Loading...

Robust Orientation Estimation Using Imu and Online Machine Learning Based Calibration in the Presence of Distortions

Golmohammad, Sadjad | 2021

1918 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 53827 (08)
  4. University: Sharif University of Technology
  5. Department: Mechanical Engineering
  6. Advisor(s): Khodaygan, Saeed
  7. Abstract:
  8. In this project an optimized and robust orientation estimation method using IMU and magnetic sensors is presented. Magnetic distortion effects in orientation estimation is also one of the main purposes. Proposed sensor fusion algorithm is based on a complementary filter which provides a quaternion estimation as the algebraic solution of a system from inertial/magnetic observations. To develop the basic sensor fusion algorithm some procedures including a simple calculation to deal better with non-gravitational accelerations, decrease the effect of magnetometer in the presence of distortions and online gyroscope bias estimation is added. Also, a method for classification the different types of magnetic distortion and eliminating them by algebraic formulation in provided. To detect the type of magnetic disturbance, a decision tree is trained using a huge number of supervised observations. The main point of detection and elimination method for magnetic disturbances is the independency of this method from sensor fusion algorithm which make it possible to be used in other different sensor fusion algorithms such as Kalman filter based methods. Using the proposed method for detection and elimination magnetic distortions in the proposed sensor fusion method decreased the RMS error of heading angle to 76% and also decreased the error to 58% in Madgwick’s algorithms
  9. Keywords:
  10. Sensor Fusion ; Decision Making Tree ; Onboard Calibration ; Magnetic Disturbance ; Attitude and Heading Reference System (AHRS) ; Machine Learning

 Digital Object List

 Bookmark

...see more