基于T-S模糊神经网络模型的钦州市主要河流水质评价

Water Quality Assessment of Main Rivers in Qinzhou City Based on T-S Fuzzy Neural Network Model

为客观评价河流水质,基于钦州市6条主要河流水质监测数据,将T-S模糊神经网络模型用于钦州市主要河流水质评价。结果表明:通过训练的T-S模糊神经网络模型具有很强的泛化能力,训练样本最大误差绝对值小于0.05;检验样本最大误差绝对值仅为0.092 4,能满足水质综合评价要求;钦州市内主要河流水质相对较好,处于Ⅱ~Ⅲ类,其中大风江、张黄江水质较好,马江较差。

In order to evaluate the water quality objectively, the T - S fuzzy neural network model was applied to analyze the water qual- ity of main rivers in Qinzhou based oll the water quality monitoring data. The results demonstrated that the trained T - S fuzzy neural network model had strong generalization ability and the maximum absolute error was less than 0.05. The maximum absolute error value of test sample was only 0. 0924, which could meet the requirements for comprehensive evaluation of water quality. Water quality of main rivers in Qinzhou city was relatively good （ among class II -III ）. However, the water quality of DaFeng river and Zhang Huang riverwas better, while Ma Jiang River was opposite.