一个矩阵作为一种货架分配方案。第二，设计了与编码相应的杂交和变异算子，并且杂交、变异都能生成可行解，不需要对解进行修正。第三，为了能够生成好的初始种群，定义了一个阀值，这个阀值不仅反映了解的适应值的信息，而且还反映解的结构的信息。第四，为了增加算法的局部搜索能力，同时又尽量不增加计算的复杂度，让模拟退火算法和一种局部搜索算法并行作用于相应的子群。通过大量的数据模拟实验及与其他的几种算法模拟结果进行比较，实验显示，该算法不论是计算结果还是算法的稳定性都优于其他算法。