Gyroscope Bias Estimation Using Moving Statistics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining the current bias of a gyroscope in mobile devices are computationally complex and require data from multiple motion sensors, making them inefficient for real-time correction.
Innovation Solution
A method that calculates a rotation rate moving average and moving standard deviation from gyroscope data, updating the gyroscope bias when the moving standard deviation falls below a threshold, using a processor to receive and process rotation rate data from a gyroscope, allowing for efficient bias estimation without relying on additional sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If combined data from multiple motion sensors (accelerometer, magnetometer, gyroscope) is used to correct gyroscope bias, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent extracts and uses only gyroscope rotation rate data to estimate bias, eliminating the need to process data from accelerometers and magnetometers. This simplifies the system by taking out the unnecessary sensors while maintaining bias correction capability through the moving average and moving standard deviation calculations applied solely to gyroscope data.
Solution Approach 2:
The gyroscope data serves multiple functions: it provides both the rotation rate information needed for bias estimation and the motion detection capability. The single sensor performs what previously required multiple sensors, making the system more universal and less complex while maintaining measurement precision through sophisticated processing of the gyroscope's own data.
2Measurement precision
If data from multiple motion sensors is processed simultaneously to detect gyroscope bias, then measurement precision is improved, but productivity decreases
Solution Approach 1:
The patent extracts only the necessary gyroscope rotation rate data for bias estimation, removing the computational burden of processing accelerometer and magnetometer data. This extraction principle improves productivity by focusing computational resources on the single most relevant data source while maintaining measurement precision through advanced processing techniques applied to this extracted data.
Solution Approach 2:
The patent applies partial action by using only the gyroscope data subset needed for bias estimation rather than processing all available sensor data. The moving average and moving standard deviation calculations are applied selectively to gyroscope rotation rates, achieving sufficient measurement precision with reduced computational effort compared to processing all sensor data simultaneously.
3Measurement precision
If sophisticated processing of multiple sensor data is used to correct gyroscope bias, then measurement precision is improved, but use of energy increases
Solution Approach 1:
The patent extracts and processes only gyroscope rotation rate data, eliminating the energy consumption associated with processing accelerometer and magnetometer data. This extraction approach maintains measurement precision through focused computational effort on the essential data source while significantly reducing overall energy usage by removing unnecessary processing operations.
Solution Approach 2:
The patent applies partial action by performing sophisticated processing (moving average, moving standard deviation) only on the necessary gyroscope data rather than all sensor data. This selective processing maintains adequate measurement precision while reducing energy consumption by avoiding excessive computational operations on redundant sensor data.
Data Source
AI summary
A method for determining a current estimated gyroscope bias of a gyroscope configured to output rotation rate data includes receiving first rotation rate data associated with a first time from the gyroscope, the first rotation rate data comprising a first rotation rate reading that indicates a rotation rate of the gyroscope about a first axis; calculating a rotation rate moving average and a rotation rate moving average associated with a second time earlier than the first time; calculating a moving standard deviation based on the first rotation rate data, and a moving standard deviation associated with the second time; determining if the moving standard deviation associated with the first time is less a threshold moving standard deviation; and in response the moving standard deviation being less than the threshold moving standard deviation, using the first rotation rate reading to update the current estimated gyroscope bias.