The performance of genetic algorithm(GA) is determined by the capability of search and optimization for satisfactory solutions. The new adaptive genetic algorithm(AGA) is built for inducing suitable search and optimiz...The performance of genetic algorithm(GA) is determined by the capability of search and optimization for satisfactory solutions. The new adaptive genetic algorithm(AGA) is built for inducing suitable search and optimization relationship. The use of six fuzzy logic controllers(6FLCs) is proposed for dynamic control genetic operating parameters of a symbolic-coded GA. This paper uses AGA based on 6FLCs to deal with the travelling salesman problem (TSP). Experimental results show that AGA based on 6FLCs is more efficient than a standard GA in solving combinatorial optimization problems similar to TSP.展开更多
The constriction factor method (CFM) is a new variation of the basic particle swarm optimization (PSO), which has relatively better convergent nature. The effects of the major parameters on CFM were systematically inv...The constriction factor method (CFM) is a new variation of the basic particle swarm optimization (PSO), which has relatively better convergent nature. The effects of the major parameters on CFM were systematically investigated based on some benchmark functions. The constriction factor, velocity constraint, and population size all have significant impact on the per- formance of CFM for PSO. The constriction factor and velocity constraint have optimal values in practical application, and im- proper choice of these factors will lead to bad results. Increasing population size can improve the solution quality, although the computing time will be longer. The characteristics of CFM parameters are described and guidelines for determining parameter values are given in this paper.展开更多
脑电图(Electroencephalogram,EEG)是诊断癫痫发作的重要依据。针对人工识别癫痫脑电信号中出现的效率低易误诊等问题,依据遗传算法和支持向量机理论,提出基于遗传算法结合支持向量机分类模型(GM-SVM)的癫痫发作脑电信号识别方法。将支...脑电图(Electroencephalogram,EEG)是诊断癫痫发作的重要依据。针对人工识别癫痫脑电信号中出现的效率低易误诊等问题,依据遗传算法和支持向量机理论,提出基于遗传算法结合支持向量机分类模型(GM-SVM)的癫痫发作脑电信号识别方法。将支持向量机相关参数设计成遗传个体,将遗传算法的适应度值设置为GM-SVM的识别准确率,通过迭代寻优获得较优的识别效果。最后,该方法在伯恩大学癫痫研究中心的脑电数据上进行训练和评估,结果表明该方法可以正确识别癫痫发作脑电信号,并达到98%的精度和99%的AUC(Area Under ROC Curve,ROC曲线下的面积),相较于其他分类算法有较优的识别性能。展开更多
The main technical problems that should be considered in the design of hydro-turbine generating units of Three Gorges Project (TGP) are analyzed;the key technical researches performed are summarized,and the parameters...The main technical problems that should be considered in the design of hydro-turbine generating units of Three Gorges Project (TGP) are analyzed;the key technical researches performed are summarized,and the parameters of hydro-turbine generating units are optimized through the study on key technical problems.The unit operation indicates that the performance of the hydro-turbine generating units is excellent,and the units can operate in a safe,stable and highly efficient mode for a long term.Therefore,it is verified effectively that the general technical design of units is scientific and rational.展开更多
文摘The performance of genetic algorithm(GA) is determined by the capability of search and optimization for satisfactory solutions. The new adaptive genetic algorithm(AGA) is built for inducing suitable search and optimization relationship. The use of six fuzzy logic controllers(6FLCs) is proposed for dynamic control genetic operating parameters of a symbolic-coded GA. This paper uses AGA based on 6FLCs to deal with the travelling salesman problem (TSP). Experimental results show that AGA based on 6FLCs is more efficient than a standard GA in solving combinatorial optimization problems similar to TSP.
基金Project (No. 20276063) supported by the National Natural Sci-ence Foundation of China
文摘The constriction factor method (CFM) is a new variation of the basic particle swarm optimization (PSO), which has relatively better convergent nature. The effects of the major parameters on CFM were systematically investigated based on some benchmark functions. The constriction factor, velocity constraint, and population size all have significant impact on the per- formance of CFM for PSO. The constriction factor and velocity constraint have optimal values in practical application, and im- proper choice of these factors will lead to bad results. Increasing population size can improve the solution quality, although the computing time will be longer. The characteristics of CFM parameters are described and guidelines for determining parameter values are given in this paper.
文摘脑电图(Electroencephalogram,EEG)是诊断癫痫发作的重要依据。针对人工识别癫痫脑电信号中出现的效率低易误诊等问题,依据遗传算法和支持向量机理论,提出基于遗传算法结合支持向量机分类模型(GM-SVM)的癫痫发作脑电信号识别方法。将支持向量机相关参数设计成遗传个体,将遗传算法的适应度值设置为GM-SVM的识别准确率,通过迭代寻优获得较优的识别效果。最后,该方法在伯恩大学癫痫研究中心的脑电数据上进行训练和评估,结果表明该方法可以正确识别癫痫发作脑电信号,并达到98%的精度和99%的AUC(Area Under ROC Curve,ROC曲线下的面积),相较于其他分类算法有较优的识别性能。
文摘The main technical problems that should be considered in the design of hydro-turbine generating units of Three Gorges Project (TGP) are analyzed;the key technical researches performed are summarized,and the parameters of hydro-turbine generating units are optimized through the study on key technical problems.The unit operation indicates that the performance of the hydro-turbine generating units is excellent,and the units can operate in a safe,stable and highly efficient mode for a long term.Therefore,it is verified effectively that the general technical design of units is scientific and rational.
