求解约束多目标区间优化问题的人工免疫算法

An Artificial Immune Algorithm for Solving Constrained Multi-objective Interval Optimization Problems

针对含区间约束的多目标区间值优化问题,设计求解该问题的人工免疫算法。借助区间可能度模型,比较两个区间数,并将区间约束转化为确定性约束。设计关于迭代次数的约束允许违背度,自适应将种群划分为约束允许群和非约束允许群;设计与抗体受支配数成反比的亲和力计算方式,由此根据亲和力大小选择抗体进行克隆及进化。对比性仿真实验表明,该算法在求解含区间约束的多目标区间值优化问题方面所获解集在质量和分布均匀性能方面均具有较好的潜力。

An artificial immune algorithm is designed to solve the multi-objective interval-valued optimization problem with interval constraints.The interval probability model is used to compare two interval numbers,and the interval constraints are transformed into deterministic constraints.Also,the constraint permissible violation degree about the number of iterations is designed,and the population is adaptively divided into constraint permissible group and constraint non-permissible group.In addition,an affinity calculation method inversely proportional to the dominant number of antibodies is designed so that antibodies could be cloned and evolved according to the affinity.The comparative simulation results show that the proposed algorithm has good potential in gaining quality and uniform distribution of the solution set for multi-objective interval value optimization problems with interval constraints.