为了应对层出不穷的逃费作弊手段,提升高速公路运营管理部门的逃费稽查水平,研究了基于车辆通行大数据的可疑逃费车辆的评价方法。在分析大量通行流水数据的基础上,给出了逃费车辆存在的11 种通行异常行为,提出了基于异常行为出现次数的车辆信用度评价模型,实现了对车辆逃费可疑度的量化。利用基于加权平均的多属性效用算法,计算出所有车辆的信用度值。经实际应用验证,正确率达33%,表明模型在较小程度上有一定的合理性和正确性。在此基础上,利用BP 神经网络算法对模型进行了改进。研究结果表明,利用改进模型推荐得到的逃费可疑车辆,稽查正确率上升至67%,可大幅提高稽查人员的工作效率,取得了比较满意的效果。
Abstract
In order to cope with the endless number of fare evasion cheating behavior, and improve the fare evasion inspection level of expressway operation and management department, an evaluation method to identify suspicious fare evasion vehicles based on vehicle traffic big data was studied. By analyzing a large number of traffic flow data, eleven kinds of abnormal traffic behavior of fare evasion vehicles were presented. A vehicle credit evaluation model based on the number of abnormal behavior was proposed, and the suspicious degree of vehicle fare evasion was quantified. Using the weighted average multi-attribute utility algorithm, the credit value of all vehicles was calculated. Being verified by actual application, the accuracy of the model reached 33%, which indicated the model was rational and correct on low degree. Then, the model was improved by using BP neural network algorithm. The research results showed that the accuracy of the improved model was up to 67%, which can improve the inspector′s work efficiency greatly and achieve satisfactory effect.
关键词
通行异常行为 /
车辆信用度 /
BP神经网络 /
逃费稽查 /
高速公路
Key words
abnormal traffic behavior /
vehicle credit /
BP neural network /
evasion inspection /
expressway
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 张友权. 浅谈高速公路车辆逃费的主要方式及其应对策略[J]. 北方交通,2011(8):66-68.
[2] 杨伟明. 高速公路联网收费逃费作弊情况分析及其对策[J]. 中国高新技术企业,2007(15):144-147.
[3] 杨淑芹. 联网收费防止利用通行卡逃漏费的途径和有效措施[J]. 交通世界(运输·车辆),2005(11):51-53.
[4] 潘亮华. 高速公路逃费手段及防治办法[J]. 中国交通信息化,2011(2):99-100.
[5] 韩慧英. 高速公路逃费探源[J]. 安全与健康, 2005(22):10.
[6] 唐州生. 高速公路车辆偷逃通行费的原因及应对措施[J]. 西部交通科技,2011(3):70-73.
[7] DELBOSC A, CURRIE G. Cluster Analysis of Fare Evasion Behaviours in Melbourne, Australia[J]. Transport Policy, 2016, 50: 29-36.
[8] GUARDA P, GALILEA P, PAGET- SEEKINGS L, et al. What is Behind Fare Evasion in Urban Bus Systems? An Econometric Approach[J]. Transportation Research Part A: Policy & Practice, 2016, 84: 55-71.
[9] TRONCOSO R, GRANGEE L D. Fare Evasion in Public Transport A: Time Series Approach[J]. Transportation Research Part A: Policy & Practice, 2017, 100: 311-318.
[10] JANKOWSKI W B. Fare Evasion and Noncompliance: A Game Theoretical Approach[J]. International Journal of Transport Economics, 1991, 18(3):275-287.
[11] 刁洪祥. ETC系统客户数据异常检测方法的研究[D]. 长沙:长沙理工大学,2004.
[12] 刁洪祥. 基于模糊C-均值聚类的ETC系统客户的逃费分析研究[J]. 企业技术开发,2005(10):8-10.
[13] 刁洪祥. 基于稳定遗传神经网络的ETC系统客户逃费分析[J]. 电脑与信息技术,2006(4):16-19.
[14] 刁洪祥,刘伟铭. 基于BP神经网络的ETC系统客户的流失分析研究[J]. 企业技术开发,2006(9):34-36.
[15] 刘勇,刁洪祥,刘伟铭. 基于改进的模糊决策树ETC系统客户欺诈分析研究[J]. 交通与计算机,2006(2):1-4.
[16] 赵彦,吴淑玲,林志恒,等. 高速公路通行卡逃费行为预测模型研究[J]. 中国科技论文,2015, 10(19):2245-2251.
[17] 张晓航. 高速公路联网收费稽查管理应用研究[D]. 西安:长安大学, 2010.
[18] 王小川. MATLAB神经网络43个案例分析[M]. 北京:北京航空航天大学出版社,2013.