【作者】 罗建科；

【导师】 罗霞；

【作者基本信息】 西南交通大学 ， 交通运输规划与管理， 2011， 硕士

【摘要】 随着我国高速公路建设的快速发展与高速公路网的逐步完善,公路管理部门的工作重心将逐步从公路网的建设转向对已有公路的养护、改建、和维护已有高速公路网上,高速公路路面管理系统的重要性也就凸显出来。因此,本文对高速公路路面管理系统展开了研究,研究了高速公路路面的破损区分类及成因分析,高速公路路面状况评价和路面使用性能预测,重点研究神经网络理论在高速公路路面管理系统中的应用。高速公路路面使用性能的评价和路面使用性能预测是路面管理系统的重要组成部分,是合理的制定路面养护维修计划和资金分配方案的基础。论文首先对高速公路路面的结构组成和划分进行分类,并分别对柔性路面和刚性路面的常规病害及成因进行了分析并给出养护方案。其次对高速公路路面使用性能单项评价指标进行了分析,论述了现行的路面使用性能综合评价方法,针对当前路面检测数据呈模糊与非线性的特点,结合已有的高速公路路面质量评价体系,建立了基于模糊神经网络的综合评价方法。最后在高速公路路面使用性能预测方面,系统总结了国内外使用性能预测模型的研究现状,分析了路面使用性能的影响因素及变化趋势,本着利用组合预测方法提高预测精度和降低预测风险的思想,首先采用灰色预测模型与BP神经网络模型进行单项预测,再建立基于熵值法的组合预测模型。通过实例验证,本文的组合预测模型在方法上和理论上是可行的,为高速公路路面管理者在进行路面的养护管理决策时能够提供新的思路和帮助。更多还原

【Abstract】 With the gradual improwement of Highway Nerwork in China, The Road management department work focus will from large-scale construction road network to conserve.rebuild,maintain and improve the existing road network.,whick highly show the importance of pavement management system.Therefore，this paper does some researches on the highway pavement pavement management system, including:study the damaged area highway pavement types and reasons, road conditions evaluation and pavement performance prediction, key research neural network theory in the highway pavement management system application.Pavement performance of the evaluation and prediction is an important part of pavement management system, it is reasonable formulate pavement maintenance maintenance plan and capital allocation scheme foundationo This paper first analyze highway pavement structure and division of classification, respectively for flexible pavement and rigid pavement of conventional diseases and causes and then give the maintenance plan.Secondly, the paper analyze the highway pavement performance single-item indicators, discusses the current pavement performance comprehensive evaluation method, in view of the current road test data is fuzzy and nonlinear characteristics, established based on fuzzy neural network of comprehensive evaluation method.Finally, in the highway pavement performance prediction, the research in China and abroad are summarized application performance prediction model, analyses the influence factors of pavement performance and changing trends.The paper use combination forecast method increase the prediction accuracy of thought, firstly by using grey prediction model and the BP neural network model, then based on entropy method of combination forecast model,use the combination-weighted coefficient。Through examples, this paper combined forecasting model on the method and theory is feasible, for highway pavement managers in pavement maintenance and management of decisions can provide a new thought and help.更多还原