摘要
智能交通系统的目的在于研究设计科学且智能的交通系统,使城市的交通更为流畅,同时必须保证人们的出行更为安全与便捷。城市移动感知基于多种渠道收集到的城市范围内的感知数据,利用理论研究方法来从中提取发现出交通系统中重要、实时、准确的信息,例如电子地图、交通流量、交通灯状态、交通事故等。本文提出了基于安装有GPS设备的探测车辆的城市移动感知系统,并且详细介绍了两个移动感知算法,分别是道路识别和交通灯状态估计。
在城市感知系统中,和以往相关的研究中车辆GPS轨迹的高采样率、高精度特性不同的是,通过分析发现探测车辆的GPS轨迹数据在采样率、地理位置精度、朝向等维度都是粗粒度的,这为移动感知算法的设计带来了很大的挑战。
在基于探测车辆的道路识别中,为了快速准确地识别城市范围内形状各异的道路,本文设计了“粗粒度GPS轨迹数据剪枝,根据所属道路对GPS数据进行分类,形状敏感的B样条曲线的道路拟合”的三步骤方案。通过实验发现基于2,000辆车1.5小时内的GPS数据,本文提出的的算法对主干道的识别率高达93%,并且比已有的地图例如OpenStreetMap更为精确。
基于探测车辆的交通灯状态估计旨在估计交通灯在任意时刻的状态。通过观察和定量地分析,发现交通灯状态与车辆的运动状态有极强的联系。因此,本文提出了瞬时状态估计和连续状态估计的策略,并且针对性地设计了相应的启发式算法。通过实验发现在拥有2,000辆车的GPS数据时,城市范围内60%左右的交通灯的状态估计错误率低达19%。
Intelligent transportation systems aim to develop scientific and intelligent transportation, which will make urban traffic fluent and promise citizens easy and secure travel experience. Based on the collected sensing data, urban mobile sensing is proposed to extract real-time and accurate information in transportation system, i.e., digital map, traffic flow, traffic light state, accident. In this paper, we propose urban mobile sensing system based on probe vehicles, which are GPS equipped and generate GPS reports periodically. Moreover, we introduce two urban sensing algorithms, i.e., road recognition and traffic light state estimation.
Related urban sensing researches mostly employ GPS trace data with high sampling rate and high accuracy. However, through analysis we find that GPS trace data generated by our probe vehicles are coarse-grained in terms of sampling rate, GPS position, heading direction, and so on. This introduces considerable challenges in designing urban sensing algorithms.
Road recognition with probe vehicle aims to recognize city roads of various types. In our designed algorithm there are mainly three steps, i.e., pruning coarse-grained GPS trace data, clustering GPS data on the same road segment, and applying shape-aware B-spline fitting to generate roads. We conduct experiments and find that the coverage of arterial roads recognized reaches 93% with GPS data generated by 2,000 vehicles in 1.5 hours.
In traffic light state estimation, we aim to detect continuous states of road traffic lights. Through observation and study, we find that there is strong relationship between traffic light state and vehicle mobility. Thus, we introduce the strategy containing snapshot state estimation and pervasive state estimation, and design heuristic algorithms to solve the problem. We conduct experimental study and find that the estimation error rate of 60% of urban traffic lights is as low as 19%.
引文
[1] J. L. M. Vrancken and O. Kruse, "Intelligent control in networks: The case of road traffic management," 2006, pp. 308-311.
[2] Y. Zhao, "Mobile phone location determination and its impact on intelligent transportation systems," Intelligent Transportation Systems, IEEE Transactions on, vol. 1, pp. 55-64, 2000.
[3] C. R. Drane and C. Rizos, Positioning systems in intelligent transportation systems: Artech House, Inc., 1998.
[4] L. Figueiredo, et al., "Towards the development of intelligent transportation systems," 2001, pp. 1206-1211.
[5] Y. Zhao and A. House, "Vehicle Location and Navigation Systems: Intelligent Transportation Systems," Artech House, pp. 221-224, 1997.
[6] Z. Li, et al., "Compressive Sensing Approach to Urban Traffic Sensing," 2011, pp. 889-898.
[7] K. Shilton, et al., "Participatory privacy in urban sensing," 2008.
[8] A. T. Campbell, et al., "People-centric urban sensing," 2006, p. 18.
[9] M. Srivastava, et al., "Wireless urban sensing systems," 2006.
[10] Urban Sensing Project. Available: http://research.cens.ucla.edu/projects/2006/ systems/urban_sensing/.
[11] B. Coifman, "Improved velocity estimation using single loop detectors," Transportation Research Part A: Policy and Practice, vol. 35, pp. 863-880, 2001.
[12] K. Liu, et al., "Feasibility of using taxi dispatch system as probes for collecting traffic information," Journal of Intelligent Transportation Systems, vol. 13, pp. 16-27, 2009.
[13] R. L. Bertini and S. Tantiyanugulchai, "Transit buses as traffic probes: Use of geolocation data for empirical evaluation," Transportation Research Record: Journal of the Transportation Research Board, vol. 1870, pp. 35-45, 2004.
[14] S. Kamijo, et al., "Traffic monitoring and accident detection at intersections," IEEE Transactions on Intelligent Transportation Systems, vol. 1, pp. 108-118, 2000.
[15] L. Cao and J. Krumm, "From GPS traces to a routable road map," 2009, pp. 3-12.
[16] A. Fathi and J. Krumm, "Detecting road intersections from gps traces," Geographic Information Science, pp. 56-69, 2010.
[17]X. Liu, et al., "POVA: Traffic Light Sensing with Probe Vehicles," in Infocom 2012, Orlando, Florida USA, 2012.
[18] U. Today. (2009). More than 50 crashes on bay bridge curve. Available: http://www.usatoday.com/news/nation/2009-11-18-bay-bridge_N.htm.
[19] S. Schroedl, et al., "Mining GPS traces for map refinement," Data mining and knowledge Discovery, vol. 9, pp. 59-87, 2004.
[20] S. Worrall and E. Nebot, "Automated process for generating digitised maps through gps data compression," 2007.
[21] L. Zhang, et al., "Integration of GPS traces with road map," 2010, pp. 17-22.
[22] Y. Chen and J. Krumm, "Probabilistic modeling of traffic lanes from GPS traces," 2010, pp. 81-88.
[23] A. F. T. Committee, "Traffic management specification version 4.0," April, vol. 7, pp. 11-28, 1996.
[24] B. Toledo, et al., "Modeling traffic through a sequence of traffic lights," Physical Review E, vol. 70, p. 016107, 2004.
[25] Portland traffic signals optimization. Available: http://www.climatetrust.org/tr afficsignals.html.
[26] Signal optimization. Available: http://www.cityofseattle.net/transportation/
[27] E. Abbott and D. Powell, "Land-vehicle navigation using GPS," Proceedings of the IEEE, vol. 87, pp. 145-162, 1999.
[28] M. A. Turk, et al., "VITS-A vision system for autonomous land vehicle navigation," Pattern Analysis and Machine Intelligence, IEEE Transactions on, vol. 10, pp. 342-361, 1988.
[29] A. Balasubramanian, et al., "Interactive wifi connectivity for moving vehicles," 2008, pp. 427-438.
[30] X. Zhang, et al., "Study of a bus-based disruption-tolerant network: mobility modeling and impact on routing," 2007, pp. 195-206.
[31] V. Bychkovsky, et al., "A measurement study of vehicular internet access using in situ Wi-Fi networks," 2006, pp. 50-61.
[32] S. C. Nelson, et al., "Encounter-based routing in dtns," 2009, pp. 846-854.
[33] J. Jeong, et al., "TBD: Trajectory-based data forwarding for light-traffic vehicular networks," 2009, pp. 231-238.
[34] J. Zhao and G. Cao, "VADD: Vehicle-assisted data delivery in vehicular ad hoc networks," 2006.
[35] F. Y. Wang, et al., "Toward intelligent transportation systems for the 2008 Olympics," Intelligent Systems, IEEE, vol. 18, pp. 8-11, 2003.
[36] J. L. Adler and V. J. Blue, "Toward the design of intelligent traveler information systems," Transportation Research Part C: Emerging Technologies, vol. 6, pp. 157-172, 1998.
[37] Openstreetmap. Available: http://www.openstreetmap.org
[38] M. Haklay and P. Weber, "OpenStreetMap: User-generated street maps," Pervasive Computing, IEEE, vol. 7, pp. 12-18, 2008.
[39] R. Fiset, et al., "Map-image matching using a multi-layer perceptron: the case of the road network," ISPRS Journal of Photogrammetry and Remote Sensing, vol. 53, pp. 76-84, 1998.
[40] D. Ben-Arieh, et al., "Geometric modeling of highways using global positioning system data and B-spline approximation," Journal of transportation engineering, vol. 130, p. 632, 2004.
[41] Statistics researchers predict road traffic conditions. Available: http://domino. watson.ibm.com/comm/research.nsf/pages/r.statistics.innovation.traffic.html.
[42] V. Naumov and T. R. Gross, "Connectivity-aware routing (CAR) in vehicular ad-hoc networks," 2007, pp. 1919-1927.
[43] J. Harri, et al., "Mobility models for vehicular ad hoc networks: a survey andtaxonomy," Communications Surveys & Tutorials, IEEE, vol. 11, pp. 19-41, 2009.
[44] K. Lee, et al., "Max-contribution: On optimal resource allocation in delay tolerant networks," 2010, pp. 1-9.
[45] H. Y. Huang, et al., "Performance evaluation of SUVnet with real-time traffic data," Vehicular Technology, IEEE Transactions on, vol. 56, pp. 3381-3396, 2007.
[46] Y. Wang, et al., "Challenges and Opportunities in Exploiting Large-Scale GPS Probe Data," 2011.
[47] V. J. Lumelsky, "On fast computation of distance between line segments," Information Processing Letters, vol. 21, pp. 55-61, 1985.
[48] D. P. Huttenlocher, et al., "Comparing images using the Hausdorff distance," Pattern Analysis and Machine Intelligence, IEEE Transactions on, vol. 15, pp. 850-863, 1993.
[49] C. R. Houck, et al., "A genetic algorithm for function optimization: A Matlab implementation," NCSU-IE TR, vol. 95, 1995.
[50] D. Whitley, "A genetic algorithm tutorial," Statistics and computing, vol. 4, pp. 65-85, 1994.
[51]A. Likas, et al., "The global k-means clustering algorithm," Pattern recognition, vol. 36, pp. 451-461, 2003.