Vehicle Positioning via Odometry Drift Correction in GPS Gaps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current vehicle positioning methods, such as satellite-based GPS and odometry, face limitations in accuracy and reliability, especially for precise localization required in autonomous driving, and are costly and resource-intensive, with errors accumulating over longer routes.
Innovation Solution
A method that corrects vehicle-specific odometry using high-precision localization data from sensors and digital maps, accounting for vehicle-specific and external influences, to enhance positioning accuracy without additional hardware, by continuously evaluating and compensating for drift in odometric measurements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If satellite-based positioning systems (GPS) are used for vehicle localization, then positioning coverage is improved, but positioning accuracy deteriorates (meter range instead of decimeter/centimeter range)
Solution Approach 1:
The patent combines satellite-based positioning (GPS) with odometric measurement methods to create a hybrid positioning system. The GPS provides coarse positioning coverage while odometry provides fine-grained accuracy, merging the strengths of both methods to achieve both wide coverage and high precision localization
Solution Approach 2:
The patent introduces odometric measurements as an intermediary method to bridge the accuracy gap between GPS (meter level) and the required precision for autonomous driving (decimeter/centimeter level). The odometry system acts as a mediator that refines GPS positions through continuous relative position measurements
2Measurement precision
If HD maps and environmental sensors are used for precise localization, then positioning accuracy is improved, but system cost and data storage requirements deteriorate
Solution Approach 1:
The patent extracts the essential positioning function from complex HD map systems and environmental sensor arrays, isolating the core odometric measurement capability. By focusing on wheel revolution counting and basic sensor fusion, the system achieves precise localization without requiring expensive HD maps or multiple environmental sensors
Solution Approach 2:
The patent uses inexpensive odometric sensors (wheel encoders, basic inertial sensors) that can be easily replaced or recalibrated, replacing the need for expensive, complex sensor suites and HD map infrastructure. The simple sensor model allows for cost-effective implementation while maintaining precision
3Device complexity
If odometry is used for position determination, then system cost is reduced, but positioning accuracy deteriorates due to error accumulation over longer routes
Solution Approach 1:
The patent implements feedback mechanisms where odometric measurements are continuously monitored and corrected based on periodic GPS position updates and detected environmental features. This feedback loop prevents error accumulation by periodically resetting and adjusting the odometric drift, maintaining accuracy over long routes
Solution Approach 2:
The patent performs preliminary calibration of odometric parameters (wheel circumference, steering geometry) before operation and continuously adjusts these parameters based on accumulated data. This preliminary setup and ongoing adjustment prevent error accumulation by maintaining accurate baseline measurements
Data Source
Figure 1
Figure 2
AI summary
The method according to the invention for determining the position of a vehicle involves the calculation of highly accurate location data for positions of the vehicle (F) while travelling on a route (1). Odometric measured variables are captured for the vehicle-individual odometrics of the vehicle (F) while travelling on the route (2). The calculated highly accurate location data and the captured odometric measured variables are evaluated jointly (3). The vehicle-individual odometrics are corrected on the basis of the evaluation (4). This involves the use of an error model for the vehicle-individual odometrics to calculate a vehicle-specific drift, and the continual correction of the vehicle-individual odometrics for as long as highly accurate location data can be calculated. The corrected vehicle-individual odometrics can then be used to determine the position of the vehicle in areas in which highly accurate location data cannot be calculated (5).