基于遗传神经网络的水质评价模型优化及应用-人文地理学专业论文.docxVIP

基于遗传神经网络的水质评价模型优化及应用专业：人文地理学 基于遗传神经网络的水质评价模型优化及应用 专业：人文地理学 研究生： 周雪倩 指导教师：丁辉 摘要：在综述国内外水质评价研究的基础上，本文将遗传算法和人工神 经网络结合起来，建立了基于GA—BP算法的水质评价模型。文章首先系统的 阐明了BP神经网络与遗传算法的基本原理和学习训练过程，分析了BP网络 存在的缺陷，介绍了通用的一些改进措施。然后利用遗传算法全局搜索能力 强的特点与人工神经网络模型学习能力强的特点，设计出遗传神经网络的水 质评价模型。该模型不仅利用遗传算法对神经网络初始权重进行了优化，而 且采用了基于黄金分割理论的优化算法对网络模型的隐含层节点数进行了选 优，这样使模型在进行了较少次的循环迭代后训练误差收敛到最小，从而在 一定程度上克服神经网络模型训练中普遍存在的局部极小点问题。最后本文 将改进后的网络模型用于某地区的水环境质量评价，以某地区地表水的实测 数据作为模型输入，运用MATLAB软件进行仿真，对GA-BP水质评价模型进行 了验证，并与传统方法结果进行比较。结果表明优化后的GA-BP网络模型具 有稳定性好、鲁棒性高、识别能力强的优点，其结果与实际水质级别具有较 高的一致性，评价方法较为合理。 关键词：水质评价BP神经网络遗传算法 Optimization Optimization and Application of Water Quality Assessmental Model based on Genetic Algorithms and Artificial Neural Networks Major：Human Geography Author：Zhou Xueqian Tutor：Ding Hui Abstract：Based on a summary of water quality assessmental model’S study， this paper tries to establish GA—BP neural network model，which combines Artificial Neuarl Network(ANN)and Genetic Algorithms(GA)．Firstly，the paper clarifies the basic principle and the study process of BP neural network and Genetic Algorithms，analyses the existing weakness and introduces some general improvement measure．Secondly，making use of excellent global searching ability of GA and fine learning ability of ANN，we design the water quality assessmental model based on GA—BP algorithms．This model not only uses GA to optimize initial weights of neural network but also uses an arithmetic based on golden sectiong theory to optimize the amount of nodes in network’S covert layer．These make training error of the model got to minimal value after less times cyclic iteration．In a sense，local optimizing problems，which is widely existed in neural network model training，can be overcome．Finnally，based on the improved arithmetic，GA—BP model for water quality assessment was built，and the water environmental monitoring results of a reservoir was put into model．We validate the model and compare with the simulation of other general method by applying Ⅱ MATLAB MATLAB s