Curve-Adaptive Collision Avoidance Control System
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing collision avoidance systems fail to consider the traveling conditions of vehicles, such as straight or curved paths, leading to unnecessary avoidance controls when predicting collision possibilities between oncoming and own vehicles.
Innovation Solution
A collision avoidance control system that detects whether the own vehicle and oncoming vehicle are traveling on curves and adjusts the transfer prediction times accordingly, setting them to shorter times when either vehicle is on a curve, to prevent unnecessary collision determinations and controls.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the transfer prediction time is set to a long time to ensure comprehensive collision detection, then the collision detection coverage is improved, but unnecessary avoidance control is triggered frequently when vehicles are traveling on curves
Solution Approach 1:
The patent applies dynamics by making the transfer prediction time adjustable based on traveling conditions. The system dynamically changes the prediction time parameter: using a first (longer) transfer prediction time when vehicles are traveling straight, and a second (shorter) transfer prediction time when vehicles are traveling on curves. This resolves the contradiction by adapting the detection coverage to the actual driving scenario, ensuring comprehensive detection when needed while preventing false alarms during curve traversal.
Solution Approach 2:
The patent implements parameter changes by modifying the transfer prediction time parameter according to the detected traveling condition. When the ECU determines that either the own vehicle or oncoming vehicle is traveling on a curve, it changes the prediction time parameter to a shorter value. This parameter adjustment directly addresses the contradiction by reducing the detection horizon in curve scenarios where long prediction times cause unnecessary avoidance control, while maintaining long prediction times for straight travel where comprehensive detection is appropriate.
2Object-generated harmful factors
If the transfer prediction time is set to a short time to avoid unnecessary avoidance control on curves, then false alarms are reduced, but collision detection accuracy may be compromised
Solution Approach 1:
The system uses dynamics to adapt the prediction time to the specific traveling condition. Rather than using a fixed short time that would compromise detection accuracy in all scenarios, the system dynamically selects between a first transfer prediction time (for straight travel) and a second transfer prediction time (for curve travel). This ensures that detection accuracy is maintained when appropriate (straight travel) while false alarms are reduced when necessary (curve travel).
Solution Approach 2:
The patent applies parameter changes by adjusting the transfer prediction time based on the traveling condition detected by the ECU. When curve travel is detected, the parameter is changed to a shorter value to reduce false alarms. When straight travel is detected, the parameter is changed to a longer value to maintain high detection accuracy. This conditional parameter adjustment resolves the contradiction between reducing false alarms and maintaining detection accuracy.
3Reliability
If the vehicle course calculation is extended to cover long prediction times, then comprehensive collision prediction is achieved, but the computational load and false collision determination increase on curves
Solution Approach 1:
The patent applies dynamics by making the course calculation duration adaptive to the traveling condition. The ECU dynamically adjusts the prediction horizon: calculating the own vehicle course and oncoming vehicle course over a first (longer) transfer prediction time when traveling straight, and over a second (shorter) transfer prediction time when traveling on curves. This reduces the computational load and false collision determinations during curve traversal while maintaining comprehensive collision prediction during straight travel.
Solution Approach 2:
The patent implements parameter changes by modifying the transfer prediction time parameter that determines the duration of vehicle course calculation. When the ECU detects curve travel, it changes this parameter to a shorter value, thereby reducing the computational burden of calculating extended vehicle courses and the associated false collision determinations. When straight travel is detected, the parameter is changed to a longer value to ensure comprehensive collision prediction.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A collision avoidance control system (1) includes an ECU (2), and the ECU (2) calculates an own vehicle course (Cm1) as a course along which an own vehicle (M) is expected to proceed during a first transfer prediction time, and time-dependent own vehicle positions (M1 - M4) on the own vehicle course (Cm1), and calculates an oncoming vehicle course (Cn1) as a course along which an oncoming vehicle (N) is expected to proceed during a second transfer prediction time, and time-depending oncoming vehicle positions (N1 - N4) on the oncoming vehicle course (Cn1). When it is determined that the own vehicle (M) or oncoming vehicle (N) is traveling on a curve, the ECU (2) sets the first transfer prediction time and the second transfer prediction time such that at least one of the first transfer prediction time and the second transfer prediction time is set to a shorter time, as compared with the case where neither of the own vehicle and the oncoming vehicle is traveling on a curve, determines whether one of the own vehicle positions (M1 - M4) overlaps a corresponding one of the oncoming vehicle positions (N1 - N4), and performs avoidance control according to the result of determination.