Bayesian Estimator for Terrain-Aided Inertial Navigation Error Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Inertial navigation systems (INS) without additional sensors, like GNSS, produce uncorrected navigation vectors with increasing errors over time, requiring periodic stops for calibration, which is undesirable and risky for moving objects.
Innovation Solution
A method using a Bayesian estimator, such as a Rao-Blackwellized Point Mass Filter, to correct inertial navigation errors with geo-mapping data, allowing for accurate position determination of a moveable object on a terrestrial surface using only inertial sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If periodic calibration stops are implemented to reduce inertial navigation errors, then navigation accuracy is improved, but operational continuity and safety are worsened
Solution Approach 1:
The patent implements feedback by using terrain matching to continuously correct inertial navigation errors. The system compares predicted terrain profiles with actual terrain measurements and uses this feedback to adjust the navigation solution, eliminating the need for periodic stops without sacrificing accuracy.
Solution Approach 2:
The patent replaces the mechanical calibration process (stopping the vehicle to allow inertial sensors to reset) with an information-based correction system. By substituting physical calibration with computational terrain matching and error correction algorithms, the system maintains continuous operation while preserving navigation accuracy.
2Measurement precision
If additional sensors (altimeter, odometer) are added to improve position determination accuracy, then navigation precision is improved, but device complexity is worsened
Solution Approach 1:
The patent makes the terrain data serve multiple functions: it provides both the navigation solution and the error correction information simultaneously. The same terrain matching process that determines position also corrects inertial navigation errors, eliminating the need for separate correction sensors.
Solution Approach 2:
The patent introduces terrain data as an intermediary that mediates between the inertial navigation system and the external environment. Instead of adding more sensors to directly measure position, the system uses terrain information as a mediator to indirectly correct navigation errors through terrain matching.
Data Source
AI summary
Using geo-mapping data and only at least one inertial sensor, a Bayesian estimator uses at least one element of an uncorrected inertial navigation vector, of a moveable object on a surface, to estimate inertial navigation error information including corresponding statistical information. The estimated inertial navigation error information is used to error correct the uncorrected inertial navigation vector. The Bayesian filter also predicts inertial navigation error information, including corresponding statistical information, for a future time instance when a next uncorrected inertial navigation vector will be obtained.