【摘要】 针对核主成分分析中核函数参数难以确定的问题,建立以第一主成分贡献率为目标函数,以核参数为优化变量的优化模型,运用SCEM-UA优化算法实现求解。将该方法应用于地下水水质评价中,通过与传统主成分分析法的对比,结果表明:基于SCEM-UA优化的KPCA法可以客观、精确地得到最优化核参数,提高了综合评价的可信度,同时也验证了该方法在水质综合评价中的可行性。更多还原

【Abstract】 As the parameters of kernel function are difficult to determine,the contribution rate of the first principal component is established as the goal function and the kernel parameters are established as optimal variables.Then SCEM-UA method is used to resolve the problem.Comparing traditional PCA(Principal Component Analysis)methods shows that SCEM-UA-based KPCA(Kernel-PCA)method can calculate objectively and precisely the optimal kernel parameters and improve the credibility of a comprehensive assessment.Meanwhile,the feasibility of this method used for water quality assessment is also verified.更多还原