Attitude Estimation for Quadrotor Based on IMU with Kalman-Filter

Abstract

Improved a quadrotor technology that is capable of fast maneuvering requires accurate attitude estimation or navigation to maintain a quadrotor stability. The GPS can provide position measurements with an accuracy of several meters, but cannot provide orientation information directly. This study aims to design a quadrotor attitude navigation system based on IMU sensors on AR Drone 2.0 with a Kalman filter using the equation of state space model. The system model was developed using the Matlab software. The Kalman filter is designed as an estimator to reduce noise on the sensor so that it can improve measurement accuracy. The test results showed that the system model can be used to estimated the orientation angle and shift of the quadrotor, while the Kalman filter that designed can reduce noise in the sensor data. At the time of tested, the system provided the measurement accuracy of above 90% when tested indoor.