改进混沌遗传算法寻优敏捷供需链动态调度时段

摘要

From the perspective of enterprise agility improvement , bottleneck time period resource scheduling was key to agile supply chain .Facing optimization design of dynamic time period scheduling for agile supply chain , a dynamic scheduling model was constructed with the lowest total cost target .Based on common defects of traditional genetic algo-rithm and limitations of heuristic algorithm , an improved chaos genetic algorithm was proposed for time period dynamic scheduling global optimization .Firstly, sectional type code was designed , then stochastic method and greedy method were used to produce better initial population , with high chromosome feasibility and excellent genetic effect .Further-more, priority reservation crossover and goal orientation mutation were chosen to ensure good genes inherited and improve genetic operation .Then, implementation of local neighborhood search and chaos search can speed up convergence .Opti-mal solution criterion was also put forward .Finally, validity of the algorithm was demonstrated by an example .Not only global optimal solution was obtained , but also higher daughter convergence and lower divergence were achieved .%本文面向敏捷供需链动态调度时段优选方案设计，构建以最低总成本为目标的动态调度模型；基于传统遗传算法的常见缺陷以及启发式算法的局限性，提出面向敏捷供需链时段资源动态调度全局寻优的改进混沌遗传算法。首先设计分节式编码，再利用随机法与贪心法产生更优良初始种群，提高染色体可行性及遗传效果；选用优先保留交叉以及贪心机制下的目标导向变异，确保优良基因继承，改善遗传操作；实施局部邻域搜索以及混沌搜索以加快收敛；提出最优解判别法。最后，实例验证算法有效性，不但取得全局最优解，而且子体更加收敛，离散度更低。