信号交叉口单车停车延误时间与停车位置的关系模型研究

发布时间：2018-06-12 02:05

本文选题：浮动车 + 信号配时参数估计　；参考：《重庆大学》2014年硕士论文

【摘要】：车辆通过信号交叉口时通常会经历一定的延误，对车辆在信号交叉延误时间的有效估计对于交通管理、公交调度、信号灯配时参数优化以及出行者的路径选择等都具有重要的意义。但是由于检测手段、技术条件等限制，，尚缺乏操作性较强的信号配时参数估计方法以及信号交叉口单车延误时间模型的分析方法。信号交叉口的单车延误时间与信号配时参数以及车辆的减速停车的位置有关，针对上述问题本文首先考虑到浮动车GPS数据的特点和便利性，通过深入分析车队经过信号交叉口的延误特性，得到车队的延误模式的描述，并在此基础上，研究了交叉口信号配时参数的估计方法。最后，在深入分析单车停车延误时间与停车位置关系的基础上，利用信号配时参数估计结果，考虑到异常数据的影响，采用稳健回归估计，建立了单车停车延误时间与停车位置的关系模型。研究工作主要包括以下三方面内容：首先，研究了基于GPS数据的单车停车延误数据获取方法，针对传统减速点法不能判断车辆多次停车的情况，做出了相应的改进。利用相邻两个GPS数据点的平均间距（平均速度）判断车辆的关键位置点，进而获取车辆每次停车的位置以及停车延误时间。结果表明了该方法的实用性和有效性，为后续的研究工作奠定了数据基础。其次，针对浮动车GPS数据覆盖率较低的情况，研究了一种基于停车延误数据的交叉口信号配时参数估计方法。通过分析车队经过信号交叉口的延误特性，得到车队延误模式三角形，并在此基础上分别讨论了未饱和与过饱和状态下的交叉口信号参数估计方法。最后通过多周期估计结果融合的方法，减少了因GPS定位误差、数据缺失以及其他偶然因素引起的估计误差。结果表明该方法具有较好的估计效果，而且能够适应于不同类型的信号交叉口。最后，针对GPS存在定位误差、数据缺失等问题以及车辆存在偶然停车等情况对延误数据造成的污染，利用稳健回归的思想，采用最小截尾二乘LTS回归，分时段分别建立了未饱和与过饱和状态下的单车的停车延误时间与停车位置之间的关系模型。结果表明该模型具有较好的估计效果。研究结果表明本文的交叉口信号配时参数估计方法和单车的停车延误时间与停车位置的关系模型估计精度较高，且具有实用价值。
[Abstract]:Vehicles usually experience certain delays when they pass signalized intersections. The effective estimation of vehicle delay time at signalized intersections is useful for traffic management and bus dispatch. The optimization of signal timing parameters and the path selection of travelers are of great significance. However, due to the limitations of means of detection, technical conditions and so on, There is still a lack of operable signal timing parameter estimation method and the analysis method of signal intersection bicycle delay time model. The signal timing parameter and vehicle deceleration stop position are related to the single cycle delay time at signalized intersection. Considering the characteristics and convenience of the floating vehicle GPS data, this paper firstly analyzes the delay characteristics of the vehicle passing signal intersection, and gets the description of the delay mode of the vehicle, and on this basis, The estimation method of signal timing parameters at intersection is studied. Finally, on the basis of deeply analyzing the relationship between parking delay time and parking position, using the estimation results of signal timing parameters and considering the influence of abnormal data, robust regression estimation is used. The relationship between parking delay time and parking position is established. The research work mainly includes the following three aspects: firstly, the acquisition method of bicycle parking delay data based on GPS data is studied, and the corresponding improvement is made in view of the traditional deceleration point method can not judge the situation of vehicle stopping many times. Using the average distance (average velocity) of two adjacent GPS data points to judge the key points of the vehicle, the location of each stop and the delay time of the vehicle are obtained. The results show the practicability and effectiveness of the method, and lay the data foundation for the further research. Secondly, aiming at the low GPS data coverage of floating vehicle, A method for estimating intersection signal timing parameters based on parking delay data is studied. Based on the analysis of the delay characteristics of the vehicle passing signalized intersection, the triangle of the vehicle delay mode is obtained, and the signal parameter estimation methods of the intersection under the condition of unsaturated and supersaturated are discussed respectively. Finally, the estimation error caused by GPS positioning error, missing data and other accidental factors is reduced by the method of fusion of multi-period estimation results. The results show that the proposed method has good estimation effect and can be applied to different signal intersections. Based on the problem of missing data and the pollution caused by accidental parking of vehicles on the delay data, using the idea of robust regression, the least truncated two times LTS regression is adopted. The relationship between the delay time and the parking position of the cycle in unsaturated and supersaturated state is established in different time periods. The results show that this model has a good estimation effect, and the research results show that the method of signal timing estimation and the relationship between stopping delay time and parking position of the vehicle in this paper have high accuracy and practical value.
【学位授予单位】：重庆大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U491.54

【参考文献】