高铁建设环境影响评价EIA方法研究

发布时间：2018-03-02 16:46

本文选题：高铁建设　切入点：LM-BP人工神经网络　出处：《石家庄铁道大学》2014年硕士论文　论文类型：学位论文

【摘要】：近年来，我国的高速铁路建设迅猛发展，运营里程成倍增长，高铁的建设和开通运营有力的带动了沿线区域经济的发展，成为了区域发展的新动力，同时高铁的发展也给沿线的环境保持带来了巨大的压力。高铁项目环境影响评价作为线路选择、控制环境负面影响的关键性决策因素日益成为研究的热点。随着研究的深入，高铁环评在研究内容和程序上都有了进一步的完善与规范，然而对于使用复杂系统科学方法进行评价研究的却很少，而这种方法十分适合解决非线性的环境系统问题，其得出的评价结果在准确性、运算效率方面都非常有优势，高铁环评就属于这样的问题。因此，研究高铁建设环境复杂系统科学评价方法具有十分重要的意义。本课题主要研究如何进行高铁建设环境影响综合评价指标体系的建立和指标数量化计算以及改进过的BP人工神经网络在高铁建设环境影响综合评价中的应用等问题。文章阐述了本课题的国内外的研究进展，以及研究目的及方法。分析了高铁建设对环境的影响，依据相关评价标准建立了高铁建设环境影响综合评价的指标评价体系。介绍了人工神经网络的相关发展，重点阐述了BP人工神经网络的相关原理，探讨了将人工神经网络应用到高铁建设环境影响综合评价中的方法。本研究根据对既往已经开通运营的铁路项目对环沿线境影响的深入调查和研究，建立了高铁建设环境影响综合评价的指标体系，确定了每个指标的量化考核因素，并对指标值的量化计算进行了说明。根据改进过的BP人工神经网络建立了高铁建设环境影响的三层前馈神经网络综合评价模型，改进过的BP神经网络算法是在传统的算法的基础上使用LM算法，，即梯度法(最速下降法)和高斯—牛顿迭代法的结合，该种方法提高了收敛速度以及精确度。利用连盐城际铁路作为实例验证，在建立好的指标体系的基础上，深入分析该工程对环境影响的情况，设计出了针对连盐铁路的LM-BP神经网络评价模型，在MATLAB软件中通过BP网络工具箱的调用进行了仿真评价，评价结果真是有效，该模型的建立可提高评价结果的客观性和准确性。
[Abstract]:In recent years, the construction of high-speed railway in our country has developed rapidly and the mileage of operation has increased exponentially. The construction and opening of high-speed railway have powerfully driven the development of regional economy along the route and become the new driving force for regional development. At the same time, the development of high-speed rail also brings great pressure to the environmental maintenance along the railway line. As a route selection, environmental impact assessment (EIA) of high-speed rail project becomes a key decision-making factor to control the negative impact of the environment. With the development of the research, environmental impact assessment has become a hot topic. The research contents and procedures of high-speed rail environmental assessment have been further improved and standardized. However, the use of complex system scientific methods for evaluation research is rare, and this method is very suitable for solving nonlinear environmental system problems. The evaluation results obtained are very advantageous in terms of accuracy and operational efficiency, which is the problem of high-speed rail EIA. Therefore, it is of great significance to study the scientific evaluation method of complex system in the environment of high-speed rail construction. This paper mainly studies how to establish the index system of environmental impact comprehensive assessment of high-speed railway construction and how to calculate the index quantificationally, and how to apply the improved BP artificial neural network in the comprehensive assessment of environmental impact of high-speed railway construction. The article expounds the research progress of this subject at home and abroad, The paper analyzes the environmental impact of high-speed railway construction, establishes an index evaluation system for environmental impact assessment of high-speed rail construction, and introduces the related development of artificial neural network. The related principle of BP artificial neural network is expounded, and the method of applying artificial neural network to the comprehensive assessment of the environmental impact of high-speed railway construction is discussed. Based on the in-depth investigation and study of the impact of the railway projects that have been opened on the environment along the ring, the paper establishes an index system for the comprehensive assessment of the environmental impact of the high-speed rail construction, and determines the quantitative assessment factors for each index. Based on the improved BP artificial neural network, a three-layer feedforward neural network comprehensive evaluation model for the environmental impact of high-speed iron construction is established. The improved BP neural network algorithm is based on the traditional algorithm using LM algorithm, that is, the combination of gradient method (the steepest descent method) and Gao Si Newton iterative method. This method improves the convergence speed and accuracy. By using the inter-city railway with salt as an example, on the basis of establishing a good index system, the paper deeply analyzes the influence of the project on the environment. The evaluation model of LM-BP neural network for Lianyan railway is designed and simulated by calling BP neural network toolbox in MATLAB software. The evaluation result is really effective. The establishment of the model can improve the objectivity and accuracy of the evaluation results.
【学位授予单位】：石家庄铁道大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：X82

【参考文献】