A Nonlinear Model Predictive Controller for Vehicle Horizontal and Longitudinal Coupling Based on Parallel Newton Solving

Abstract

The invention discloses a predictive controller based on the parallel Newton solution of the nonlinear model of the vehicle's transverse and longitudinal coupling. The controller obtains the nonlinear control model of the transverse and longitudinal coupling from the vehicle's 3-degree-of-freedom dynamics model, and adopts the steering angle of the front wheels and the driving force of the front and rear wheels. As a control quantity, according to the model predictive control algorithm, the physical constraints of the vehicle are considered and a cost function is constructed. The invention mainly aims at the problem of vehicle path tracking control, uses vehicle dynamics to obtain a control model of horizontal and vertical coupling, designs a nonlinear model predictive controller with this model, and uses parallel Newton method to realize fast solution of the nonlinear controller. The non-linear model predictive controller of vehicle transverse and longitudinal coupling path tracking is derived from the vehicle three-degree-of-freedom dynamics model, and the interaction between the transverse and longitudinal directions is considered. The nonlinear predictive controller is designed based on this model, which retains the vehicle system Non-linearity ensures model accuracy.

Description

technical field [0001] The invention relates to a horizontal and vertical coupling control system in the field of intelligent vehicle automatic driving and a fast calculation method thereof, in particular to a vehicle horizontal and vertical coupling nonlinear model predictive controller based on parallel Newton solution. Background technique [0002] The automatic driving system mainly includes three parts: perception, planning and control. Among them, the vehicle path tracking control system is one of the core links, which has strong research significance and application value. Path tracking control achieves precise tracking of the desired path by controlling the corresponding chassis actuators according to the desired driving path given by the planning system and the state information fed back by the vehicle in real time. Therefore, how to design the controller so that the vehicle can quickly realize the tracking of the desired path, while ensuring real-time performance, ...

AI Technical Summary

Problems solved by technology

[0003] 1. The structure of the vehicle system is complex, especially under complex extreme working conditions, it shows strong nonlinearity, multi-dimensional motion and high coupling, etc. The controller established based on the vehicle linear model ignores many dynamic characteristics, and it is difficult to meet the control requirements of the system. Accuracy requirements

[0004] 2. Although the model predictive controller designed based on the nonlinear model of the vehicle retains the nonlinear and dynamic characteristics of the vehicle, it causes a large computational burden

In the traditional solution method, due to the coupling relationship between the state quantities, the solution in the prediction time domain cannot be independent, and the overall calculation is required. With the increase of the prediction time domain, the solution time is difficult to meet the real-time requirements of the system

Images