Multiobjective Optimization of Lane-Changing Strategy for Intelligent Vehicles in Complex Driving Environments

• Published in: IEEE transactions on vehicular technology Vol. 69; no. 2; pp. 1291 - 1308
• Main Authors: Zhou, Jian, Zheng, Hongyu, Wang, Junmin, Wang, Yulei, Zhang, Bing, Shao, Qian
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.02.2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Clustering algorithms; Clusters; Collision avoidance; Collision dynamics; Collisions; Computer simulation; Heuristic algorithms; Intelligent vehicle; Intelligent vehicles; Lane changing; lane-changing strategy optimization; Multiple objective analysis; Optimization; path planning; Performance indices; Polynomials; Strategy; TOPSIS algorithm; Trajectory; Trajectory planning; Tuning; Vehicle dynamics; Vehicles
• ISSN: 0018-9545; 1939-9359
• DOI: 10.1109/TVT.2019.2956504

This paper describes an optimal lane-changing strategy for intelligent vehicles under the constraints of collision avoidance in complex driving environments. The key technique is optimization in a collision-free lane-changing trajectory cluster. To achieve this cluster, a tuning factor is first derived by optimizing a cubic polynomial. Then, a feasible trajectory cluster is generated by adjusting the tuning factor in a stable handling envelope defined from vehicle dynamics limits. Furthermore, considering the motions of surrounding vehicles, a collision avoidance algorithm is employed in the feasible cluster to select the collision-free trajectory cluster. To extract the optimal trajectory from this cluster, the TOPSIS algorithm is utilized to solve a multiobjective optimization problem that is subject to lane change performance indices, i.e., trajectory following, comfort, lateral slip and lane-changing efficiency. In this way, the collision risk is eliminated, and the lane change performance is improved. Simulation results show that the strategy is able to plan suitable lane-changing trajectories while avoiding collisions in complex environments.