Barometric Sensor Calibration Filtering for Accurate Altitude Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing barometric air pressure sensors in mobile devices suffer from sensor drift, leading to inaccurate altitude estimates, especially in urban environments or indoor locations, which can delay emergency responses and impair navigation applications.
Innovation Solution
A method to identify and filter out erroneous calibration values from an initial dataset using calibration metrics, such as Z-metrics and indoor/outdoor filters, to generate a filtered calibration dataset for improved sensor calibration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If calibration values are derived using outdoor atmospheric pressure data, then the calibration process can be performed more frequently and with more data availability, but erroneous calibration values may be introduced due to terrain height errors, vertical displacement, HVAC pressurization, or erroneous 2D position estimates
Solution Approach 1:
The patent extracts and removes erroneous calibration values from the calibration dataset by computing a calibration metric (Z-metric) for each value and filtering out those that fall outside acceptable ranges. This separation of good data from bad data resolves the contradiction by maintaining high productivity through extensive data collection while ensuring reliability through systematic removal of erroneous values.
Solution Approach 2:
The patent implements a feedback mechanism where calibration values are evaluated against terrain database heights and 2D position estimates to compute confidence metrics. This feedback loop allows the system to identify and correct erroneous calibrations, enabling frequent calibration operations while maintaining accuracy through continuous validation.
2Ease of operation
If calibration values are derived using terrain database heights and 2D position estimates, then the calibration process can be simplified, but errors propagate from inaccurate position estimates and terrain data
Solution Approach 1:
The patent uses feedback by comparing barometric altitude estimates against terrain database heights at the device's 2D position to compute calibration metrics. This feedback mechanism validates whether the simplified calibration approach produced accurate results, allowing the system to maintain operational simplicity while correcting for errors in position estimates and terrain data through metric-based filtering.
Solution Approach 2:
The patent introduces the calibration metric (Z-metric) as an intermediary that mediates between the simplified calibration inputs (2D position, terrain height) and the final calibration output. This intermediary validates the calibration process, ensuring that simplicity does not compromise precision by filtering out calibrations where the intermediary metric indicates error.
3Measurement precision
If multiple calibration values are collected to improve calibration accuracy, then the calibration becomes more robust, but the dataset may contain erroneous values that degrade the calibration quality
Solution Approach 1:
The patent extracts erroneous calibration values from the dataset by computing calibration metrics and removing values that fall outside acceptable ranges. This extraction process resolves the contradiction by maintaining the benefits of multiple calibration values while eliminating the harmful erroneous ones that would otherwise degrade calibration quality.
Solution Approach 2:
The patent changes the parameter set by adding calibration confidence values and Z-metrics to the calibration data structure. These additional parameters enable the system to evaluate and filter calibration values, allowing multiple values to be collected for improved accuracy while identifying and excluding erroneous entries through parameter-based validation.
Data Source
AI summary
A method involves identifying multiple calibration values and corresponding calibration confidence values of an initial calibration dataset in which one or more calibration values were derived using an atmospheric pressure measurement from a barometric air pressure sensor of a mobile device. A calibration metric is determined for each calibration value. One or more of the calibration values are filtered out from the initial calibration dataset based on the calibration metric to generate a filtered calibration dataset. A filtered calibration value is determined using the filtered calibration dataset. The barometric air pressure sensor of the mobile device is calibrated using the filtered calibration value.