A new mutation operator with the ability to adjust exploration and exploitation for DE algorithm
Mingcheng Zuo, Guangming Dai, Lei Peng, Maocai Wang, Pan Peng, Changchun Chen · Proceedings of the Genetic and Evolutionary Computation Conference Companion · 2019
This paper proposes a new mutation operator Current --- to --- better-AC with an angle control rule for DE algorithm. This new mutation operator has the potential to automatically adjust the exploration and exploitation based on the angle control rule. Under the control of this rule, the individuals pair for generating mutation vector is tendentiously selected. In Current --- to --- better-AC, only high-ranking individuals are permitted to use the angle control rule, while low-ranking individuals are encouraged to move to a random high-ranking individual. This method is embedded into JADE as a new algorithm AC-JADE, and its performance is tested by COCO benchmarks. Experimental results show that AC-JADE performs better than JADE.