Position control system based on inertia measurement unit sensor fusion with Kalman filter

Abstract

This paper proposes position control system based on measurements of an inertia measurement unit (IMU) sensor (composed of a gyro sensor and an acceleration sensor) attached on the tip position of a 2-link planar manipulator. To estimate joint angle from only one IMU sensor, velocity applied to end effector is required. However, it is difficult to measure accurate velocity from integration of measurements of the acceleration sensor due to noise, offset and drift error. Therefore, Kalman filter and sensor fusion with acceleration sensor and gyro sensor are introduced to estimate the velocity with high accuracy. In addition to that, Disturbance observer (DOB) is used in the position control system, and the estimated angular velocity information is utilized in DOB. To confirm the performance of proposed control system, 3 types of simulation of position control are conducted. Kalman filter can reduce the noise effect and position control is achieved by proposed control system.