Sensor Data Verification via Paired Comparison and Quality Counter
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sensor systems in the automobile industry face challenges in effectively utilizing redundancies for error detection and verification of measured data, leading to suboptimal robustness against random faults and systematic errors, especially due to differences in acquisition rates, synchronization, and latency among sensors.
Innovation Solution
A method and system that compares at least three values describing an identical physical quantity in pairs, using both parallel and analytical redundancies, with a quality counter system to identify and reject erroneous values, and employs an error state space Kalman filter for fusion, independent of the system's stochastic model, to enhance robustness and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensor fusion methods are used to correct measured data, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The verification process is segmented into distinct stages: initial plausibility check using stochastic models, paired comparison of values, quality counter evaluation, and final verification decision. This segmentation allows complex sensor fusion to be broken down into manageable, independent verification steps that can be implemented systematically
Solution Approach 2:
A quality counter serves as an intermediary mechanism between the paired comparison process and the final verification decision. The quality counter accumulates conformity information from multiple comparisons and provides a quantitative basis for the verification module to determine whether measured data should be accepted or rejected, simplifying the decision-making process
2Reliability
If redundancies are utilized for error detection, then reliability is improved, but loss of time occurs due to additional processing
Solution Approach 1:
Plausibility checks using stochastic models are performed preliminarily before the main paired comparison verification process. This preliminary action filters out obviously erroneous data early, reducing the computational burden of subsequent verification steps and minimizing overall processing time while maintaining reliability
Solution Approach 2:
The system performs paired comparisons for all available sensor values to maximize reliability, but the verification module selectively accepts or rejects data based on quality counter thresholds. This partial action approach ensures thorough error detection while avoiding unnecessary processing of already-verified data
3Measurement precision
If paired comparison of at least three values is performed, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The verification process is segmented into distinct stages: initial plausibility check using stochastic models, paired comparison of values, quality counter evaluation, and final verification decision. This segmentation allows complex sensor fusion to be broken down into manageable, independent verification steps that can be implemented systematically
Solution Approach 2:
The quality counter provides feedback from the paired comparison process to the verification module. By accumulating conformity information from multiple comparisons and comparing it against thresholds, the system creates a feedback mechanism that automatically adjusts verification decisions based on the consistency of measured data across multiple sensors
Data Source
AI summary
The disclosure relates to a method for verifying measured data from at least one sensor system. The measured data directly or indirectly describe values of physical quantities The values of indirectly described physical quantities are calculated from the measured data and/or from known physical and/or mathematical relationships. At least three values describing an identical quantity are subjected in pairs to a mutual comparison. In addition, at least two of the at least three values describing an identical quantity are determined independently of one another by at least one sensor system. A third value describing an identical quantity is determined by a basic sensor system. The disclosure further relates to a corresponding system and a use of the system.