多种群人工免疫算法在多峰函数上的优化

Optimization of multiple populations artificial immune algorithm on multimodal functions

针对人工免疫算法在多峰函数优化上存在优化精度低的缺点,提出了多种群人工免疫算法(MAIA)对多峰函数进行优化。MAIA包含多个独立的抗体种群,独立的抗体种群各自进行抗体选择、克隆和变异等免疫操作,在每个独立种群更新和评价完后选择每个种群中最好的抗体进行多种群评价,然后将当前的最佳抗体共享给每个单独种群最终生成各个种群新一代的抗体群。仿真实验结果表明:相比于人工免疫算法,MAIA求解精度更高,提高了多峰函数寻优的精度。

In view of the disadvantages of artificial immune algorithm with low optimization precision in multi peak function optimization, the optimization of multimodal functions by MAIA was proposed in this study.The MAIA contained multiple independent antibody populations.Independent antibody populations performed antibody selection and cloning operations.Each individual population was updated and evaluated, and then the best antibody in each population was selected for the multiple population evaluation. After that, the current best antibody was shared to each individual population and finally the new generation of antibody groups was generated.The simulation results showed that MAIA solved higher accuracy and improved the accuracy of multi-peak function optimization compared with artificial immune algorithm.