Autonomous Vehicle Control System for Dynamic Capability Assessment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous and semi-autonomous vehicles face challenges in efficiently merging onto freeways, making turns across busy traffic, and navigating through congested roads due to limitations in available information about their own capabilities, leading to conservative and sluggish maneuvers.
Innovation Solution
An autonomous vehicle control system that utilizes high-fidelity models of vehicle subsystems to predict and communicate maximum acceleration, braking, and curvature capabilities to the supervisory controller, enabling more precise and timely path planning and control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conservative powertrain models are used for path planning, then vehicle safety is improved, but merging efficiency and responsiveness deteriorate
Solution Approach 1:
The system dynamically changes planning parameters (acceleration, braking, curvature) based on real-time vehicle capability assessments. Instead of using fixed conservative values, the controller continuously adjusts these parameters according to actual vehicle performance data, allowing optimal balancing of safety and efficiency for each specific maneuver context.
Solution Approach 2:
The system performs preliminary assessment of vehicle capability before executing maneuvers. By evaluating acceleration capabilities, braking performance, and curvature limits in advance, the controller can plan more aggressive and efficient paths while maintaining safety margins, rather than defaulting to conservative approaches.
2Measurement precision
If high-fidelity vehicle models are implemented, then path planning precision is improved, but computational complexity increases
Solution Approach 1:
The vehicle model is segmented into distinct subsystems (powertrain, braking, steering) with dedicated capability assessments for each. This modular approach allows the controller to evaluate specific maneuver requirements against relevant subsystem capabilities without processing the entire vehicle model comprehensively, reducing computational burden while maintaining precision.
Solution Approach 2:
The system performs capability assessments selectively based on maneuver type. For acceleration-intensive maneuvers, detailed powertrain modeling is applied; for braking scenarios, braking system capabilities are emphasized. This partial application of high-fidelity modeling reduces overall computational complexity while maintaining precision where most needed.
3Loss of time
If real-time capability assessment is performed, then maneuver timing is improved, but information processing load increases
Solution Approach 1:
The system performs preliminary capability assessments during periods when maneuvers are not actively being planned, pre-computing vehicle capability parameters and storing them for rapid retrieval. This allows real-time maneuver planning to use pre-prepared capability data, reducing processing load during time-critical decision-making moments.
Solution Approach 2:
The vehicle controller autonomously assesses its own capability without requiring external information processing. By using onboard sensors and pre-stored vehicle parameters, the system generates capability assessments independently, minimizing the information processing burden on external systems while maintaining real-time responsiveness.
Data Source
AI summary
A current state of a vehicle can be identified. At least a minimum acceleration capability of the vehicle is determined. A desired acceleration profile to follow is determined based at least in part on the minimum acceleration capability. An acceleration of the vehicle is controlled based at least in part on the desired acceleration profile.