基于模糊RBF神经网络的地表水环境质量评价

Surface Water Quality Assessment Based on Fuzzy RBF Artificial Neural Network

针对现有水质评价方法评价结论过保护、存在人为因素影响、表达模糊信息能力弱等问题,结合模糊评价和RBF神经网络的优点,构建了模糊RBF神经网络水质综合评价模型。模型应用于淮河安徽段,并与单因子评价法进行对比,以验证模型的合理性。评价结果表明,两种方法的评价结果基本一致,模型客观合理,淮河安徽段部分监测点水质Ⅴ类、劣Ⅴ类居多,总体水质状况较差,评价结果符合实际情况。

To solve the problems in existing water quality evaluation methods, such as over-protecting, artificial interference and poor ability to express fuzzy information, a fuzzy RBF artificial neural network model for the comprehensive evaluation of water quality is established by combining the advantages of fuzzy recognition and RBF artificial neural network. The model is applied to evaluate the water quality of the reach of Huaihe River in Anhui Province, and the results is compared with single factor assessment method to prove the rationality of model. The evaluation results show that two methods basically have same evaluating results, and the new model is objective and reasonable. The water quality in many monitoring sites of the reaches of Huaihe River in Anhui Province is ClassVor poor ClassV, and the evaluation results are is accordance with actual situation.