Automotive Forward Collision Threat Evaluation Using Longitudinal Distance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional automotive forward collision threat evaluation systems have limited accuracy due to inaccurate estimation of lateral velocity and lateral acceleration of oncoming vehicles, leading to false alarms and unnecessary interventions, as they rely on predicted paths that may change before actual collisions occur.
Innovation Solution
The method evaluates the threat by estimating a free passing distance for the oncoming vehicle based on the host vehicle's position and the oncoming vehicle's longitudinal velocity, determining if this distance is sufficient for safe passing, without relying on the predicted path of the oncoming vehicle, and signaling collision avoidance only if the distance is insufficient.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional sensors (RADAR, LIDAR, cameras) are used to estimate lateral velocity and lateral acceleration of oncoming vehicles, then the system can detect the presence and basic motion parameters of oncoming vehicles, but the estimation accuracy of lateral velocity and lateral acceleration is limited
Solution Approach 1:
The patent extracts and eliminates the problematic component of predicting the oncoming vehicle's future path, which relies on inaccurate lateral velocity and acceleration estimates. Instead, the system focuses only on longitudinal parameters (longitudinal velocity, longitudinal distance) that can be measured with high accuracy by conventional sensors, and combines these with the host vehicle's own motion parameters to determine collision threat.
Solution Approach 2:
The patent changes the set of parameters used for collision evaluation from including lateral velocity and lateral acceleration to using only longitudinal velocity, longitudinal distance, and host vehicle motion parameters. This parameter substitution eliminates the accuracy problem while maintaining the ability to assess collision threat.
2Loss of time
If the system predicts the future path of the oncoming vehicle several seconds before collision to enable timely intervention, then collision avoidance actions can be initiated early, but the uncertainty of future maneuvers increases and prediction accuracy decreases
Solution Approach 1:
The patent removes the oncoming vehicle's future path prediction from the collision evaluation process entirely. By eliminating this unreliable predictive element, the system avoids the trade-off between early intervention and prediction reliability, achieving both timely detection and high reliability through direct comparison of current longitudinal parameters.
3Reliability
If the system compares predicted future paths of oncoming and host vehicles to evaluate collision threat, then it can identify potential collisions, but false alarms and unnecessary interventions occur due to limited prediction accuracy
Solution Approach 1:
The patent extracts and removes the source of false alarms by eliminating the comparison of predicted future paths. Instead, it uses a direct evaluation method based on current longitudinal parameters and host vehicle motion, which provides reliable collision threat assessment without generating false alarms.
Solution Approach 2:
The patent changes the evaluation parameters from predicted path coordinates to fundamental motion parameters (longitudinal velocity, longitudinal distance, host vehicle acceleration). This parameter transformation provides a more reliable and direct assessment of collision threat, eliminating false alarms caused by prediction errors.
Data Source
AI summary
A method and system operable for evaluation of an automotive vehicle forward collision threat. The presence of an oncoming vehicle in front of a host vehicle is established and a passing position of the host vehicle for the moment when the vehicles reach each other is predicted. In addition, a free passing distance for free passing of the oncoming vehicle on a first side of the host vehicle is predicted for the situation when the host vehicle is positioned in the passing position. The free passing distance is the minimum distance between the first side of the host vehicle and a road edge. A width of the oncoming vehicle is also estimated. If the free passing distance is equal to or less than a comparison value, it is decided that a forward vehicle collision threat exists. The comparison value is at least equal to the width of the oncoming vehicle.