Jian Ou

In order to solve the problem of emergency collision avoidance of autonomous vehicles when the front vehicle is unstable and sliding under high-speed conditions, a research method for the state recognition of the front side-skid vehicle and the trajectory prediction of the front side-skid vehicle was proposed. By extracting the key features of the vehicle in front of the vehicle in danger of sliding to build a skidding recognition model of the vehicle in front, a skidding recognition strategy of the vehicle in front was designed based on the extracted skidding feature indexes to judge the skidding state of the vehicle in front. The state quantity of the sliding vehicle in front is selected, and the constant rotation rate and acceleration model (CTRA) is established to predict the trajectory of the sliding vehicle in front in a short time. Considering the simplified assumptions of the model and the noise in the process of sensor perception information, the Unscented Kalman Filter (UKF) is used to deal with the uncertainty in the trajectory prediction process, the possible position and covariance of the front sideslipping vehicle are calculated, and the possible future area of the front sideslipping vehicle is estimated under the condition of a probability of 0.9. Through the established Carsim and Simulink co-simulation platform, the effectiveness of the front vehicle skidding state recognition strategy and the accuracy of the trajectory prediction of the sliding vehicle are verified under the condition of high speed and low attachment.

In emergency situations, it is difficult to meet the requirements of safe driving only by relying on the braking system, and the probability of accidents can be reduced by employing an emergency lane-changing mode. To improve the adaptability of the distributed electric vehicle adaptive cruise control (ACC) strategy to complicated and volatile conditions, a multimode ACC strategy with emergency lane-changing function is proposed. Firstly, the ACC is divided into four modes aimed at the problem of complex conditions, and a switching strategy is designed to control the switching of them. Simultaneously, the car-following mode is divided in greater detail based on time to collision (TTC), and the acceleration weighted average algorithm is adopted for accuracy and output continuity during switching. Then, the ACC is established with a hierarchical control framework, in which a PID-based cruise mode and a multi-objective optimized car-following mode based on model predictive control (MPC) are devised. The target brake wheel cylinder pressure is selected as the emergency brake pressure in takeover mode. In addition to the MPC-based system, the emergency lane-changing mode incorporates a yaw moment controller in the upper-level controller to improve body stability during emergency lane changing in the upper-level controller. In the lower-level controller, the upper-level output is converted into driving torque, wheel cylinder pressure, and front wheel angle to control vehicle travel and generate additional yaw moment. Finally, the results indicate that the presented multimode switching strategy can adapt to complex and instable transportation environments. In the cruise control scenario, the host vehicle can rapidly reach cruising speed within 5 s. In the car-following scenario, the host vehicle can stably follow the preceding vehicle with an acceleration of −5–3.5 m/s² and a jerk of −2–2 m/s³ throughout the entire process, maintaining a safe distance from the preceding vehicle. In emergency lane-changing scenarios, vehicles with body stability control can better follow the lane-changing trajectory, and tracking accuracy is improved by 65%. Simultaneously, parameters such as front wheel angle, yaw rate, sideslip rate, and lateral acceleration remain within the normal range. In mixed switching scenarios, each mode can be correctly switched according to diverse operating conditions, and obstacle avoidance can be accomplished through horizontal and vertical strategies, which also verify the effectiveness and rationality of the control strategy proposed.

The vehicle particle model was built to compare and analyze the effectiveness of three different collision avoidance methods. The results show that during vehicle high-speed emergency collision avoidance, lane change collision avoidance requires a smaller longitudinal distance than braking collision avoidance and is closer to that with a combination of lane change and braking collision avoidance. Based on the above, a double-layer control strategy is proposed to avoid collision when vehicles change lanes at high speed. The quintic polynomial is chosen as the reference path after comparing and analyzing three polynomial reference trajectories. The multiobjective optimized model predictive control is used to track the lateral displacement, and the optimization objective is to minimize the lateral position deviation, yaw rate tracking deviation, and control increment. The lower longitudinal speed tracking control strategy is to control the vehicle drive system and brake system to track the expected speed. Finally, the lane changing conditions and other speed conditions of the vehicle at 120 km/h are verified. The results show that the control strategy can track the longitudinal and lateral trajectories well and achieve effective lane change and collision avoidance.

Aiming at the problem of body instability caused by actuator failure in a distributed electric vehicle drive system, a fault-tolerant control strategy of longitudinal and lateral force cooperative reconstruction with active steering control was proposed, and a layered control structure was adopted based on the vehicle model. In the upper controller, the resultant force and torque are calculated according to the vehicle parameter state and MPC algorithm; the lower controller is the cooperative reconfiguration allocation layer, and the minimum tire load rate, longitudinal and lateral force constraints and front wheel angle control are considered. Finally, offline simulation experiments and hardware-in-the-loop experiments are completed to verify the effectiveness and real-time performance of the designed strategy. The results show that the designed strategy can significantly improve the driving stability and safety of the vehicle when the actuator fails.

Autonomous urban driving navigation is still an open problem and has ample room for improvement in unknown complex environments. This paper proposes an end-to-end autonomous driving approach that combines Conditional Imitation Learning (CIL), Mask R-CNN with DDPG. In the first stage, data acquisition is first performed by using CARLA, a high-fidelity simulation software. Data collected by CARLA is used to train the Mask R-CNN network, which is used for object detection and segmentation. The segmented images are transformed into the backbone of CIL to perform supervised Imitation Learning (IL). DDPG means using Reinforcement Learning for further training in the second stage, which shares the learned weights from the pre-trained CIL model. The combination of the two methods is an innovative way of considering. The benefit is that it is possible to speed up training considerably and obtain super-high levels of performance beyond humans. We conduct experiments on the CARLA driving benchmark of urban driving. In the final experiments, our algorithm outperforms the original MP by 30%, CIL by 33%, and CIRL by 10% in the most difficult tasks, dynamic navigation tasks, and in new environments and new weather, demonstrating that the two-stage framework proposed in this paper shows remarkable generalization capability in unknown environments on navigation tasks.

Autonomous driving technology in urban environments is a very important avenue of research. Notably, the question of how to plan safe lane-changing trajectories is a challenge in multi-vehicle traffic environments. In our research, three kinds of polynomial lane changing mathematical models were analyzed and compared. It was found that the fifth polynomial is the most suitable for lane changing trajectories; it is defined as a generalized lane-changing trajectory cluster, whereby the minimum lane change time is determined by the vehicle lateral stability threshold. Here, a collision avoidance algorithm is proposed to eliminate unsafe trajectories. Finally, the TOPSIS algorithm is used to solve the multi-objective optimization problem, and the optimal lane-changing expected trajectory is obtained from the safe trajectory cluster. The simulation results showed improvements in lane-changing efficiency of 6.67% and no collisions in the overtaking condition. In general, the proposed method of identifying the optimal lane changing trajectory can achieve safe, efficient and stable lane changing.