Predictive control optimization method and device for autonomous vehicle

Abstract

The invention discloses a predictive control optimization method and device for an autonomous vehicle, and relates to the technical field of computers. One embodiment of the method comprises the steps of obtaining an optimization function based on a vehicle dynamics model, wherein decision variables of the optimization function comprise a state variable and a control increment; obtaining vehicle real-time driving parameters and a reference trajectory, and obtaining decision variables of the optimization function according to the vehicle real-time driving parameters and the vehicle reference trajectory; determining constraint conditions to constrain the optimization function according to vehicle operation indexes, wherein the constraint conditions comprise a front wheel rotation angle increment constraint, a front wheel rotation angle constraint, an acceleration increment constraint, an acceleration constraint, a transverse acceleration constraint and a tire slip angle constraint; and constructing an optimization model according to the optimization function and the constraint condition, and optimizing the decision variable by using the optimization model to enable the result of the optimization function to meet a preset requirement so as to perform predictive control on the vehicle. According to the invention, the track following precision and stability of the autonomous vehicle are improved, and the control effect is improved.

Description

technical field [0001] The present invention relates to the field of computer technology, in particular to a method and device for predictive control optimization of automatic driving vehicles. Background technique [0002] The trajectory tracking controller of autonomous vehicles is divided into upper and lower controllers in the current mainstream applications. The main function of the upper controller is to process the reference trajectory and generate control instructions, and the main function of the lower controller is to execute the control instructions output by the upper controller. Drive the vehicle. At present, when the upper controller controls the vehicle, it mostly adopts the model predictive control MPC (Model Predictive Control) method, which can control the vehicle laterally and longitudinally at the same time. At present, the commonly used model predictive control methods are mostly control methods based on kinematics models. [0003] In the course of rea...

AI Technical Summary

Problems solved by technology

[0004] The control method based on the kinematic model does not consider factors such as tire dynamics and the time-varying nature of the vehicle model with speed, resulting in low control accuracy and high rollover risk when the vehicle is driving at high speed when detours and corners

Images