A Sensor Fusion Algorithm in Humanoid Robot PD Balancing Control for Walking on Slope

Abstract

Gait is one of the main factors in build humanoid robot. Exactly, it will be different if the humanoid robot is walking on slope. This domain problem can be overcome by shifting the center of gravity (COG) of the robot, either leaning forwards or backwards to maintain its balanced body position. In this case, tilt detection system is needed. But it is not enough, because the sensor data also have a noise. So it needs an algorithm to minimize the noise of the sensor data, which it can be solved by adding a sensor fusion algorithm. Sensor fusion has the concept of combining data from several sensors to minimize uncertainty due to the noise. The IMU (Measurement Inertia Unit) is used as a tilt detection sensor. There are accelerometer sensor and gyroscopic sensor that is used in sensor fusion algorithm of the humanoid robot. This paper implements sensor fusion algorithm in balancing the robot on the slope. The test results show the addition of the sensor fusion algorithm in reading data, gives the robot capability of walking on the slope with maximum tilt 12°.