Intelligent Real-Time MEMS Sensor Fusion and Calibration

Abstract

This paper discusses an innovative adaptive heterogeneous fusion algorithm based on estimation of the mean square error of all variables used in real time processing. The algorithm is designed for a fusion between derivative and absolute sensors and is explained by the fusion of the 3-axial gyroscope, 3-axial accelerometer and 3-axial magnetometer into attitude and heading estimation. Our algorithm has similar error performance in the steady state but much faster dynamic response compared to the fixed-gain fusion algorithm. In comparison with the extended Kalman filter the proposed algorithm converges faster and takes less computational time. On the other hand, Kalman filter has smaller mean square output error in a steady state but becomes unstable if the estimated state changes too rapidly. Additionally, the noisy fusion deviation can be used in the process of calibration. The paper proposes and explains a real-time calibration method based on machine learning working in the online mode during run-time. This allows compensation of sensor thermal drift right in the sensor's working environment without need of re-calibration in the laboratory.

Attitude Determination Method by Fusing Single Antenna GPS and Low Cost MEMS Sensors Using Intelligent Kalman Filter Algorithm

“…It has obvious advantages over the traditional centralized EKF algorithm, but heavy calculation burden could not satisfy the demands of real time processing. The other most widely researched method is that the attitude estimate using CF fusion algorithm is utilized as the state coefficient [26] or measurement vector [16,17,25] in modified Kalman filter, which could be regarded as one benchmark algorithm in the field of GPS and INS data fusion. Among this category algorithm, the quaternion estimation algorithm (QUEST) [3,11,25] is usually used as attitude algorithm for Micro IMU, which could remove the adverse effect by linear acceleration of the navigation system to attain good dynamic performance.…”

Improving the Measurement Accuracy of SINS Through Calibration on Horizontal Three-axis Turntable

“…To minimize its impact on navigation accuracy, the output errors must be modelled and calibrated before use, and compensated during the utilization. The common calibration methods are classified into the separated and the systematic calibration method [5,6].…”

“…In systematic calibration, the output errors of the IMU, such as attitude error, velocity error and position error, are treated as observed quantities [7][8][9][10], and used to calibrate the parameters in the IMU error model [11][12][13][14][15], reducing the calibration dependence on the turntable's accuracy. This method is a desirable way to carry out filed calibration, and a hotspot in relevant research [16].…”

New methods in the application of inertial and magnetic sensors in online pattern recognition problems

“…These errors have especially high effect in applications which utilize accelerometers and gyroscopes to obtain position and orientation, since the measurements need to be integrated. Because of the integration process, even very small errors at the output accumulate very rapidly and the position error becomes considerably large [50]. Although the measurements in the proposed system are not integrated during feature extraction, it is very important to calibrate the sensors before use, since the errors can affect also the movement classification process.…”

“…The first one uses measurements acquired in specific stationary positions or during specified movements, which are utilized to compute calibration parameters based on different basic principles. The second is the Kalman filter-based approach [50,[55][56][57], which is a widely known state estimation technique which tries to obtain the unknown parameter based on the system model and observations accrued over a period of time. These techniques aim to estimate navigation states along with calibration parameter together in one framework.…”