考虑换道意图的LSTM-AdaBoost车辆轨迹预测模型

doi:10.3778/j.issn.1002-8331.2111-0052

摘要/Abstract

摘要： 不合理的车辆的换道行为是导致交通事故发生的主要原因之一，提前预知换道车辆的轨迹并及时做出相应调整有助于减少事故的发生。针对换道车辆轨迹预测问题，采用将深度学习和集成学习相结合的轨迹预测方法，并考虑了换道意图的影响。建立连续隐马尔可夫模型对车辆进行换道意图检测，提前判别车辆的换道状态，并输入至相应的轨迹预测模型中；将LSTM（long short term memory）作为AdaBoost算法（adaptive boosting）的基预测器，建立LSTM-AdaBoost模型，在多个基预测器同时进行轨迹预测的基础上，通过训练调整各个基预测器的权重并将结果加权集成，提升预测模型的精度和稳定性；通过NSGIM（next generation simulation）数据集对模型进行训练和测试，结果显示意图预测模型在变道前一秒的准确率在90%以上，LSTM-AdaBoost集成轨迹预测模型与单一的LSTM模型相比精度和稳定性显著提升，且预测结果中异常数据更少，具有较好的稳定性；同时预测对比结果也表明增加意图预测模块有助于提升换道轨迹预测的精度。

关键词: 车辆换道轨迹预测, 换道意图识别, 隐马尔可夫模型, 长短期记忆网络, AdaBoost集成算法

Abstract: Improper vehicle lane-changing behavior is one of the main causes of traffic accidents. Predicting the trajectory of lane-changing vehicles in advance and making corresponding adjustments in time will help to reduce the occurrence of accidents. A trajectory prediction method combining deep learning and ensemble learning is proposed for the lane-changing vehicles, the influence of lane-chang intention on trajectory prediction is considered at the same time. Firstly, establish continuous hidden Markov model（CHMM） to detect the lane-changing intention of the vehicles through their past trajectory data in real time. The driving state of the vehicle is predicted in advance through the past trajectory data of the vehicle, and the trajectory sequence predicted as lane-change is input into the corresponding trajectory prediction model. Secondly, taking LSTM（long short-term memory） as the base predictor of adaptive boosting algorithm, the LSTM-AdaBoost model is established. Based on the trajectory prediction of multiple base predictors at the same time, the weights of each base predictor are adjusted through training and the results are weighted and integrated to improve the accuracy and stability of the prediction model. Finally, the model is trained and tested through the NGSIM（next generation simulation） data set collected in the real driving environment. The results show that the accuracy of the intention prediction model is more than 90% one second before lane change, the accuracy and stability of the LSTM-AdaBoost integrated trajectory prediction model are significantly improved compared with the single LSTM prediction model. And there are fewer abnormal data samples in the prediction results, indicate that the model has good stability. At the same time, the prediction comparison results also show that adding the intention prediction module is helpful to improve the accuracy of lane-changing trajectory prediction.

Key words: vehicle lane-changing trajectory prediction, lane-changing intention identification, hidden Markov model（HMM）, long short-term memory network（LSTM）, AdaBoost ensemble algorithm

孟宪伟, 唐进君, 王喆. 考虑换道意图的LSTM-AdaBoost车辆轨迹预测模型[J]. 计算机工程与应用, 2022, 58(13): 280-287.

MENG Xianwei, TANG Jinjun, WANG Zhe. Trajectory Prediction of Vehicles Based on LSTM-AdaBoost Model Considering Lane-Changing Intention[J]. Computer Engineering and Applications, 2022, 58(13): 280-287.

参考文献

[1] 于扬，梁军，陈龙等.基于高斯混合隐马尔科夫模型与人工神经网络的紧急换道行为预测方法[J].中国机械工程，2020，31（23）：106-114.
YU Y，LIANG J，CHEN L et al.Vehicle emergency lane-changing behavior prediction method based on GMM-HMM and ANN[J].China Mechanical Engineering，2020，31（23）：106-114.
[2] YUAN L，LI S E，LI G.Detection of driver cognitive distraction：An SVM based real-time algorithm and its comparison study in typical driving scenarios[C]//Proceedings of Intelligent Vehicles Symposium，2016.
[3] KEMAR P，PREEOLLAZ M，LEFERE S.Learning-based approach for online lane change intention prediction[C]//Proceedings of IEEE Intelligent Vehicles Symposium，2013：797-802.
[4] ZYNER A，WORRALL S，WARD J.Long short term memory for driver intent prediction[C]//Proceedings of Intelligent Vehicles Symposium，2017.
[5] HEINE J，SYLLA M，LANGER I.Algorithm for driver intention detection with fuzzy logic and edit distance[C]//Proceedings of IEEE International Conference on Intelligent Transportation Systems，2015：1022-1027.
[6] 杨诚.基于隐马尔可夫理论的驾驶人换道意图识别研究[D].长春：吉林大学，2016.
YANG C.Research on drivers lane change intention recognition based on hidden Markov model[D].Changchun：Jilin University，2016.
[7] 张可心.智能汽车换道行为及危险性识别方法研究[D].哈尔滨：哈尔滨工业大学，2017.
ZHANG K X.Recognition method of lane-changing behavior and danger level for intelligent vehicle[D].Harbin：Harbin Institute of Technology，2017.
[8] HOUENOU A，BOMMIFAIT P，CHERFAOUI V.Vehicle trajectory prediction based on motion model and maneuver recognition[C]//Proceedings of IEEE/RSJ International Conference on Intelligent Robots & Systems，2013.
[9] BARTH A，FRANKE U.Where will the oncoming vehicle be the next second?[C]//Proceedings of IEEE Intelligent Vehicles Symposium，2008.
[10] TOLEDO-MOERO R，ZAMORA-IZQUIERDO M A.IMM-based lane-change prediction in highways with low-cost GPS/INS[J].IEEE Transactions on Intelligent Transportation Systems，2009，10（1）：180-185.
[11] 詹盛，徐远新，石涌泉，等.基于模糊着色Petri网的车辆运动轨迹预测[J].计算机工程与应用，2014，50（3）：227-231.
ZHAN S，XU X Y，SHI Y Q，et al.Prediction of vehicle trajectory based on fuzzy colored Petri net[J].Computer Engineering and Applications，2014，50（3）：227-231.
[12] PARK S H，KIM B D，KANG C M.Sequence-to-sequence prediction of vehicle trajectory via LSTM encoder-decoder architecture[C]//Proceedings of Intelligent Vehicles Symposium，2018.
[13] HOU L，XIN L，LI S E，et al.Interactive trajectory prediction of surrounding road users for autonomous driving using structural-LSTM network[J].IEEE Transactions on Intelligent Transportation Systems，2020，21（11）：4615-4625.
[14] DAI S，LI L，LI Z.Modeling vehicle interactions via modified LSTM models for trajectory prediction[J].IEEE Access，2019，7：38287-38296.
[15] ZHANG X，SUN J，QI X.Simultaneous modeling of car-following and lane-changing behaviors using deep learning[J].Transportation Research：Part C Emerging Technologies，2019，104：287-304.
[16] 秦胜君，李婷.多交互车辆轨迹预测研究[J].计算机工程与应用，2021，57（11）：232-238.
QIN S J，LI T.Research on multi-interaction vehicle trajectory prediction[J].Computer Engineering and Applications,2021，57（11）：232-238.
[17] 季学武，费聪，何祥坤，等.基于LSTM网络的驾驶意图识别及车辆轨迹预测[J].中国公路学报，2019，32（6）：34-42.
JI X W，FEI C，HE X K，et al.Intention recognition and trajectory prediction for vehicles using LSTM network[J].China Journal of Highway and Transport，2019，32（6）：34-42.
[18] TOLEDO T，ZOHAR D.Modeling duration of lane changes[J].Transportation Research Record Journal of the Transportation Research Board，1999（1）：71-78.
[19] CHEN Q，HUANG H，LI Y.Modeling accident risks in different lane-changing behavioral patterns[J].Analytic Methods in Accident Research，2021（6）：100159.
[20] 卢锡凤.基于隐马尔可夫模型的城市道路路段不良驾驶行为鉴别[D].哈尔滨：哈尔滨工业大学，2016.
LU X F.Risky driving behavior indentification of urban road sections based on hidden Markov model[D].Harbin：Harbin Institute of Technology，2016.
[21] XIAO C，CHEN N，HU C.Short and mid-term sea surface temperature prediction using time-series satellite data and LSTM-AdaBoost combination approach[J].Remote Sensing of Environment，2019，233：111358.
[22] 李达，张照生，刘鹏，等.基于改进长短时记忆神经网络-自适应增强算法的多天气车辆分类方法[J].汽车工程，2020，42（9）：1248-1255.
LI D，ZHANG Z S，LIU P，et al.Vehicle classification method in multi-climates based on modified LSTM-AdaBoost algorithm[J].Automotive Engineering，2020，42（9）：1248-1255.
[23] BAI Y，XIE J，WANG D，et al.A manufacturing quality prediction model based on AdaBoost-LSTM with rough knowledge[J].Computers & Industrial Engineering，2021，155（5）：107227.
[24] THIEMANN C，TREIBER M，KESTING A.Estimating acceleration and lane-changing dynamics based on NGSIM trajectory data[J].Transportation Research Record Journal of the Transportation Research Board，2008，2088：90-101.
[25] TIJERINA L，GARROTT W，GLECKER M，et al.Van and passenger car driver eye glance behavior during the lane change decision phase[C]//Proceedings of the 84th Annual Meeting of the Transportation Research Board，2005.