论文部分内容阅读
针对项目实践中目标解常需要满足多种目标约束的难题,提出一种改进的多目标遗传算法。该方法对遗传算法进行改进,选择算子在下一代群组的选择上引入精英保留策略和小生境的概念,根据拥挤距离的排序确定下一代的个体,变异算子在适应度值小于平均适应度时,采用一种自适应的方式确定算子。最后将所提出的算法应用于解决排课问题,相比遗传算法计算速度更快,排课效果更好,表明了该算法的有效性。
Aiming at the problem that target solution often needs to satisfy many kinds of objective constraints in project practice, an improved multi-objective genetic algorithm is proposed. This method improves the genetic algorithm. The selection operator introduces the concept of elitist retention strategy and niche in the selection of the next generation group, and determines the individuals of the next generation according to the congestion distance rankings. The fitness of the mutation operator is less than the average fitness Degrees, an adaptive way to determine the operator. Finally, the proposed algorithm is applied to solve the scheduling problem, compared with the genetic algorithm to calculate faster, better timetable effect, indicating the effectiveness of the algorithm.