AEB Parameter Calibration for Personalized Braking and Steering Timing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current autonomous emergency braking systems use a standard 'time-to-activate braking' (TTAB) that does not account for individual user differences, leading to suboptimal collision avoidance strategies.
Innovation Solution
A method and apparatus that self-adaptively calibrate TTAB and 'time-to-activate steering' (TTAS) using a deep learning model, incorporating vehicle kinematics, user operation, object ahead, and road attributes to determine personalized corrected values based on driving behaviors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a standard TTAB value is used for all users, then the system is simple and easy to operate, but it does not account for individual user differences leading to suboptimal collision avoidance
Solution Approach 1:
The system dynamically adjusts TTAB and TTAS parameters based on real-time detection of user driving characteristics. The calibration module continuously learns and adapts to individual user behaviors, transforming static standard values into dynamic personalized parameters that evolve with user preferences.
Solution Approach 2:
The system performs self-calibration by automatically detecting user driving patterns and adjusting parameters without requiring manual user input or configuration. The calibration module autonomously analyzes driving data and determines optimized TTAB and TTAS values, enabling the system to serve itself in the parameter customization process.
2Measurement precision
If deep learning methods are used to determine calibrated values, then personalized calibration accuracy is improved, but computational complexity and processing time increase
Solution Approach 1:
The system performs preliminary calibration during periods when the vehicle is not in emergency braking mode, accumulating driving data and pre-computing personalized parameters. This advance calibration preparation reduces the computational burden during critical braking situations, as the deep learning model has already processed the data and determined optimized parameters beforehand.
Solution Approach 2:
The patent replaces traditional mechanical calibration methods with deep learning-based computational approaches. Instead of manual parameter adjustment or simple rule-based systems, the invention uses neural networks to automatically analyze driving patterns and determine optimal TTAB and TTAS values, achieving higher precision through computational intelligence.
Data Source
Figure 1
Figure 2~4
AI summary
Disclosed are a method and apparatus for calibrating vehicle anti-collision parameters and a computer storage medium for implementing the method. The method for calibrating the vehicle anti-collision parameters comprises the following steps: setting a standard value of the latest time to activate braking TTAB and a standard value of the latest time to activate steering TTAS of a present vehicle; determining a calibrated value of TTAB and a calibrated value of TTAS; and combining the standard value of TTAB and the standard value of TTAS with respective corresponding calibrated values to determine a corrected value of TTAB and a corrected value of TTAS. Since the corresponding corrected value of TTAB and the corresponding corrected value of TTAS may be self-adaptively determined, driving safety and user experience are improved.