Self-adaptive MPC parking transverse control method based on positions and postures

Abstract

The invention, which belongs to the technical field of intelligent vehicle transverse control, discloses a self-adaptive MPC parking transverse control method based on positions and postures. The method comprises the steps: step one, setting a range of a parking vehicle speed and selecting a predicted step length and sampling time according to the speed; step two, on the basis of a nonlinear vehicle kinematics model, designing a model prediction control algorithm and calculating a transverse distance deviation and a yaw angle deviation; step three, designing a performance index function according to the transverse distance deviation and the yaw angle deviation obtained in the step two; and step four, designing a fuzzy controller for adaptively adjusting a weight coefficient. According to the invention, the fuzzy controller adjusts the weight coefficient of the performance index function in real time based on the vehicle speed and the transverse distance deviation. Therefore, the vehicle posture can be effectively controlled; a problem that the vehicle body is not parked properly after the vehicle enters the parking space is solved; and the high-precision control can be realized byusing a few of prediction step lengths, so that the calculation burden is reduced.

Description

technical field [0001] The invention relates to the technical field of intelligent vehicle lateral control, in particular to an adaptive MPC parking lateral control method based on position and attitude. Background technique [0002] In recent years, the number of cars has continued to increase, the ability of transportation technology has continued to improve, and the demand for assisted driving technology for cars has also increased. Based on actual needs, research and exploration on technologies related to automatic parking control have been carried out. Due to the limitation of parking control and the limitation of the body structure, the vehicle is required to park according to the specification of the parking space. Based on these limitations, avoiding the collision between the main vehicle and other vehicles and realizing the safe storage of vehicles is the difficulty in the development of automatic parking technology. The path planning method for parking generally u...

AI Technical Summary

Problems solved by technology

[0003] In the automatic parking scene, although some common vehicle lateral control methods can achieve high error accuracy, the tracking effect on the expected attitude of the vehicle is not good, and collisions are prone to occur during the parking process, resulting in parking failure or the need to restart the vehicle. Plan the path for parking, especially in parallel parking scenarios. The path planning uses curves. If the posture of the vehicle body is not adjusted in time, it will cause the vehicle to be placed incorrectly after entering the parking space, making it difficult to adjust again.

Images