基于事例推理的装配顺序规划中的遗传算法改进设计

An Improved Genetic Algorithm Designed for CBR-Based Assembly Sequence Planning

针对基于事例推理(CBR)的装配顺序推理的特点,建立了相应的遗传算法模型.通过大量的算法实验,分析比较了传统进化、阶段进化和自适应进化策略的优劣,以及算法参数对推理效率和推理质量的影响.在此基础上,设计了改进的遗传算法.最后的实例分析验证了该算法的有效性.

In case based reasoning （CBR）-based assembly sequence planning （ASP）, genetic algorithm （GA） can be utilized to improve the efficiency and quality of the reasoning process. A GA model was constructed according to the characteristic of CBR-based ASP. Meanwhile, to optimize the GA algorithm, different evolution strategies （including conventional evolution strategy, steps evolution strategy, and selfadaptation evolution strategy） and the operation parameters （crossover rate and mutation rate） were explored thoroughly based on a large number of calculation tests. Thereafter, an improved genetic algorithm was proposed and verified using a real engineering application.