Inertial Navigation Filter Architecture for Defect Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing positioning systems with multiple inertial measurement units struggle to detect defective units or strong drifts, as existing methods fuse signals before comparison with GPS data, limiting accuracy and reliability.
Innovation Solution
A positioning system with independent navigation filters for each inertial measurement unit, where a mean estimate is calculated without reinjecting it into individual filters, allowing for detection of defective units and achieving an optimum estimate with limited calculation resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If inertial measurement units are combined by averaging signals before processing, then measurement precision is improved, but the ability to detect defective units is lost
Solution Approach 1:
The patent divides the signal processing into separate stages: first processing each inertial measurement unit independently through its own navigation filter, then combining the results. This segmentation allows defective units to be identified at the individual filter level before final combination, resolving the contradiction between maintaining precision and enabling defect detection.
Solution Approach 2:
The patent introduces an intermediary fusion stage that combines results from multiple independent navigation filters. This intermediary layer allows the system to detect anomalies in individual units while maintaining the benefits of multiple measurement sources, enabling both precision and defect detection capabilities.
2Reliability
If multiple inertial measurement units are used to improve reliability, then system reliability is improved, but calculation resources are increased
Solution Approach 1:
The patent segments the calculation workload by assigning independent navigation filters to each inertial measurement unit, allowing parallel processing. This segmentation maintains reliability through multiple units while reducing overall calculation complexity through distributed computation rather than centralized processing.
Solution Approach 2:
The patent processes each inertial measurement unit independently with its own navigation filter, then combines only the necessary outputs. This partial processing approach maintains reliability through comprehensive unit analysis while avoiding the excessive computational burden of fully integrating all units simultaneously.
3Difficulty of detecting and measuring
If independent navigation filters operate without reinjecting mean estimate, then defective unit detection is enabled, but estimation optimality may be compromised
Solution Approach 1:
The patent implements feedback through the fusion stage, where the mean estimate from combining multiple independent filters is fed back to improve individual filter outputs. This feedback mechanism maintains estimation optimality while preserving the ability to detect defective units through the independent filter architecture.
Solution Approach 2:
The patent merges the outputs of multiple independent navigation filters through a fusion stage that calculates a mean estimate. This merging process combines the strengths of independent filtering (defect detection) with the optimality of combined estimation, resolving the contradiction between detection capability and estimation quality.
Data Source
AI summary
Disclosed is a positioning system including: several inertial measurement units; at least one common sensor, providing a measurement of a positioning parameter of a system; for each inertial measurement unit, a navigation filter configured to: a) determine an estimate of the positioning parameter, on the basis of an inertial signal provided by the inertial measurement unit; and to b) correct the estimate, as a function of the measurement and of a correction gain that is determined on the basis of an augmented variance higher than the variance of a measurement noise of a common sensor; and—at least one fusion module determining a mean of the estimates, the mean being not reinjected at the input of the navigation filters. Also disclosed is an associated positioning method.