Quadrotor as one kind of Unmanned Aerial Vehicle (UAV) can fly manually using remote controller or autonomously. Quadrotor that flies autonomously will get interference that affects the sensor readings. This condition causes the system cannot reach steady state and requires a long settling time when controlled. Sensory readings error can be reduced using Kalman Filter, but Kalman Filter has heavy computation. So Kalman Filters can be combined with Madgwick quaternion fusion that has light computing. The output of Madgwick quaternion fusion can be optimised for processing using Kalman Filter. Therefore, it takes Madgwick quaternion Kalman Filter approach to reduce read error from the fusion sensor then the result becomes an input state for LQR control to maintain quadrotor balance. The results of this research indicate that quadrotor can overcome the noise and reach steady-state quickly. The system responses show that it takes rise time for 0.4 seconds, 0.3 seconds, and rise time for 1.6 seconds to stabilise roll, pitch, and yaw motion respectively. These results mean that the quadrotor control system has a reasonably fast response. The control system is also capable of eliminating excessive overshoot.
Auto VTOL System on Quadrotor Using Madgwick Quaternion Kalman Filter and LQR
2018-08-01
708412 byte
Conference paper
Electronic Resource
English
Novel Quaternion Kalman Filter
Online Contents | 2006
|Novel quaternion Kalman filter
IEEE | 2006
|IEEE | 2015
|A Novel Quaternion Kalman Filter
AIAA | 2002
|Quadrotor dual quaternion control
IEEE | 2015
|