Sistem Kendali Penerbangan Quadrotor pada Keadaan Melayang dengan Metode LQR dan Kalman Filter

Abstract

Quadrotor is a type of UAV (Unmanned Aerial Vehicle) with four propellers and four rotor. Quadrotor as flying robots has the advantage to take off and land vertically. In addition quadrotor also has the ability to fly hovered near a stationary state. However quadrotor had some difficulties to operate. One of these difficulties is to make quadrotor be able to fly and maintain the stationary state of the Euler angles (roll, pitch, and yaw). Linear Quadratic Regulator (LQR) as one of the modern control method which has the advantage of maintaining the conditions on the ground. This method can be combined with Kalman filter algorithm. It aims to reduce measurement error from the process sensor fusion and maintain Euler angles (roll, pitch and yaw).

Kalman filter aims to reduce the measurement error of the sensor fusion. Then the output of Kalman filter algorithm becomes the input state for control LQR the roll angle and pitch angle. Input state is multiplied with the negative feedback  as process systems. The results are converted into pulses to rotate the brushless motor so quadrotor can fly stably.

The test results showed quadrotor while maintaining stability against roll angle has overshoot of 0.35 ° and the pitch angle has overshoot of 2 °.