Curve Velocity Control Using Lateral Acceleration Confidence
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous driving systems struggle with maintaining effective velocity control when lane markings are not clearly identifiable, leading to reduced confidence in trajectory prediction and control.
Innovation Solution
A vehicle system that determines a trajectory based on predicted and actual lateral acceleration, adjusts a confidence score based on the difference between these accelerations, and uses this score to select an appropriate longitudinal velocity profile for control signals, even in conditions where lane markings are unclear.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If adaptive cruise control and lane following features are used on curvy roads, then velocity control and lane centering are improved, but the system becomes unreliable due to limited sensing ability of the roadway ahead
Solution Approach 1:
The system performs preliminary actions by predicting the trajectory and calculating expected lateral acceleration before the vehicle reaches curved sections. The confidence score is computed in advance based on the agreement between predicted and actual lateral acceleration, allowing the system to proactively adjust velocity profiles before sensing limitations cause control failures.
Solution Approach 2:
The system continuously monitors the agreement between predicted lateral acceleration (from trajectory) and actual lateral acceleration (from steering data) to compute a confidence score. This feedback mechanism allows the system to detect when sensing ability is insufficient and adjust velocity control accordingly, maintaining reliability despite limited roadway ahead sensing.
2Reliability
If the system disables autonomous driving features on curvy roads due to limited sensing, then safety is maintained, but velocity control and lane following capabilities are lost
Solution Approach 1:
The system dynamically adjusts velocity control based on real-time confidence scores computed from lateral acceleration agreement. Rather than statically disabling features on curvy roads, the system continuously adapts velocity profiles to match confidence levels, maintaining both safety and velocity control capability across varying road conditions.
Solution Approach 2:
The system changes the parameter of velocity profile selection based on the confidence score. When confidence is high (good agreement between predicted and actual lateral acceleration), more aggressive velocity profiles are used. When confidence is low (poor agreement or lost lane markings), conservative velocity profiles are selected, maintaining safety while preserving adaptability.
3Measurement precision
If the system uses predicted lateral acceleration to determine confidence score, then velocity control accuracy is improved, but the system complexity increases
Solution Approach 1:
The confidence score acts as an intermediary metric that simplifies the relationship between complex trajectory predictions and velocity control decisions. Instead of directly using complex predicted lateral acceleration calculations to control velocity, the system first computes a simplified confidence score from the agreement between predicted and actual lateral acceleration, then uses this score to select appropriate velocity profiles.
Data Source
AI summary
Methods and systems for providing driving assistance in a vehicle. In one embodiment, a method includes: determining, by a processor, a trajectory of the vehicle along a roadway; determining, by the processor, a confidence score based on a predicted lateral acceleration and an actual lateral acceleration at a point along the trajectory; determining, by the processor, longitudinal velocity data based on the confidence score; and generating, by the processor, control signals to vehicle actuators to control the vehicle based on the longitudinal velocity data.