An Evaluation Method of Vehicle Handling Consistency Based on Improved DTW Algorithm

doi:10.12382/bgxb.2023.1104

Abstract

Abstract:

Time series matching technology is widely used in vehicle handling consistency evaluation.An evaluation method of vehicle throttle control action consistency based on segmented dynamic time warping (DTW) is proposed for the motion consistency evaluation of a coach vehicle in the process of cooperative driving.On the basis of the inconsistent number of sample data points in the cooperative control test of coach vehicle and the slope constraint of dynamic bending path, the traditional DTW rectangular search area is transformed into a parallelogram search area to reduce the area of the search area by changing the slope of search path, thus greatly reduce the calculation amount.Four groups of typical throttle action curves are selected to carry out 50 rounds of iterative experiments for verification, and the DTW distance matrix between the action curves of real vehicles A and B is calculated by segmented DTW method.A minimum deviation leveling method is used to combine the cluster objects for the actions of vehicles A and B, so as to complete the consistency evaluation of throttle action data.The experimental results show that the average matching accuracy of the improved DTW algorithm in each throttle action can reach 89.2%, which is about 3.2% higher than that of the single DTW algorithm, and the average matching time is about 92.45s, which is about 12.6% lower, thus verifying the feasibility and superiority of the segmented DTW algorithm in the consistency evaluation of throttle action.

Key words: vehicle, cooperative driving, throttle action curve, consistency evaluation, segmented DTW

CLC Number:

ZHONG Yue, XU Feng, YAN Mengfei, LIU Yile. An Evaluation Method of Vehicle Handling Consistency Based on Improved DTW Algorithm[J]. Acta Armamentarii, 2025, 46(1): 231104-.

Figures/Tables 14

Fig.1 Throttle action curve of vehicle A

Fig.2 Throttle action curve of vehicle B

Fig.3 Dynamic bending distance path diagram

Fig.4 DTW parallelogram search area

Fig.5 Original action curves of Groups A1 and A2 and their action curves after improved dynamic bending

Fig.6 Raw data and DTW distances of Groups B1 and B3 action curves

Fig.7 Raw data and DTW distances of Groups A3 and B1 action curves

Fig.8 Raw data and DTW distances of Groups A2 and B2 action curves

Fig.9 Comparison of matching accuracies before and after DTW improvement

Fig.10 Comparison of matching times before and after DTW improvement

Table 1 The matching accuracies of DTW algorithm before and after improvement

匹配算法	平均匹配精度
匹配算法	A1A2	B1B3	A3B1	A2B2
DTW	0.866	0.876	0.844	0.85
分段式DTW	0.894	0.912	0.88	0.882

Table 2 The matching times of DTW algorithm before and after improvement

匹配算法	平均匹配时间/s
匹配算法	A1A2	B1B3	A3B1	A2B2
DTW	96.48	79.88	126.86	119.9
分段式DTW	89.5	76.46	105.36	98.46

Table 3 DTW distances between the throttle actions of two vehicles

曲线编号	A1	A2	A3	A4	B1	B2	B3	B4
A1	0				A组曲线的平均DTW距离:21.761 B组曲线的平均DTW距离:2.547 AB曲线的平均DTW距离:26.692
A2	9.382	0
A3	9.544	12.734	0
A4	38.835	36.698	23.374	0
B1	9.76	27.777	15.559	52.053	0
B2	6.6019	20.42	15.297	52.633	2.104	0
B3	8.2896	24.6453	16.960	53.585	1.757	1.375	0
B4	12.439	29.7243	20.105	61.220	2.116	4.669	3.260	0

Fig.11 DTW distance clustering spectrum of throttle action curve

References 38

[1]	JAFARI B, MASHADI B. Valve control of ahydraulically interconnected suspension systemto improve vehicle handling qualities[J]. Vehicle System Dynamics, 2023, 61(4):1011-1027.
[2]	RATHORE G S, CHAWLA S. Design and development of rack and pinion steering gear assembly for passenger car vehicle[C]//Proceedings of Symposium on International Automotive Technology.Chicago,IL, US: Society of Automotive Engineers, 2024.
[3]	ISO 17288-2: 2004.Passenger cars-free-steer behaviour-Part2:steer-pulse open-loop test method[S]. Geneva, Switzerland: International Organization for Standardization,2004.
[4]	ISO 3888-1:1999.Passenger cars-test track for a severe lane-change manoeuvre Part1:double-lane change[S]. Geneva, Switzerland: International Organization for Standardization,1999.
[5]	VÖRÖS I, OROSZ G, TAKÁCS D. On the global dynamics of path-following control of automated passenger vehicles[J]. Nonlinear Dynamics, 2023, 111(9):8235-8252.
[6]	LO S C. Classification of driving behavior by pattern recognition in multiclass users traffic flow[C]//Proceedings of AIP Conference Proceedings.American Institute of Physics, 2007, 963(2):989-992.
[7]	WANG J, WANG Q N, WANG P Y, et al. Adaptive shift control strategy based on driving style recognition[C]//Proceedings of SAE/KSAE 2013 International Powertrains,Fuels & Lubricants Meeting. Seoul,Korea: Society of Automotive Engineers,2013.
[8]	ZHANG H, WANG L K, SHI W X. Seismic control of adaptive variable stiffness intelligent structures using fuzzy control strategy combined with LSTM[J]. Journal of Building Engineerng, 2023,78:107549.
[9]	LEE T K, FILIPI Z. Real-world driving pattern recognition for adaptive HEV supervisory control:Based on representative driving cycles in Midwestern US[C]//Proceedings of SAE 2012 World Congress & Exhibition:Detroit, MI,US: Society of Automotive Engineers,2012.
[10]	XU B, WANG H C. A comparative analysis of adaptive energy management for a hybrid electric vehicle via five driving condition recognition methods[J]. Energy, 2023,269:126732.
[11]	ZONTONE P, AFFANNI A, BERNARDINI R, et al. Analysis of physiological signals for stress recognition with different car handling setups[J]. Electronics, 2022, 11(6):888.
[12]	WEBB G I, PETITJEAN F. Tight lower bounds for dynamic time warping[J]. Pattern Recognition, 2021,115:107895.
[13]	GE L, CHEN S. Exact dynamic time warping calculation for weak sparse time series[J]. Applied Soft Computing, 2020,96:106631.
[14]	程前, 李建良. 基于改进动态时间规整的相似性度量及轨迹聚类[J]. 电子测量技术, 2021, 44(23):1-5.
	CHENG Q, LI J L. Similarity measurement and trajectory clustering based on improved dynamic time warping[J]. Electronic Measurement Technology, 2021, 44(23):1-5. (in Chinese)
[15]	邓青, 薛青, 罗佳. 基于支持向量机的坦克驾驶模拟训练结果分析[J]. 兵工学报, 2019, 40(9):1953-1960. doi: 10.3969/j.issn.1000-1093.2019.09.021
	DENG Q, XUE Q, LUO J. Analysis of tank driving simulation training results based on support vector machine[J]. Acta Armamentarii, 2019, 40(9):1953-1960. (in Chinese) doi: 10.3969/j.issn.1000-1093.2019.09.021
[16]	LOBATO P, RAYNO M, DAILY J, et al. Quantifying repeatability of real-world on-road driving using dynamic time warping[C]//Proceedings of WCX World Congress Experience.Chicago,IL, US: Society of Automotive Engineers, 2022.
[17]	郭沐锋, 彭文成, 李平华. 某型坦克驾驶教练车在役考核指标体系研究[J]. 工程与试验, 2022, 62(3):40-42.
	GUO M F, PENG W C, LI P H. Study on in-service assessment index system of a coach tank[J]. Engineering & Test, 2022, 62(3):40. (in Chinese)
[18]	郭孔辉, 金凌鸽, 曹宇, 等. 车辆操纵稳定性客观评价指标的降维处理[J]. 汽车技术, 2010(2):1-4.
	GUO K H, JIN L G, CAO Y, et al. Dimensional reduction of objective evaluation criteria of vehicle handling behavior[J]. AutomobileTechnology, 2010(2):1-4. (in Chinese)
[19]	刘义乐, 刘峻岩. 坦克驾驶协同动作模式表示与匹配方法[J]. 装甲兵工程学院学报, 2019, 33(3):63-67.
	LIU Y L, LIU J Y. Representation and matching method of tank driving cooperative action mode[J]. Journal of Academy of Armored Force Engineering, 2019, 33(3):63-67. (in Chinese)
[20]	CHIBA S, SASAOKA H. Effectiveness of transfer learning in autonomous driving using model car[C]//Proceedings of the 2021 13rd International Conference on Machine Learning and Computing.New York,NY, US: Association for Computing Machinery,2021:595-601.
[21]	龚诗雄, 王旭, 孔国杰, 等. 多车协同目标跟踪方法[J]. 兵工学报, 2022, 43(10):2429-2442.
	GONG S X, WANG X, KONG G J, et al. Methods for multi-vehicle cooperative object tracking[J]. Acta Armamentarii, 2022, 43(10):2429-2442. (in Chinese) doi: 10.12382/bgxb.2021.0462
[22]	JOHNPAUL C I, PRASAD M V N K, NICKOLAS S, et al. Fuzzy representational structures for trend based analysis of time series clustering and classification[J]. Knowledge-Based Systems, 2021, 1(7):106991.
[23]	李正欣, 张凤鸣, 李克武, 等. 一种支持DTW距离的多元时间序列索引结构[J]. 软件学报, 2014, 25(3):560-575.
	LI Z X, ZHANG F M, LI K W, et al. Inde--x structure for multivariate time series under DTW distance metric[J]. Journal of Software, 2014, 25(3):560-575. (in Chinese)
[24]	BERNTD D J, CLIFFORD J. Using dynamic time warping to find patterns in time series[J]. Workshop on Knowledge Knowledge Discovery in Databases, 1994,398:359-370.
[25]	刘美, 王全民. 基于DTW的时间序列相似度量方法的优化[J]. 计算机与数字工程, 2023, 51(4):814-819,938.
	LIU M, WANG Q M. Optimization of time series similarity measurement based on DTW[J]. Computer & Digital Engineering, 2023, 51(4):814-819,938. (in Chinese)
[26]	杨雄, 郭佳豪, 方鑫, 等. 基于改进动态时间规整算法的终端波形比对方法[J]. 电子测量技术, 2023, 46(6):178-184.
	YANG X, GUO J H, FANG X, et al. Method based on optimized dynamic time warping algorithm for waveform comparison in terminal[J]. Electronic Measurement Technology, 2023, 46(6):178-184. (in Chinese)
[27]	陈乾, 胡谷雨. 一种新的DTW最佳弯曲窗口学习方法[J]. 计算机科学, 2012, 39(8):191-195.
	CHEN Q, HU G Y. New learning method for optimal warping window of DTW[J]. Computer Science, 2012, 39(8):191-195. (in Chinese)
[28]	朱洪涛, 李姗, 肖勇, 等. 基于动态时间弯曲的轨道波形匹配方法[J]. 振动与冲击, 2018, 37(11):246-251.
	ZHU H T, LI S, XIAO Y, et al. WEI H. Rail waveform matching method based on dynamic time warping[J]. Journal of Vibration and Shock, 2018, 37(11):246-251. (in Chinese)
[29]	杨靖. 基于动态时间弯曲的时间序列相似性搜索技术的研究[D]. 哈尔滨: 哈尔滨工业大学, 2015.
	YANG J. Research of similarity search based on dynamic time warping in time aeries data mining[D]. Harbin:Harbin Institute of Technology, 2015. (in Chinese)
[30]	JANG M K, HWANG S J, KIM J H, et al. Overturning and rollover characteristics of a tractor through dynamic simulations:effect of slope angle and obstacles on a hard surface[J]. Biosystems Engineering, 2022,219:11-24.
[31]	SOLEIMANI G, ABESSI M. DLCSS:a new similarity measure for time series data mining[J]. Engineering Applications of Artificial Intelligence, 2020,92:103664.
[32]	夏磊. 基于行为识别匹配的用户日常运动检测[J]. 计算机测量与控制, 2018, 26(2):264-267.
	XIA L. Detection of user’s daily motion based on behavior recognition matching[J]. Computer Measurement & Control, 2018, 26(2):264-267. (in Chinese)
[33]	马晓蒙. 基于DTW模板匹配的外骨骼机器人步态识别与预测研究[D]. 武汉: 武汉理工大学, 2019.
	MA X M. Research on gait recognition and prediction of exoskeleton Robot based on DTW template matching[D]. Wuhan: Wuhan University of Technology, 2019. (in Chinese)
[34]	FOTSO V S S, NGUIFO E M, VASLIN P. Grasp Heuristic for time series compression with piecewise aggregate approximation[J]. Rairo-Operations Research, 2019, 53(1):243-259.
[35]	崔文锋. 智能汽车综合纵向控制的统一驾驶员模型研究[D]. 长春: 吉林大学, 2018.
	CUI W F. Research on unified driver model of comprehensive longitudinal control for intelligent vehicles[D]. Changchun: Jilin University, 2018. (in Chinese)
[36]	KEOGH E J, PAZZANI M J. Scaling up dynamic time warping to massive datasets[C]//Proceedings of Principles of Data Mining and Knowledge Discovery: 3 rd European Conference.Berlin,Heidelberg,Gemany:Springer,1999:1-11.
[37]	ZHAO J, ITTI L. shapeDTW:Shape dynamic time warping[J]. Pattern Recognition, 2018,74:171-184.
[38]	HONG J Y, PARK S H, BAEK J G. SSDTW:Shape segment dynamic time warping[J]. Expert Systems with Applications, 2020,150:113291.