摘要
Due to the increasing complexity of optimization problems, distributed differential evolution (DDE) has become a promising approach for global optimization. However, similar to the centralized algorithms, DDE also faces the difficulty of strategies' selection and parameters' setting. To deal with such problems effectively, this article proposes an adaptive DDE (ADDE) to relieve the sensitivity of strategies and parameters. In ADDE, three populations called exploration population, exploitation population, and balance population are co-evolved concurrently by using the master-slave multipopulation distributed framework. Different populations will adaptively choose their suitable mutation strategies based on the evolutionary state estimation to make full use of the feedback information from both individuals and the whole corresponding population. Besides, the historical successful experience and best solution improvement are collected and used to adaptively update the individual parameters (amplification factor F and crossover rate CR) and population parameter (population size N), respectively. The performance of ADDE is evaluated on all 30 widely used benchmark functions from the CEC 2014 test suite and all 22 widely used real-world application problems from the CEC 2011 test suite. The experimental results show that ADDE has great superiority compared with the other state-of-the-art DDE and adaptive differential evolution variants.
学科与关键词
- Distributed Differential Evolution With Adaptive Resource AllocationLi, Jian-Yu; Du, Ke-Jing; Zhan, Zhi-Hui*; Wang, Hua; Zhang, Jun*IEEE Transactions on Cybernetics, 2023, 53(5): 2791-2804.被引 29
- Investigation and Improvement of Distributed Differential Evolution Algorithm ClouddeLuo, Liu-Yue; Shi, Lin*; Zhan, Zhi-Hui*13th International Conference on Advanced Computational Intelligence (ICACI), Wanzhou, PEOPLES R CHINA, 2021-05-14 To 2021-05-16.
- Adversarial Differential Evolution for Multimodal Optimization ProblemsJiang, Yi; Chen, Chun-Hua*; Zhan, Zhi-Hui*; Li, Yun; Zhang, JunIEEE Congress on Evolutionary Computation (CEC), Padua, ITALY, 2022-07-18 To 2022-07-23.被引 2
