A Cascaded Approach for Quadrotor's Attitude Estimation

Abstract

This paper presents a new methodology to estimate the orientation of a quadrotor using single low cost IMU sensor. The proposed solution uses two extended Kalman filters (EKF) along with a Direction Cosine Matrix (DCM) algorithm. An EKF is used to filter the noise signal of the angular rates measured by a 3-axis gyroscope sensor. Subsequently, a DCM algorithm uses the filtered gyro signal along with the reading from a 3-axis accelerometer and a 3-axis magnetometer sensor to compute the Euler angles. Finally, another EKF is presented to improve the estimation of the Euler angles. A complete simulation platform was developed using MatLab software to test the performance of the proposed method and compare it with two alternative methods.