Orientation Estimation Using Preprocessed MEMS Sensor Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating the orientation of objects using MEMS sensors are inefficient, noisy, and biased, particularly when dealing with magnetic disturbances, and require significant computing power, making them unsuitable for microcontrollers.
Innovation Solution
A method that preprocesses measurements to detect disturbances and estimates orientation using a simplified approach by directly incorporating corrected sensor measurements into the state vector, reducing the need for complex Kalman filter calculations and prioritizing reliable sensor outputs based on detected disturbances.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If complex Kalman filter calculations are used for orientation estimation, then measurement precision is improved, but computing power requirements increase
Solution Approach 1:
The method segments the orientation estimation process into distinct phases: a preprocessing phase that detects disturbances and corrects measurements, and a main estimation phase that uses simplified calculations. This segmentation allows complex filtering operations to be performed only when necessary, reducing overall computational load while maintaining precision.
Solution Approach 2:
The preprocessing step performs preliminary disturbance detection and measurement correction before the main orientation estimation. By preparing corrected measurements in advance, the subsequent estimation phase can use simpler, less computationally intensive algorithms while still achieving high precision results.
2Measurement precision
If disturbance detection and correction steps are added, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The method extracts and isolates the disturbance detection and correction operations into a separate preprocessing module. This extraction allows the complex disturbance handling logic to be separated from the main estimation algorithm, making both parts simpler and more manageable while improving overall measurement precision.
Solution Approach 2:
The preprocessing step acts as an intermediary between raw sensor measurements and the main orientation estimation algorithm. It transforms noisy, disturbed measurements into corrected measurements that are more suitable for simplified estimation algorithms, thereby improving precision without significantly increasing the complexity of the main estimation process.
3Measurement precision
If multiple sensor modalities are fused, then orientation estimation accuracy is improved, but ease of operation deteriorates
Solution Approach 1:
The method dynamically adapts the sensor fusion process based on detected disturbance conditions. During normal operation, simplified fusion rules are applied for ease of implementation. When disturbances are detected, the preprocessing step activates more sophisticated correction mechanisms. This dynamic adaptation maintains accuracy while preserving implementation simplicity during typical operation.
Data Source
Figure 1
Figure 2
Figure 2b
AI summary
The invention relates to a method for estimating the orientation of an object in space at a moment k using measurements of the total acceleration (yA), of the magnetic field (yM), and of the rotational speed (yG) of said object on the three spatial axes, comprising a step of preprocessing (210, 220, 230) said measurements (yA, yM, yG) at a moment k so as to detect the existence of an interference in said measurements, and calculate corrected measurements at the moment k, and a step of estimating the orientation (formula I) at the moment k by an observer from the corrected measurements (formula II) at the moment k from the preprocessing step. The estimation step advantageously includes the use of matrices for transforming the object frame of reference into a fixed frame, said matrices being applied, in the event of an interference, to a quaternion deduced from the integration of a rotational-speed measurement made with an orientation quaternion calculated at the moment k-1.