高速公路交通量预测及山区高速公路后评价研究

发布时间：2018-01-06 02:27

本文关键词：高速公路交通量预测及山区高速公路后评价研究　出处：《大连理工大学》2015年博士论文　论文类型：学位论文

【摘要】：山区高速公路后评价有利于发现项目建设过程中的缺点和不足,对于探索解决问题,改善项目管理现状,提高决策、建造及管理水平具有重要的现实意义；对路段上的交通量进行准确预测与分析,提高远景交通量的预测精度,可以在很大程度上减少因规划设计失误而造成的损失和浪费。为了提高山区高速公路建设项目决策、建造及运营管理水平,本文对高速公路交通量预测模型、山区高速公路后评价指标体系及综合评价方法进行探索和研究,具体研究内容如下： 1)基于Cat映射理论和GA (Genetic Algorithm)算法演化机制,提出新混沌自适应遗传算法(New Chaotic Adaptive Genetic Algorithm, NCAGA),用于RSVR (Support Vector Regression)模型参数优选,设计新混沌自适应遗传算法优化RSVR模型参数机制,利用主成分分析法(Principal Component Analysis, PCA)和相关性分析法,确定RSVR模型的输入向量,提出PRSVR-NCAGA高速公路交通量预测模型。基于高速公路交通量数据,开展预测研究,检验提出算法和预测模型的可行性和优越性。 2)针对山区高速公路自然生态环境脆弱和交通安全评价特殊性,建立了包含3个二级评价指标、14个三级评价指标的山区高速公路环境影响后评价指标体系和包含7个二级评价指标、30个三级评价指标的山区公路建设项目交通安全评价指标体系。结合各评价指标值的特点,基于隶属度函数,设计用于确定各评价指标的调查问卷,采用调查问卷的方式获取评价指标值,通过开展实证评价,检验建立的评价指标体系的合理性和评价指标获取方式的可行性。 3)基于AHP (Analytic Hierarchy Process)法和模糊理论,提出AHP-FUZZY多级山区高速公路评价模型。结合后评价指标体系建立面向山区特性的山区高速公路后评价方法。基于实际工程开展实证评价研究,完善提出的评价方法,验证提出的评价方法的可行性以及评价结果的可靠性。 4)基于提出的山区高速公路后评价方法,对实际山区高速公路进行后评价实证研究,结合评价结果,揭示评价对象存在的问题,给出具体合理化应对措施。
[Abstract]:The mountainous area highway post evaluation for the discovery of the project construction process of the shortcomings and deficiencies, to explore ways to solve problems, improve project management, improve decision-making, has important practical significance to the construction and management level; the accurate prediction and analysis of traffic on the road, improve the prediction accuracy of traffic in the future, can be reduced due to the planning and design of failure loss and waste to a great extent. In order to improve the highway construction project in decision-making, construction and operation management level, this article on highway traffic volume forecasting model of mountain highway post evaluation index system and comprehensive evaluation method of exploration and research, the main research contents are as follows:
1) Cat mapping theory and GA (Genetic Algorithm) algorithm based on chaotic evolution mechanism, put forward a new adaptive genetic algorithm (New Chaotic Adaptive Genetic Algorithm, NCAGA), RSVR (Support Vector Regression) is used for parameter optimization, parameter optimization mechanism of RSVR model to design new chaotic adaptive genetic algorithm, using principal component analysis (Principal Component Analysis PCA), and correlation analysis, to determine the input vector of RSVR model, the PRSVR-NCAGA highway traffic forecast model. Based on the data of highway traffic, to carry out prediction research, test the feasibility and superiority of the proposed algorithm and the prediction model.
2) evaluation according to the particularity of the natural ecological environment is fragile and the mountainous area highway traffic safety, including 3 level two indexes, the index system and contains 7 level two indexes evaluation of highway environment of the mountain area 14 level three indexes, the traffic safety evaluation index system of highway construction project 30 three grade evaluation according to the characteristics of each index. The evaluation index values, based on membership function is designed to determine the evaluation index of the questionnaire, using the questionnaire for the evaluation index value, through empirical evaluation, feasibility test to establish evaluation index system is reasonable and evaluation index of access methods.
3) based on AHP (Analytic Hierarchy Process) method and fuzzy theory, and presents the AHP-FUZZY evaluation model is established. The multistage mountain mountain mountain highway oriented characteristics of the post evaluation index system evaluation combined. Research on empirical evaluation of actual engineering based on the evaluation method of improving the reliability evaluation method to validate the feasibility and evaluation of the results.
4) based on the post evaluation method of mountainous expressway, the post evaluation empirical study of real mountainous expressway is carried out, combined with the evaluation results, the problems of evaluation objects are revealed, and specific and reasonable countermeasures are given.

【学位授予单位】：大连理工大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：U491

【参考文献】