Transverse control adjusting method cooperating with longitudinal controller

Abstract

The invention relates to the technical field of intelligent automobiles, in particular to a transverse control adjusting method cooperating with a longitudinal controller. According to the technical scheme, the method comprises the steps of starting an automatic parking function, obtaining the longitudinal speed, the acceleration, the yaw velocity, the transverse error, the course error and curvature information of a planned path of a vehicle, constructing a transverse control state vector, building a state-space equation based on a vehicle model, constructing a transverse controller through an LQR algorithm, and calculating an optimal feedback gain matrix. And outputting an initial steering wheel angle instruction, introducing a cooperative input adjustment mechanism, dynamically adjusting a state weight matrix of the LQR through a nonlinear adjustment function according to the longitudinal speed of the vehicle, realizing adaptive optimization of control gain, and dynamically limiting an angular speed output upper limit of the steering wheel according to the longitudinal acceleration and path curvature of the vehicle. And performing speed limiting processing on the steering wheel instruction, outputting a smooth control instruction, and detecting a large curvature or steering saturation risk.

Images