Key points are not available for this paper at this time.
Sensor Fusion (Complementary and Kalman filters) and Moving Average filter are implemented on an Arduino microcontroller based data acquisition of rotation degree from Inertial Measurement Unit (IMU) sensor for stabilized platform application. Stable platform prototype is designed to have two degrees of freedom, roll and pitch rotation. Output data from gyro and accelerometer were combined to take the advantage of each sensor. Digital filter algorithm was embedded into microcontroller programming. This paper analyzes overshoot percentage, rise time, and data series smoothness of Sensor Fusion (Complementary and Kalman filter) and Moving Average filter response in IMU data acquisition from step input of 20-degreerotation. Moving-average filter resulted in the smallest overshoot percentage of 0% but produce theslowest responsewith 0.42 second rise time. Overall best results are obtained using Complementary filter (alpha value 0.95) by overshoot percentage of 14.17%, 0.24 second rise time, and 0.18 data series smoothness.
Redhyka et al. (Thu,) studied this question.