Vehicle Path Prediction Restriction for Traffic Indicator Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle exterior environment recognition devices struggle to accurately identify traffic indicators like traffic lights and signs, especially in environments where lane lines are difficult to recognize, leading to potential exclusion of control inputs due to mismatched predicted and actual traveling paths.
Innovation Solution
A vehicle exterior environment recognition device that includes a traveling path predicting module, a restricting module, and a control input identifying module, which predicts and restricts the traveling path based on vehicle speed, blinker state, angular speed, and steering angle to improve the accuracy of traffic indicator identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the traveling path is predicted based on current traveling condition (steering angle, angular speed), then the predicted path reflects temporary steering operations, but the predicted path does not agree with the actual traveling path
Solution Approach 1:
The system dynamically adjusts the traveling path prediction based on vehicle speed. When speed exceeds the threshold (50 km/h), the prediction is restricted to a straight line, ignoring temporary steering operations. When speed is at or below the threshold, the prediction incorporates actual steering angle and angular speed, allowing the path to curve. This dynamic adjustment resolves the contradiction by adapting the prediction method to the driving context.
Solution Approach 2:
The system changes the prediction parameters based on vehicle speed. At high speeds (>50 km/h), it uses a simplified model with zero steering angle and zero angular speed, forcing a straight path prediction. At low speeds (≤50 km/h), it uses the actual steering angle and angular speed from sensors, allowing curved path prediction. This parameter change strategy resolves the contradiction between responsiveness and accuracy.
2Productivity
If traffic indicators are identified based on predicted traveling path, then control inputs can be determined, but traffic indicators may be unintentionally excluded when prediction is inaccurate
Solution Approach 1:
The system dynamically adjusts the traveling path prediction based on vehicle speed. When speed exceeds the threshold (50 km/h), the prediction is restricted to a straight line, ignoring temporary steering operations. When speed is at or below the threshold, the prediction incorporates actual steering angle and angular speed, allowing the path to curve. This dynamic adjustment resolves the contradiction by adapting the prediction method to the driving context.
Solution Approach 2:
The system changes the prediction parameters based on vehicle speed. At high speeds (>50 km/h), it uses a simplified model with zero steering angle and zero angular speed, forcing a straight path prediction. At low speeds (≤50 km/h), it uses the actual steering angle and angular speed from sensors, allowing curved path prediction. This parameter change strategy resolves the contradiction between responsiveness and accuracy.
Data Source
AI summary
A vehicle exterior environment recognition device includes a traveling path predicting module that predicts a traveling path on which a vehicle travels, based on a current traveling condition of the vehicle, a traveling path restricting module that restricts the predicted traveling path in the width direction of the vehicle, according to at least one or more parameters selected from the group consisting of a traveling speed of the vehicle, an indicating state of a blinker, an angular speed of the vehicle, and a steering angle, and a control input identifying module that identifies a traffic indicator that exists ahead of the vehicle based on the restricted traveling path and that is to be used as the control input.