Automated Vehicle Path Planning and Trajectory Tracking Control Based on Unscented Kalman Filter Vehicle State Observer

For automated driving vehicles, path planning and trajectory tracking are the core of achieving obstacle avoidance. Real-time external environment perception and vehicle state monitoring play the important role in the decision-making of vehicle operation. Sensor measuring is an important way to obtain vehicle state parameters, but some parameters cannot be measured due to sensor cost or technical reasons, such as vehicle lateral velocity and side-slip angle. This disadvantage will adversely affect the monitoring of vehicle self-condition and the control of vehicle running, even it will lead to erroneous decision-making of vehicles. Therefore, this paper proposes an automated driving path planning and trajectory tracking control method based on Kalman filter vehicle state observer. Some of vehicle state data can be measured accurately by sensors. The vehicle state observer combined with Adaptive Neural Fuzzy Interference System (ANFIS) and Unscented Kalman filter (UKF) is used to estimate the lateral velocity of the vehicle in real time. The vehicle body state and tire constraints are considered. The model predictive control (MPC) method is used to predict the vehicle and obstacle trajectory and control strategy. Thus, reasonable local path planning is obtained. Finally, some complex scenarios are build including road and dynamic obstacles in scenario builder. Then the path planning strategy is established with Simulink to verify the feasibility of the local path planning theory proposed in this paper.