摘要
在模糊逻辑系统及控制的设计中,为了避免生成过多的模糊规则影响系统的效果,有必要进行模糊规则的筛选。本文将遗传算法与模糊逻辑系统相结合,使用遗传算法进行规则筛选,并调整模糊神经网络系统的参数,以此来提高模糊逻辑系统的精确性。将选择出来适应度最高的规则作为规则基,设计模糊逻辑系统。为了检验系统的可行性,将设计的模糊逻辑系统应用于丹麦克朗/人民币汇率的预测中,并将遗传算法与最小二乘法进行比较,仿真结果表明,所提出的方法是有效的,取得了更好的效果。
To avoid generate too many fuzzy rules in fuzzy logical system,it necessary to select fuzzy rules.Through the study of genetic algorithms(GA)and fuzzy logic system knowledge and uses this algorithm to select rules of the fuzzy logical system to improve the accuracy of fuzzy logical system (FLS).The selected optimum performance as a rule of the rule base,and together with the parameters into the neural network and fuzzy logic system corresponding design.In order to test the performance of the system,the design of the fuzzy logic system used to predict the DKK/CNY exchange rate in comparison,by comparing GA algorithm and the least square method,the simulation results show the effectiveness of the proposed design method and can be realized better performance.
作者
王怡杰
王涛
兰洁
WANG Yi-jie;WANG Tao;LAN Jie(College of Science,Liaoning University of Technology,Jinzhou121001,China)
出处
《模糊系统与数学》
北大核心
2018年第5期78-86,共9页
Fuzzy Systems and Mathematics
基金
辽宁省高校基本科研业务资助项目(JL201615410)
辽宁省自然科学基金指导项目(20180550056)
