Recognition Model of Highway Toll Evasion Behavior Considering Cost-Sensitivity

doi:10.12141/j.issn.1000-565X.230078

Abstract

Abstract:

In order to effectively improve the efficiency of highway vehicle toll evasion inspection, based on ETC （Electronic Toll Collection） toll data, this paper proposed a highway vehicle evasion recognition model by combining KNN (K-Nearest Neighbor), adaptive boosting (Adaboost) algorithm and cost-sensitive learning mechanism. Firstly, in view of the large volume and redundancy of the original ETC toll flow data, data discretization and standardization processing rules were developed to repair and standardize the data form, and then two types of toll evasion features were extracted. Secondly, seven types of toll evasion, such as large vehicles with small tags, were selected as the main research objects by analyzing the ETC data set. Thirdly, to address the problem of inefficient model classification due to the “high-dimensional” characteristics of the evasion data, the best subset of features showing the evasion characteristics was selected by Pearson and Spearman correlation analysis and ReliefF importance analysis. Fourthly, to address the model overfitting problem caused by the class “imbalance” between toll evasion vehicles and normal vehicles, KNN was used as the base classifier in the Adaboost algorithm, and the boundary ambiguity of different categories was alleviated through TomekLinks undersampling, then a cost-sensitive learning mechanism was introduced to improve the model’s emphasis on the minority class (toll evasion vehicles) to alleviate the tendency to discriminate the majority class (normal vehicles). Finally, the performance of the KNN-Adaboost model incorporating cost-sensitive learning mechanisms was verified by comparing the recognition effects of different classification models for various types of evasion events. The results show that the precision of the proposed model is 0.98, Recall is 0.96, F1-score is 0.97, and Kappa coefficient is 0.95, indicating that the proposed model can better solve the sample class imbalance problem than other models and has higher recognition accuracy for minority class，and it can be a reference for improving the efficiency of highway toll inspection.

Key words: highway transport, ensemble learning, machine learning, cost-sensitivity, feature selection

CLC Number:

U491

ZHAO Jiandong, XU Huiling, LÜ Xing, et al. Recognition Model of Highway Toll Evasion Behavior Considering Cost-Sensitivity[J]. Journal of South China University of Technology(Natural Science Edition), 2024, 52(5): 10-19.

Figures/Tables 14

Table 1

Table 2

Fig.1

Table 3

Fig.2

Table 4

Fig.3

Fig.4

Fig.5

Fig.6

Fig.7

Fig.8

Table 5

Fig.9

References 18

1	陈海亮，吴旭明．广东省高速公路ETC防逃费判别系统方案探讨［J］．中国交通信息化，2014（12）：62-65．
	CHEN Hailiang， WU Xuming ．Discussion on the scheme of ETC anti-evasion discriminating system for Guangdong expressway［J］．China ITS Journal，2014（12）： 62-65．
2	李小运．数据挖掘技术在高速公路联网收费稽查上的研究与应用［D］．广州：华南理工大学，2014．
3	马千惠．基于高速公路大数据的偷逃费车辆发现方法研究［D］．北京：北方工业大学，2019．
4	杨祥．基于高速公路多源大数据的双向倒卡车辆在线稽核方法研究［J］．中国市政工程，2022（3）：59-62．
	YANG Xiang ．Research on online audit method of two-way change cards vehicles based on multi-source big data of expressway［J］．China Municipal Engineering，2022（3）：59-62．
5	杨阳，李石磊，唐博文．高速公路套牌车稽查系统技术方法和策略研究［J］．汽车与安全，2022（3）：78-82．
	YANG Yang， LI Shilei， TANG Bowen ．Analysis on inspection methods and development trend of highway fake plate vehicles［J］．Auto & Safety，2022（3）：78-82．
6	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．
	ZHAO Yan， WU Shuling， LIN Zhiheng，et al ．Study on the prediction model of toll fraud behavior for highway pass card［J］．China Sciencepaper，2015，10（19）：2245-2251．
7	李松江，周舟，李岩芳，等．基于IGA-IBP算法的高速公路逃费预测［J］．计算机工程与设计，2018，39（12）：3840-3845．
	LI Songjiang， ZHOU Zhou， LI Yanfang，et al ．Prediction of highway escape cost based on IGA-IBP algorithm［J］．Computer Engineering and Design，2018，39（12）：3840-3845．
8	申长春．基于BP神经网络Bagging集成的高速公路绿通车稽查模型研究［D］．西安：长安大学，2018．
9	向红艳，杨朋涛，伊佳佳．基于RF-LR的高速公路逃费车辆状态预测模型［J］．重庆师范大学学报（自然科学版），2020，37（1）：75-80．
	XIANG Hongyan， YANG Pengtao， YI Jiajia ．State prediction model of expressway escaping vehicle based on RF-LR［J］．Journal of Chongqing Normal University （Natural Science），2020，37（1）：75-80．
10	杨朋涛．基于大数据的通行费异常车辆识别算法研究［D］．重庆：重庆交通大学，2020．
11	刘昱岗，郑帅，徐旭东，等．基于历史通行数据的假冒绿通车逃费行为预测［J］．公路交通科技，2021，38（4）：92-102，141．
	LIU Yugang， ZHENG Shuai， XU Xudong，et al ．Prediction of fake toll-free vehicle based on historical traffic data［J］．Journal of Highway and Transportation Research and Development，2021，38（4）：92-102，141．
12	MATEUSZ B， ATSUTO M， MACIEJ A M ．A systematic study of the class imbalance problem in convolutional neural networks［J］．Neural Networks，2018，106：249-259．
13	ZULFIQAR A， WASEEM S ．Comparative study of discretization methods on the performance of associative classifiers［C］∥Proceedings of 2016 International Conference on Frontiers of Information Technology （FIT）．Pakistan：IEEE，2016：87-92．
14	HOU X D， ZHANG T， JI L，et al ．Combating highly imbalanced steganalysis with small training samples using feature selection［J］．Journal of Visual Communication and Image Representation，2017，49：243-256．
15	ASIM K M， IDRIS A， IQBAL T，et al ．Seismic indicators based earthquake predictor system using genetic programming and AdaBoost classification［J］．Soil Dynamics and Earthquake Engineering，2018（111）：1-7．
16	陈琼，谢家亮．基于自适应采样的不平衡分类方法［J］．华南理工大学学报（自然科学版），2022，50（4）：26-34，45．
	CHEN Qiong， XIE Jialiang ．An imbalanced classification method based on adaptive sampling［J］．Journal of South China University of Technology （Natural Science Edition），2022，50（4）：26-34，45．
17	KANG D，OH S ．Balanced training/test set sampling for proper evaluation of classification models［J］．Intelligent Data Analysis，2020，24（1）：5-18．
18	KARCIOĞLU A A， BULUT H ．Performance evaluation of classification algorithms using hyperparameter optimization［C］∥Proceedings of 2021 6th International Conference on Computer Science and Engineering （UBMK）．Ankara：IEEE，2021：354-358．

字段名
进站名称	出站ETC卡类型	基础费用	支付类型
进站时间	通行介质	费用	跨省数量
进站车牌号	出站车种	折扣费用	通行次数
进站车牌颜色	出站名称	收费里程	OBU标识
进站车种	出站时间	费额	是否跨省
进站车型	出站车牌号	识别车牌号	交易类型标识
进站车货总重	出站车牌颜色	识别车牌颜色	超载率
进站车轴	出站车型	最短路径里程	行程时间
出口收费类型	出站车货总重	最小费额	行程速度
出站车轴	车辆标识	通行类型	通行时段

异常特征	统计规则
车轴一致性	出口车轴与入口车轴一致为0，不一致为1
车种一致性	出口车种与入口车种一致为0，不一致为1
车牌一致性	出口车牌号与入口车牌号一致为0，不一致为1
车型一致性	出口车型与入口车型一致为0，不一致为1
车牌颜色一致性	出口车牌颜色与入口车牌颜色一致为0，不一致为1
出入收费站一致	出入口收费站一致为1，不一致为0
短途重载	里程<50 km，车货总重大于空车和核定载重的80%
长途轻载	里程>300 km，车货总重小于空车和核定载重的30%
长途低费	除减免车，每公里费用小于收费标准为1，否则为0
超时车辆	时间超过里程/60 km/h+休息时间为1，否则为0

序号	逃费类型	逃费原理
1	大车小标（货）	逃车型、车种
2	大车小标（客）	逃车型、车种
3	倒卡逃费	逃距离
4	网内循环	逃距离
5	U形行驶	逃距离
6	套牌换卡	逃距离、逃车型、车种
7	甩挂逃费	逃距离、逃车型、车种

序号	取值范围		取值范围
1	0.00~0.23	5	32.25~32.28
2	0.23~0.53	6	32.28~154.00
3	0.53~1.00	7	154.00~∞
4	1.00~32.25

模型	评价指标
模型	AUC	K_a	精确率	召回率	F₁
NB	0.98	0.07	0.52	0.96	0.52
DT	0.86	0.68	0.85	0.83	0.83
RF	0.94	0.69	0.88	0.81	0.84
SVM	0.92	0.76	0.79	0.8	0.84
KNN	0.95	0.83	0.95	0.89	0.92
Adaboost	0.99	0.88	0.91	0.85	0.88
代价敏感的KNN-Adaboost	0.99	0.95	0.98	0.96	0.97