Field-Based Torque Steering Control for Autonomous Vehicles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles using traditional steering methods struggle to replicate human driving behavior, which allows for deviation from a centerline while maintaining control, and often require more energy to steer than human-driven vehicles.
Innovation Solution
A field-based torque steering system that uses potential fields to guide the vehicle along a nominal path, allowing for deviation within a defined corridor by applying torque to the steering column based on the vehicle's pose relative to the field, incorporating attractive and repulsive potentials to maintain the vehicle within the driving corridor.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional steering control methods are used to track the centerline of a path, then the vehicle maintains precise path following, but the control effort and energy consumption increase compared to human driving behavior
Solution Approach 1:
The patent transforms the steering control problem from direct steering angle control to torque control by changing the control parameter. The potential field gradient provides a torque command that naturally replicates human driving behavior, allowing the vehicle to follow the path centerline with appropriate deviation while consuming less energy. This parameter transformation resolves the contradiction by enabling energy-efficient control that maintains path following precision.
Solution Approach 2:
The patent replaces traditional mechanical steering control systems with a field-based torque control system. Instead of directly controlling steering wheel angle through mechanical means, the system uses a virtual potential field to generate torque commands that act on the steering system. This substitution enables more natural and energy-efficient steering behavior while maintaining path tracking accuracy.
2Measurement precision
If autonomous vehicles steer to precisely track the centerline of a path, then path following accuracy is improved, but the driving behavior becomes unnatural compared to human drivers who allow more deviation
Solution Approach 1:
The potential field-based torque steering system enables the autonomous vehicle to steer itself in a manner that naturally replicates human driving behavior. The attractive potential centered on the path guide automatically generates appropriate steering torques that allow for natural deviation from the centerline while maintaining overall path following. This self-service approach eliminates the need for complex control algorithms to simulate human behavior.
Solution Approach 2:
By changing from direct steering angle control to torque control based on potential field gradients, the system naturally produces human-like steering behavior. The torque commands derived from the potential field allow the vehicle to deviate from the centerline in a controlled and natural manner, improving the ease of operation and naturalness of driving behavior while maintaining adequate path following accuracy.
3Use of energy by moving object
If field-based torque steering control is implemented to allow deviation from the centerline, then energy consumption is reduced, but path following precision may be compromised
Solution Approach 1:
The potential field-based torque steering system continuously calculates the gradient of the attractive potential field based on the vehicle's position relative to the path guide. This feedback mechanism ensures that the torque commands naturally adjust to maintain the vehicle within the driving corridor while following the path centerline with appropriate deviation. The feedback control resolves the contradiction by maintaining path tracking precision while enabling energy-efficient operation.
Data Source
AI summary
A system includes a computer programmed to determine, along a nominal path to be traversed by a vehicle, a potential field representing a driving corridor for the vehicle. The computer is further programmed to identify a position of the vehicle relative to the potential field at a current time, and apply a torque to q steering column of the vehicle. The torque is based at least in part on the position. The potential field includes an attractive potential that guides the vehicle to remain within the corridor.