Gyro Sensor Output Correction Using Quaternion Posture Calculation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional positioning systems using gyro sensors struggle to accurately calculate the azimuth and inclination angles of a moving body, especially when indoors or in tunnels where GPS signals are unavailable, and when the moving body is stationary, leading to reduced accuracy of GPS measurement data.
Innovation Solution
A positioning system that includes a gyro sensor, a geomagnetic sensor, and an acceleration sensor, where the output of the gyro sensor is corrected by converting it into a rotation matrix or quaternion based on the outputs of the geomagnetic and acceleration sensors, performing normalization and orthogonalization, and using correction coefficients to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GPS signal is used for positioning, then positioning accuracy is improved, but positioning reliability deteriorates when GPS signals are unavailable (indoors or in tunnels)
Solution Approach 1:
The patent combines GPS positioning data with inertial sensor data (gyro sensor and acceleration sensor) to create a hybrid positioning system. The arithmetic controller integrates multiple data sources including position/velocity from GPS, posture from quaternion calculation, and error corrections from extended Kalman filter to maintain positioning reliability when GPS signals are unavailable.
Solution Approach 2:
The system implements feedback through error estimation and correction mechanisms. The extended Kalman filter continuously estimates errors in position, velocity, posture, and sensor biases, then applies corrections to maintain accuracy. This feedback loop ensures positioning reliability by compensating for drift and errors in real-time.
2Ease of operation
If gyro sensor data is integrated over time to calculate posture, then posture detection capability is improved, but integration error accumulation deteriorates measurement precision
Solution Approach 1:
The extended Kalman filter provides continuous feedback to estimate and correct gyro bias errors and integration errors in posture calculation. By comparing inertial sensor data with GPS positioning data and quaternion-based posture calculations, the system identifies and corrects drift accumulation, maintaining posture measurement precision over time.
Solution Approach 2:
The patent replaces pure mechanical integration of gyro data with a computational approach using extended Kalman filter and quaternion mathematics. This substitution transforms the error-prone integration process into a corrected computational model that maintains precision by continuously estimating and correcting errors based on multiple data sources.
3Measurement precision
If azimuth sensor is used to detect true geomagnetism, then azimuth detection accuracy is improved, but device complexity increases due to magnetic field cancellation requirements
Solution Approach 1:
The system uses the acceleration sensor and gyro sensor data to calculate posture and azimuth through quaternion operations, making the azimuth sensor optional rather than essential. This multi-functionality allows the system to achieve azimuth detection through alternative means, reducing device complexity while maintaining accuracy through the same extended Kalman filter correction mechanism.
Data Source
AI summary
A positioning system includes: a gyro sensor configured to detect an angular velocity of a moving body; a geomagnetic sensor configured to detect a direction in which the moving body is placed; an acceleration sensor configured to detect an acceleration of the moving body; and a gyro sensor output correction part configured to correct an output of the gyro sensor, wherein the gyro sensor output correction part is configured to convert output components of the gyro sensor into at least one of a rotation matrix and a quaternion based on output components of the geomagnetic sensor and output components of the acceleration sensor, and to calculate a posture of the moving body based on the rotation matrix or the quaternion.