Sensor Fusion Using Kalman Filter for a Quadrotor-Attitude Estimation

Quad rotors are types of unmanned air vehicles that became an attractive topic to an enormous number of people worldwide. This book focuses on sensor fusion of low-cost inertial sensors: a tri-axis accelerometer, a tri-axis gyroscope and a tri-axis magnetometer. Three types of filters were implemented in this book: two models of linear complementary filter (LCF), two models of linear Kalman filter (KF) and a quaternion-based extended Kalman filter (EKF). A total of five filter models were tested and verified using experimental data sets through three distinct scenarios. Six different data sets were chosen, each with antithetic characteristics, to test the filters on them. On the basis of the results of this research, it can be concluded that taking into consideration accelerometer bias improves the performance of the filter. KF shows the best performance as it combines the characteristics of short computing time and accurate estimation results suiting the aim of the book which is to allow the quad rotor to hover.