Abstract: To solve the problems of variousness and irregular quantity of Cut Order Planning(COP)sizes in apparel mass customization, the mathematical model was built and two stage optimization method based on probability search and genetic algorithm was proposed. In the first stage, several initial cut table layout plans which satified the production constriction were generated randomly. Combined searching algorithm with probability, the optimal sizes combination plan as well as initial cut table layout plan were obtained according to the principle of minimum number of cut table. In the second stage, the plans obtained from first stage were optimized again by using genetic algorithm, and the optimal cut table layout plan was received according to the principle that the producting proportion of redundant apparel not exceeding the allowable maximum value. A practical production case was given and computed separately by two stage method and manual empirical method. The results and comparison showed that the two stage optimization method could rapidly get the apparel cutting plan, decrease the number of cut table, save fabric and reduce cutting costs in the same production condition.

Key words: mass customization, cut order planning, two stage optimization, probability search, genetic algorithms

摘要： 为解决大批量定制服装生产中裁剪分床计划尺码较多且各尺码数量不规则的问题，建立了裁剪分床计划的数学模型;提出基于概率搜索和遗传算法的两阶段优化方法进行求解，第一阶段随机生成若干满足生产约束的初始裁床铺料层数方案，利用搜索算法结合概率，按投入裁床数量最少的原则得到最优尺码组合方案和相应初始裁床铺料层数方案，第二阶段基于前一阶段得到的最优尺码组合方案，按照满足订单情况下生产多余服装的比例不超过企业允许的最大值原则，利用遗传算法再次优化得到最优裁床铺料层数方案。针对实际生产案例，分别利用本算法和人工经验算法求解并进行比较，结果表明在相同的生产条件下，两阶段优化方法能快速求解出服装裁剪分床方案，减少铺床数、节省面料并降低成本。

关键词: 大批量定制, 裁剪分床计划, 两阶段优化, 概率搜索, 遗传算法