【作者】 李熙；

【导师】 朱晓宁； 蔡国强；

【作者基本信息】 北京交通大学 ， 交通运输规划与管理， 2011， 博士

【摘要】 当前,大力发展城市轨道交通,建立“多层次、立体化的综合交通体系”已成为我国交通发展的主要方向。从1969年10月开始,经过40多年的发展,我国已经进入大规模城市轨道交通建设阶段。目前,已有10余个城市建成城市轨道交通线路,50余个城市制定并在实施城市轨道交通规划,规划里程超过7500公里。“十二五”期间,规划建设城市快速轨道交通项目总长度将达1700公里。但是,在城市轨道交通快速发展的同时,安全形势却不容乐观——列车安全问题不断出现、事故常有发生,这些故障不但严重的影响到城轨车辆的行车安全,一旦引发事故将会带来巨大的人员伤亡和经济损失。城市轨道交通车辆安全运行的基础在于车辆的关键部件,而走行部更是重中之重。为提高城市轨道交通运营的安全性,本文在对走行部原理及数据分析基础上,对其安全评估展开以下研究：1.系统地分析了影响城市轨道交通车辆走行部安全的人员及组织管理因素、环境因素、线路基础设施、走行部内部原因、维护检修及构建了由评估层次、方法体系、安全等级以及评估流程等方面构成的走行部安全评估体系结构。2.建立了城市轨道交通车辆走行部的改进概率影响图模型,该方法提高了运算针对性和效率,取得良好的效果,针对分析结果反映出的问题对城市轨道交通公司的工作尤其是运营维护工作提出了合理化建议。3.采用改进层次分析法分析轮对轴箱装置、构架、弹簧悬挂装置、中央牵引装置、基础制动装置、驱动装置及相互之间的重要度,采用模糊综合评价方法建立城市轨道交通车辆走行部综合安全评估模型进行评估。在改进层次分析法中,提出采用聚类分析采集模糊风险评价集中的风险隶属度矩阵。针对评估结果探讨了采用风险控制措施降低风险的方法。4.提出了基于概率统计推理理论的双向安全评估方法并建立概率安全模型,并通过相应模型算法实现安全概率计算。该方法基于可靠性分析GO法计算分析过程中的状态流概率,对系统正向进行安全评估,逆向进行故障概率判断定位,实现运行系统-部件的综合概率安全评估。利用提出的方法对城市轨道交通车辆走行部进行了安全概率计算分析。5.针对城市轨道交通车辆走行系统的特点,提出通过采用传感器获取在城市轨道交通车辆走行部的振动信号,利用小波分析法、快速傅里叶变换和神经网络结合对振动速度、加速度等信号进行特征提取的方法,经过实际验证效果显著。更多还原

【Abstract】 At present, it becomes the main trend of transport development to vigorously develop urban rail transit and to establish a "multi-level, three-dimensional integrated transport system ". From October 1969, through 40 years, China has stepped onto the stage of large-scale urban rail transit construction. Currently, rail transit lines have been built in more than 10 cities. Urban rail transportation planning has been developed and under implementation in more than 50 cities with the planning mileage over 7500 km. In the " Twelfth Five-Year " period, the planning and construction of urban rail transit lines will reach the total length of 1,700 kilometers.However, with the rapid development of urban rail transit, the situation of its safety needs more attention, because safety issues and accidents often occur. These failures will not only seriously affect the safety of urban rail traffic vehicles, but also soon lead to huge accident human and economic losses. Safety of urban rail transit vehicles is on the basis of key components of vehicles with running gear as the top priority. In order to improve the safety of urban rail transit operations, through historical data and structural principle analysis, this article takes the vehicle running gear as the research and safety assessment object and carries out the following research:1.The paper systematic analyzes the safety influence parameters of urban rail transit including staff and management, environments, infrastructures, running gear internal factors and maintenance. Meanwhile, in this paper the author constructs the safety assessment architecture of running gear which includes assessment layers, assessment methodologies, assessment flow and safety levels.2.Improved probabilistic influence diagram is applied to construct safety assessment model of urban rail transit vehicle running gear, which improved the targeting and efficiency of operations. This paper proposes reasonable proposals to urban rail companies especially the maintenance department according to the problems reflected in the analysis results.3.Improved analytic hierarchy process (AHP) was implemented to analyze the importance of different parts of running gear. On this basis, it establishes the urban rail transit vehicles running gear fuzzy safety assessment model and gathers risk matrix of fuzzy risk assessment through clustering analysis in AHP. Finally, risk control methods are discussed to reduce risks of the assessment results.4.It proposes a safety assessment method based on the theory of probability statistics reasoning, establishes probability safety model and calculated safety probability through the appropriate model algorithm, which is on the basis of state flow probability in the analysis process of reliability analysis Go methodology which can help judge and locate of inverse probability for system failure and realized a comprehensive probabilistic safety assessment of operational system-parts. And it calculates the safety probability analysis of urban rail transit vehicles running gear in the proposed method.5.Aiming at the characteristics of urban rail transit vehicle running systems, it proposes to acquire vibration signals of urban rail transit vehicles running gear through the sensors, combines wavelet analysis with Fast Fourier Transform and neural network to analyze the vibration velocity or acceleration signals for feature extraction, realizes the purpose of safety assessment of running gear, and achieves remarkable results by actual tests.更多还原