改进遗传复合形算法及其在并行公差优化设计中的应用

Improved Genetic Polygon Algorithm and Its Application in Concurrent Tolerance Optimization Design

提出了一种将遗传算法与复合形算法相结合而构成的遗传复合形算法.在遗传算法初始化之后,用复合形法选出较好的个体,再用遗传算法寻求最优解,并首次将加权制造成本和质量损失成本同时应用到公差并行设计的优化数学模型中.研制出面向公差并行设计的新的优化软件.经过试验验证较以前的算法有较大改进.

A new genetic algorithm（ GA）--genetic polygon algorithm（GPA） is proposed. It is formed by the combination of the genetic algorithm and the polygon algorithm （PA）. In the first place, the population is initialized using the genetic algorithm. Then the better individuals are selected using the polygon algorithm. At last, the best individual is found using the genetic algorithm. For the first time, the weighted cost and the quality loss cost are applied synchronously in the optimized mathematical model of the concurrent tolerance optimization design. New optimal software facing concurrent tolerance design was developed. The examination shows that the new arithmetic is better than the former.