研究目的
Investigating the effectiveness of a cooperative differential evolution algorithm with multiple populations for multiobjective optimization.
研究成果
The proposed cooperative differential evolution algorithm with multiple populations (CMODE) demonstrates superior performance compared to state-of-the-art multiobjective DE algorithms and other popular multiobjective evolutionary algorithms on benchmark problems with two, three, and many objectives. The algorithm effectively balances convergence and diversity, and its online search behavior and parameter sensitivity are well-understood.
研究不足
The study does not address the scalability of the algorithm to very high-dimensional problems or its performance on real-world applications. The computational cost increases with the number of objectives and subpopulations.
1：Experimental Design and Method Selection:
The study employs a cooperative differential evolution (DE) algorithm with multiple populations for multiobjective optimization. It includes M single-objective optimization subpopulations and an archive population for an M-objective optimization problem. Adaptive DE is applied to each subpopulation and the archive population.
2：Sample Selection and Data Sources:
Benchmark problems with two, three, and many objectives from CEC2009 and WFG test suites are used for evaluation.
3：List of Experimental Equipment and Materials:
The study utilizes computational resources for running the algorithms and evaluating their performance on the benchmark problems.
4：Experimental Procedures and Operational Workflow:
The algorithm is tested on various benchmark problems to evaluate its performance in terms of convergence and diversity. The online search behavior and parameter sensitivity are also investigated.
5：Data Analysis Methods:
Performance is evaluated using inverted generational distance (IGD) and hypervolume (HV) indicators. Statistical tests, including Wilcoxon and Friedman tests, are conducted to compare the algorithm's performance with state-of-the-art multiobjective DE algorithms and other popular multiobjective evolutionary algorithms.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
