文档介绍：南京工业大学
毕业设计
题目:基于神经网络改进算法的水质评价
学生姓名:
学号:
专业:自动化
班级:浦自0806
指导教师:
2012年5月
基于神经网络改进算法的水质评价
摘要
水质评价是根据评价标准,通过所建立的数学模型,对水体的水质等级进行综合评判。如今水质的水质分析方法主要有因子评价法、综合指数评价法、模糊综合评价法、神经网络评价法等。此次课题曾尝试应用单因子分析法对水质进行评价,但发现此方法过于武断,对水质的评价不够客观。后采用BP神经网络对水质进行分析,介绍BP神经网络模型的结构原理和算法,分析了在建模过程中可能出现的问题并提出了解决方案,根据国家环保总局发布的地表水环境质量标准和南湾水库水质的实测数据,建立了具有较好泛化能力的三层BP神经网络。同时表明,由于BP神经网络模型高度非线性以及输出结果以连续函数形式表达,其对水质进行的综合评价更客观。
关键词:神经网络改进算法水质评价 MATLAB
Based on work Algorithm design of water quality assessment
Abstract
Water quality assessment is prehensive evaluation of water quality, water bodies, through the established mathematical model based on the evaluation criteria. The water quality of the water quality analysis methods factor evaluation method, posite index evaluation method, the prehensive evaluation method, the work evaluation. The subject have tried to single-factor analysis to evaluate the water quality, but this method is too arbitrary and evaluation of water quality is not objective enough. Introduced after the BP work to analyze the water quality, introduced the principle and algorithm of BP work model structure, analysis of the problems that may arise during the modeling process and proposed solutions, surface water environmental quality standards according to the State Environmental Protection Administration released South Bay Reservoir water quality measured data, the BP work has good generalization ability. Also shows that the highly nonlinear and the output of BP work model is the expression of the form of a continuous function, the water quality evaluation more objective.
Keywords:work Improved algorithm Water Quality Assessment MATLAB
目录
摘要 I
ABSTRACT II
第一章绪论 1
水质评价概况 1
研究背景及意义 1
发展的现状 2
本课题研究的内容 3
第二章课题相关的理论基础 4
BP神经网络的内容 4
人工神经网络的内容 4
基本BP算法公式推导 5
BP网络的设计 8
BP神经网络在水质评价中的应用 9
BP神经网络算法改进 9
BP神经网络算法改进措施 9
本文BP神经网络算法的改进措施 11
单因子分析法的内容 12
单因子分析法的概念 12
因子分析法国内外发展、实践