A co-evolutionary lane-changing trajectory planning method for automated vehicles based on the instantaneous risk...

Lane-changing trajectory planning (LTP) is an effective concept to control automated vehicles (AVs) in mixed traffic, which can reduce traffic conflicts and improve overall traffic efficiency. To enhance the lane change safety for AVs, a co-evolutionary lane-changing trajectory planning (CLTP) method is proposed to describe the risk minimization process that co-evolves with the dynamic traffic environment in the limited literature. Firstly, the natural driving data of vehicle trajectory on the expressway provided by the High dataset are used to construct the lane-changing samples. To obtain the future traffic environment information, a deep learning neural network is adopted to capture trajectory dynamics in mobility of surrounding vehicles around a lane-changing vehicle. Secondly, the safe interaction between the subject vehicle and the surrounding vehicles is considered to establish a mathematical model for the temporal and spatial risk identification of a lane change event based on the fault tree analysis method. Subsequently, the risk minimization of lane change is considered as the objective. Based on the acceleration and deceleration overtaking rules and the trapezoidal acceleration method, the longitudinal and lateral displacement schemes during a lane change are designed. Finally, the motion parameters of longitudinal and lateral displacement are acquired to form an ideal lane change trajectory using a genetic algorithm. The results show that this method can effectively achieve higher safety of the lane-changing process, and reduce the traffic conflicts and traffic turbulence caused by dangerous lane-changing behaviors. The findings can provide theoretical support for lane change trajectory planning algorithm design of intelligent vehicles.