基于种群个体可行性的约束优化进化算法

提出一种新的求解约束优化问题的进化算法.该算法在处理约束时不引入惩罚因子,使约束处理问题简单化.基于种群中个体的可行性,分别采用3种不同的交叉方式和混合变异机制用于指导算法快速搜索过程.为了求解位于边界附近的全局最优解,引入一种不可行解保存和替换机制,允许一定比例的最好不可行解进入下一代种群.标准测试问题的实验结果表明了该算法的可行性和有效性.

Constrained optimization evolutionary algorithm based on individual feasibility of population

A novel constrained optimization evolutionary algorithm is proposed for solving constrained optimization problem, which does not introduce penalty parameters to deal with constraints. In the process of opulation evolution, the proposed algorithm searches the solution space of the problem through three different crossover methods based on population feasibility. A mixed mutation strategy is used to guide the process fast toward the feasible region of the search space. In addition, an infeasible solution diversity conservation and replacement strategy is used to keep a certain number of infeasible
solutions in each generation so as to enforce the evolutionary search toward an optimal solution from both sides of feasible and infeasible regions. The proposed algorithm is tested on eight well-known constrained optimization problems, and the experiment result shows the effectiveness and feasibility of the method.