Autonomous Vehicle Lateral Control With Curvature and Inclination Compensation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing self-driving vehicle lateral motion control methods require significant computational resources for online calculation of optimal control gains and fail to consider road inclination, leading to potential steady-state errors and reduced precision.
Innovation Solution
A method and apparatus for controlling lateral motion that sets control parameters offline, uses polynomial fitting for real-time gain calculation, and incorporates feedback and feedforward compensation to account for road curvature, inclination, and vehicle uncertainties, ensuring high precision and reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If optimal control parameter gain is obtained by solving the Riccati equation online, then control precision is improved, but calculation amount increases and computing resources are consumed
Solution Approach 1:
The patent pre-calculates and stores optimal control parameter gains for different vehicle speeds before actual operation. During real-time control, the system only needs to query the pre-computed gain corresponding to the current speed, avoiding the need to solve the Riccati equation online. This preliminary action transfers the heavy computational burden from real-time operation to an offline preparation phase, resolving the contradiction between control precision and calculation amount.
2Measurement precision
If control parameters are adjusted adaptively based on path tracking errors, then tracking precision is improved, but controller complexity increases
Solution Approach 1:
The patent changes the control parameter (steering angle) based on the path tracking error magnitude. When the error is large, a larger steering correction is applied; when the error is small, a smaller correction is applied. This parameter change approach allows the controller to adapt to different tracking conditions without requiring a complex adaptive control algorithm, thus improving tracking precision while maintaining relatively simple controller structure.
3Speed
If real-time control gain is acquired through polynomial fitting, then calculation speed is improved, but control accuracy may be reduced compared to exact Riccati solution
Solution Approach 1:
The patent creates a simplified copy of the optimal control gain characteristics by using polynomial fitting to approximate the relationship between vehicle speed and optimal control gain. Instead of computing the exact Riccati solution in real-time, the system uses the polynomial model (copy) to quickly estimate the appropriate gain. This copying approach achieves real-time calculation speed while maintaining sufficient control accuracy for practical applications.
Data Source
Figure 1~2
Figure 3~4
Figure 5~7
AI summary
A method and an apparatus for controlling lateral motion of self-driving vehicle, and a self-driving vehicle. The method for controlling lateral motion of self-driving vehicle comprises: setting control parameters of a linear quadratic controller (S102); acquiring vehicle overall parameters (S104); acquiring a control parameter gain (S106) according to the control parameters and the vehicle overall parameters; acquiring a real-time control parameter gain (S108) according to the control parameter gain; acquiring a state error feedback amount, a trajectory curvature, and a vehicle inclination angle (S110); acquiring a control amount and a compensation amount (S112) according to the real-time control parameter gain, the state error feedback amount, the trajectory curvature, and the vehicle inclination angle; and acquiring an expected steering wheel angle, and outputting the expected steering wheel angle to a steering-by-wire system (S114) according to the control amount and the compensation amount. The present disclosure can achieve the optimal control of the self-driving vehicle lateral motion without static error, greatly reduce the calculation amount, and ensure convergence of an optimal control gain, achieving highly reliable and highly accurate tracking of an expected driving trajectory.