Autonomous Vehicle Lane Tracking via Inertial Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current lane tracking systems for autonomous vehicles suffer from low update frequencies of lane detection information, leading to reduced control precision and performance, especially in curved roads and varying light conditions, due to the time lag in processing camera images.
Innovation Solution
A lane tracking method and system that utilizes a processing unit to store reference and past location data, calculates estimated yaw rates and lateral accelerations, and estimates future lane line data based on vehicle motion information, allowing for more frequent and accurate lane tracking updates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If lane detection information is updated at a low frequency (10-20 Hz) due to image processing time, then the system can maintain stable operation, but the control precision and lane tracking performance deteriorate, especially in curved roads or when the vehicle is traveling with larger lateral speed
Solution Approach 1:
The system pre-calculates and stores historical lane line data and vehicle location data at high frequency before they are needed for control decisions. By having this data prepared in advance, the system can provide high-frequency update information to the control system without requiring real-time image processing at that frequency, thus resolving the contradiction between stable operation and control precision.
Solution Approach 2:
The system dynamically adjusts the lane tracking approach by combining low-frequency actual lane detection data with high-frequency predicted lane line data. The prediction model adapts to different driving conditions (curved roads, lateral speed variations) to maintain control precision while allowing the image processing to operate at a stable, lower frequency.
2Ease of operation
If the vehicle control system operates at a frequency limited by the lower update frequency of lane detection information, then the system can maintain simplicity in operation, but the resolution and precision of control instructions are reduced
Solution Approach 1:
The system introduces an intermediate data processing layer that generates high-frequency predicted lane line data based on low-frequency actual lane detection and vehicle motion information. This intermediary layer acts as a bridge, allowing the control system to receive high-resolution control instructions without requiring the image processing system to operate at high frequency, thus maintaining operational simplicity while improving control precision.
3Device complexity
If conventional lane tracking methods are used, then the system can maintain simplicity, but the system cannot provide correct lane line information when lane lines have deteriorated, are unclear or absent, or present abnormal color contrast due to variations in light conditions
Solution Approach 1:
The system prepares compensatory measures in advance by maintaining historical lane line data and vehicle location data. When lane markings become unclear or absent, the system can rely on the predicted lane line data generated from this pre-prepared information, cushioning against the failure of conventional detection methods without requiring complex real-time alternative detection systems.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enhances the robustness and precision of lane tracking by increasing the refresh rate of lane line data from 10 Hz to 100 Hz, improving control accuracy and handling temporary failures in lane detection, such as deteriorated or unclear lane markings.
Implementation Method 1
calculating an estimated yaw rate and an estimated lateral acceleration that correspond to the current time point based on an angular speed and an acceleration of the autonomous vehicle which are measured by an inertial measurement unit of the autonomous vehicle at the current time point
Data Source
AI summary
A lane tracking method is proposed for use by an autonomous vehicle running on a lane. A future location of the autonomous vehicle that corresponds to a future time point is estimated based on a current location and a measurement result of an inertial measurement unit of the autonomous vehicle. The future location of the autonomous vehicle, and a reference lane line data and a reference past location that correspond to a reference past time point are used to estimate a future lane line data that corresponds to the future time point.