Vehicle Localization via Environment-Class Adaptive Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The determination of a high-accuracy vehicle position required for partly, highly, or fully automated operation is hindered by interfering objects in the environment, which existing methods struggle to filter out effectively, impacting safety and efficiency in automated vehicle operation.
Innovation Solution
A method that detects environmental data values, determines the environment class, and applies a suitable filter to remove interfering objects, allowing for the precise localization and automated operation of vehicles by selecting the appropriate filter based on the environment class, thereby ensuring accurate positioning and enhanced safety.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple filters are applied to remove interfering objects, then the accuracy of position determination is improved, but the computational capacity requirement increases
Solution Approach 1:
The patent applies different filtering strategies and levels of complexity depending on the specific environment class detected. For example, in pedestrian zones with many static objects, simpler filters may suffice, while in dynamic environments like construction sites, more sophisticated filters are applied. This selective approach maintains high position accuracy while optimizing computational resource usage based on local environmental characteristics.
Solution Approach 2:
The system dynamically adjusts filtering parameters such as threshold values, search windows, and filter sensitivity based on the detected environment class. When transitioning from a rural environment to an urban environment, the system modifies its filtering parameters to match the characteristic interference patterns of each environment type, thereby maintaining accuracy without requiring maximum computational power in all scenarios.
2Productivity
If environment-specific filters are selected based on environment class, then the position determination speed is improved, but the system complexity increases
Solution Approach 1:
The system performs preliminary classification of the environment into predefined environment classes before applying specific filters. By categorizing the environment first (e.g., rural, urban, pedestrian zone, construction site), the system can then select from a prepared set of environment-specific filters, avoiding the need to evaluate all possible filters and thereby improving processing speed while managing complexity through structured classification.
Solution Approach 2:
The system dynamically adapts its filtering approach by continuously monitoring environmental characteristics and adjusting the selected filter set accordingly. When the environment class changes during vehicle operation, the system transitions between different filter configurations, optimizing performance for each environment type while maintaining a manageable system architecture through dynamic reconfiguration rather than static complexity.
3Reliability
If interfering objects are filtered out to ensure safe automated operation, then the safety is improved, but the time required for position determination increases
Solution Approach 1:
The system applies safety-critical filtering only in environment classes where interfering objects pose significant risks to automated operation. For example, in pedestrian zones or construction sites where false detections could lead to unsafe decisions, more thorough filtering is applied. In contrast, in open rural environments with fewer hazard types, the system uses lighter filtering, thereby maintaining safety where needed while minimizing time loss in lower-risk environments.
Data Source
AI summary
A method and an apparatus for localizing and automatically operating a vehicle, including the task of detecting environmental data values that represent an environment of the vehicle, the environment of the vehicle including at least one interfering object; determining an environment class of the environment of the vehicle; determining a high-accuracy position of the vehicle based on the environmental data values, at least one filter being applied to filter the at least one interfering object out of the environmental data values, the at least one filter being selected as a function of the environment class, and the high-accuracy position being determined after the at least one interfering object is filtered out; and automatically operating the vehicle as a function of the high-accuracy position.