Nonlinear Vehicle Steering Control for Curved Roadways
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous vehicle steering systems rely on complex linear control theories, requiring numerous vehicle parameters to be tuned, which is time-consuming and inefficient.
Innovation Solution
A nonlinear control function is used to generate steering control commands for autonomous vehicle steering, utilizing fewer vehicle parameters and incorporating sensed data such as lane markers, curvature, and vehicle velocity to calculate feed-forward and feedback control outputs, along with a disturbance compensator, to autonomously steer the vehicle.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If linear control theories (PID control loops, Jacobian, Taylor series expansions) are used for autonomous vehicle steering, then steering control can be achieved, but the system complexity and number of parameters to be tuned increase significantly
Solution Approach 1:
The patent transitions from linear control theories requiring multiple vehicle parameters to a nonlinear control function that uses fewer parameters. This parameter reduction directly addresses the contradiction by maintaining steering control capability while significantly reducing system complexity and tuning requirements
Solution Approach 2:
The patent extracts and removes unnecessary complex linear control components (PID loops, Jacobian calculations, Taylor series expansions) from the steering system, retaining only the essential nonlinear control function that achieves the same steering objective with reduced complexity
2Reliability
If linear control theories with numerous vehicle parameters are used, then steering control can be achieved, but the time required to tune parameters increases
Solution Approach 1:
By changing from linear to nonlinear control methodology, the patent reduces the number of parameters that require tuning, directly reducing the time loss associated with parameter adjustment while maintaining steering control reliability
3Extent of automation
If existing linear control functions with many vehicle parameters are used, then autonomous steering can be implemented, but computational efficiency decreases
Solution Approach 1:
The patent removes computationally intensive linear control calculations (PID loops, Jacobian matrices, Taylor series) from the autonomous steering system, replacing them with a more efficient nonlinear control function that maintains full autonomous steering capability while improving computational efficiency
Data Source
AI summary
A method and system for controlling a vehicle along a curved roadway, including receiving vehicle sensed data relating to vehicle operation and/or vehicle position and orientation; identifying first and second lane markers from the sensed data; determining a reference path for the vehicle using the lane markers; calculating a feed-forward control output based upon the curvature of the lane markers; calculating a lateral position of the vehicle relative to the reference path, and a lateral position error based upon the lateral position; and calculating a vehicle heading angle based upon the vehicle sensed data, and a heading angle error. A feedback control output is calculated using a nonlinear control function based upon the lateral position error, the heading angle error, and the sensed velocity. A steering control output is generated for steering the vehicle, based upon the feed-forward control output and the feedback control output.