Probabilistic Lane-Change Decision-Making and Planning for Autonomous Heavy Vehicles Wen Hu, Zejian Deng, Dongpu Cao, Member, IEEE, Bangji Zhang, Amir Khajepour, Member, IEEE, Lei Zeng, and Yang Wu Abstract—To improve the safety and driving stability of the autonomous heavy truck, it is necessary to consider the differ- ences of driving behavior and drivable trajectories between the heavy trucks and passenger cars. This study proposes a proba- bilistic decision-making and trajectory planning framework for the autonomous heavy trucks. Firstly, the driving decision pro- cess is divided into intention generation and feasibility evalua- tions, which are realized using the utility theory and risk assess- ment, respectively. Subsequently the driving decision is made and sent to the trajectory planning module. In order to reflect the greater risks of the truck to other surrounding vehicles, the aggressiveness index (AI) is proposed and quantified to infer the asymmetrical risk level of lane-change maneuver. In the planning stage, the lateral and roll dynamics stability domains are devel- oped as the constraints to exclude the candidate trajectories that would cause vehicle instability. Finally, the simulation results are compared between the proposed model and the artificial poten- tial filed model in the scenarios extracted from the naturalistic driving data. It is shown that the proposed framework can pro- vide the human-like lane-change decisions and truck-friendly tra- jectories, and performs well in dynamic driving environments. Index Terms—Autonomous heavy truck, decision-making, driving aggressiveness, risk assessment, trajectory planning.    I. Introduction A LMOST one third of accidents are caused by trucks while they only account for 10% of the road vehicles, and the fatality of accidents caused by trucks is also much higher than that in the passenger cars related crashes [1]. The applications of autonomous trucks could help to reduce the accidents caused by the fatigue and distracted of driver which are com- mon in the long-distance transport. However, their typical fea- tures are not stressed well in the decision-making and plan- ning process [2], [3]. Therefore, this study aims at proposing the lane-change decision making and trajectory planning model for the heavy trucks by considering their unique char- acteristics. To address the lane-change (LC) of autonomous heavy trucks, the differences between the heavy trucks and the pas- senger cars should be introduced. Firstly, the motivation and frequency of lane-change are different. The passenger car drivers are likely to change lane to obtain higher driving speed while the truck drivers may not in the same scenario with more weightings on the driving safety [4]. Secondly, the lat- eral and rollover dynamics stabilities of trucks are inferior to them of the passenger cars because of the higher center of mass and worse braking capability [1], [5]. Thus, the lateral speed and lane change spans differ in large range, leading to various distributions of lane change trajectories [6]. Moreover, the trucks have greater impact on other traffic participants due to larger size and weight. Their wider perception blind areas also increase the driving risks [7]. Lane-change behavior can be modeled based on integrated or hierarchical frameworks [8]. The decision-making process is weakened with lane keeping and lane changing trajectories not distinguished in the integrated framework [9], [10]. On the contrary, the hierarchical architecture focuses more on under- standing the lane-change behavior of real human drivers [8], which can be divided into lane-change decision, trajectory generation and optimization [11], [12]. Toledo modeled the lane-change behavior decision using utility theory and gap acceptance model [13]. However, the acceptable gaps vary greatly for different types of vehicles and drivers. In the exist- ing publications about lane-change decision [14], [15], only a few involved with the driving behavior of heavy trucks at the individual vehicle level. Chen et al. [4] and Aghabayk et al. [6] revealed that the car-following behavior and lane-change decision of trucks are different with cars by analyzing the real driving data. Moridpour et al. developed a lane-changing deci- sion model of trucks for the microscopic traffic simulation based on fuzzy logic [3]. Nevertheless, the characteristics of trucks are not highlighted with mathematic model. Moreover, the above microscopic traffic models ignore the feasibility of lane-change trajectory. Trajectory generation and optimization is part of the motion planning research [16], where plenty of methods can be applied [17], such as interpolating curve, sampling-based approach, graph search, artificial potential field [9], etc. Poly- Manuscript received March 22, 2022; accepted June 30, 2022. This work was supported by the National Natural Science Foundation of China (51870 51675). Recommended by Associate Editor Zhen Song. (Corresponding author: Zejian Deng and Bangji Zhang.) Citation: W. Hu, Z. J. Deng, D. P. Cao, B. J. Zhang, A. Khajepour, L. Zeng, and Y. Wu, “Probabilistic lane-change decision-making and planning for autonomous heavy vehicles,” IEEE/CAA J. Autom. Sinica, vol. 9, no. 12, pp. 2161–2173, Dec. 2022. W. Hu, B. J. Zhang, and L. Zeng are with the State Key Laboratory of Advanced Design and Manufacturing for Vehicle Body, College of Mechanical and Vehicle Engineering, Hunan University, Changsha 410082, China (e-mail: huxiaowen@hnu.edu.cn; bangjizhang@hnu.edu.cn; s2002002 34@hnu.edu.cn). Z. J. Deng and A. Khajepour are with the Department of Mechanical and Mechatronics Engineering, University of Waterloo, Waterloo ON N2L3G1, Canada (e-mail: z49deng@uwaterloo.ca; akhajepour@uwaterloo.ca). D. P. Cao and Y. Wu are with the School of Vehicle and Mobility, Tsinghua University, Beijing 100084, China (e-mail: DP_Cao2016@163.com; yangwu@mail.tsinghua.edu.cn). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/JAS.2022.106049 IEEE/CAA JOURNAL OF AUTOMATICA SINICA, VOL. 9, NO. 12, DECEMBER 2022 2161