摘要
水环境质量评价中,评价因子和评价等级之间是一种复杂的非线性关系。人工神经网络权重因子可以通过学习自动调节,不需要人为设置权重,能较好的处理复杂的非线性关系。通过构建基于BP(反向传播)算法的人工神经网络模型,对抚河水质进行评价。评价结果与模糊聚类方法得出的结果进行对比可发现,神经网络法可以客观、准确的得出评价等级,同时还可以分析出同一类别河段污染程度的不同,实例证明它是一种较好的水环境质量评价方法。
The evaluation gene and water quality grades have a complicated non-linear relationship. Artificial neural network model can easily solve the complex non-linear relationship by self-learning which can change weights of every layers without man involving. One BP( Back propagation) neural network model on the water quality of Fu River assessment is built. Compared to the fuzzy cluster algorithm method, the model can be impersonally and exactely obtain the assessment levels. Moreover, it can analyze the contaminated degree. Some instances show it is an effective water quality assessment method.
出处
《东华理工大学学报(自然科学版)》
CAS
2008年第1期85-88,共4页
Journal of East China University of Technology(Natural Science)
基金
东华理工大学校长基金项目(DHXK0717)
关键词
水质评价
BP算法
神经网络模型
抚河
water quality assessment
BP algorithm
neural network model
Fu River