【摘要】 就径向基函数人工神经网络(RBF)在水质评价中的应用做了探讨。分别从网络结构的选择、可调参数的优化方法和学习样本的代表性3方面作了详细的分析、阐述,并提出了一些解决办法,为RBF网络在水质评价中的应用提供了一定的参考材料。最后以博斯腾湖为例,利用RBF网络建立水质评价模型,模型评价结果趋势基本符合实际,效果较好。更多还原

【Abstract】 Applying RBF in water quality assessment is a new development. This paper discussed several key steps in such application, including selecting RBF network structure, optimizing adjustable parameters and improving representativities of learning samples. Suggestions are given to improve the performance in solving problems. Then the RBF method is used in a case study about water quality assessment of Bosten Lake, and the results are in good agreement with monitoring data.更多还原