Autonomous Steering Control Stabilization via Kalman Filter Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern automobiles equipped with lane-centering features may unexpectedly move or jump laterally due to changes in roadway conditions, particularly when lane markers become unidentifiable, leading to instability in autonomous driving modes.
Innovation Solution
A steering control system that utilizes a combination of polynomial path data and estimation path data, where the estimation path data is generated using a bicycle model and a Kalman filter, to maintain vehicle positioning within lane markers, even when the vehicle-lane confidence level drops below a threshold, thereby preventing sudden lateral movements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the system uses polynomial path data for lane-centering control, then the vehicle can maintain positioning within lane markers under normal conditions, but the vehicle may unexpectedly jump laterally when lane markers become unidentifiable or confidence level drops
Solution Approach 1:
The patent introduces an estimation filter as an intermediary component that generates estimation path data when polynomial path data becomes unreliable. The estimation filter acts as a mediator between the unreliable polynomial path data and the steering control system, providing smoothed lateral position estimates that prevent sudden lateral jumps while maintaining lane-centering functionality.
Solution Approach 2:
The system dynamically changes the parameter being used for control based on confidence level. When polynomial path data confidence drops below a threshold, the system switches from using polynomial-derived lateral positions to using estimation filter-derived lateral positions. This parameter switching resolves the contradiction by adapting to changing roadway conditions.
2Reliability
If the system switches to estimation path data when confidence is low, then lateral stability is improved, but the system complexity increases due to additional filtering components
Solution Approach 1:
The system implements dynamic switching between two control data sources (polynomial path data and estimation path data) based on real-time confidence level assessment. This dynamic approach allows the system to maintain simplicity under normal conditions while automatically activating the estimation filter only when needed, thus improving reliability without permanently increasing system complexity.
Solution Approach 2:
The patent segments the control system into distinct functional modules: polynomial path generation, estimation filter, confidence level monitoring, and switching logic. This segmentation allows each component to be independently optimized and managed, reducing overall system complexity while enabling the sophisticated dual-path control strategy.
Data Source
AI summary
A system and a method are described. The method includes: receiving sensed input from a vehicle sensor suite; using the input, providing a first output; determining that a vehicle-lane confidence level is less than a threshold; and then instead, providing a second output, wherein the first and second outputs comprise lane-correction data, wherein the second output is determined using an estimation filter.