Abstract :
Orientation estimation with sensors of accelerometer and gyroscope shows quite
large errors under certain conditions. Calculations with the sensor accelerometer show
reliability when static but have an error when dynamic. Calculations with the sensor
gyroscope show reliable static and dynamic conditions but errors continue to accumulate
during data recording which makes the orientation calculation less accurate.
This study aims to produce a more accurate orientation with zero acceleration
compensation algorithm. The method used is to measure the acceleration value of thesensor
covariance value accelerometer. The combination of the sensors accelerometer and
gyroscope (IMU) using the method Kalman filter plus the algorithm is zero acceleration
compensation intended to compensate for the orientation calculation on the accelerometer
when the acceleration is static and the orientation calculation using the gyroscope sensor
when it is dynamic.
The results of this study are orientation quadrotor using the Kalman filter, which can
reduce errors and changes in orientation results with zero acceleration compensation
algorithm compared to HP orientation results. The orientation results produced an error in
the attitude roll with an accuracy of 96.77%, and an error in the attitude pitch with an
accuracy of 98.81%.