{"title":"差分进化中的精英强化策略","authors":"Chun-Ling Lin, Sheng-Ta Hsieh","doi":"10.1109/IRCE.2019.00027","DOIUrl":null,"url":null,"abstract":"Differential evolution (DE) is one of powerful optimizers. It has various mutation strategies for solving widely applications. In order to draw on the strong points of different mutation strategy and to offset their weaknesses, the elitist reinforcement strategy is proposed. Thus, in the proposed method, there are more than one vector-group will be involved in DE, the mutation strategy of each vector-group should be different. If one of vector-group cannot find better solution, elitist reinforcement strategy will join a few elitist vectors to replace vectors which with poor performance. The elitist vectors can guide other vectors toward to potential solution space. In order to test proposed method, the CEC 2005 test functions are adopted for experiments. Also, three DE variants are involved for comparison. From the results, it shows that the proposed method performs better than other three DE methods.","PeriodicalId":298781,"journal":{"name":"2019 2nd International Conference of Intelligent Robotic and Control Engineering (IRCE)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2019-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Elitist Reinforcement Strategy for Differential Evolution\",\"authors\":\"Chun-Ling Lin, Sheng-Ta Hsieh\",\"doi\":\"10.1109/IRCE.2019.00027\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Differential evolution (DE) is one of powerful optimizers. It has various mutation strategies for solving widely applications. In order to draw on the strong points of different mutation strategy and to offset their weaknesses, the elitist reinforcement strategy is proposed. Thus, in the proposed method, there are more than one vector-group will be involved in DE, the mutation strategy of each vector-group should be different. If one of vector-group cannot find better solution, elitist reinforcement strategy will join a few elitist vectors to replace vectors which with poor performance. The elitist vectors can guide other vectors toward to potential solution space. In order to test proposed method, the CEC 2005 test functions are adopted for experiments. Also, three DE variants are involved for comparison. From the results, it shows that the proposed method performs better than other three DE methods.\",\"PeriodicalId\":298781,\"journal\":{\"name\":\"2019 2nd International Conference of Intelligent Robotic and Control Engineering (IRCE)\",\"volume\":\"3 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2019-08-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2019 2nd International Conference of Intelligent Robotic and Control Engineering (IRCE)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/IRCE.2019.00027\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2019 2nd International Conference of Intelligent Robotic and Control Engineering (IRCE)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/IRCE.2019.00027","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Elitist Reinforcement Strategy for Differential Evolution
Differential evolution (DE) is one of powerful optimizers. It has various mutation strategies for solving widely applications. In order to draw on the strong points of different mutation strategy and to offset their weaknesses, the elitist reinforcement strategy is proposed. Thus, in the proposed method, there are more than one vector-group will be involved in DE, the mutation strategy of each vector-group should be different. If one of vector-group cannot find better solution, elitist reinforcement strategy will join a few elitist vectors to replace vectors which with poor performance. The elitist vectors can guide other vectors toward to potential solution space. In order to test proposed method, the CEC 2005 test functions are adopted for experiments. Also, three DE variants are involved for comparison. From the results, it shows that the proposed method performs better than other three DE methods.