Performance improvement of an industrial servopositioner using load side acceleration measurements

Nowadays, industrial servopositioning systems are used in many applications that demand high positioning accuracy, while requiring at the same time high dynamic movements. However, these systems often suffer from mechanical non-idealities, primarily the elasticity induced by the trasmission, which leads to load oscillation when using a standard collocated controller approach. While load side position sensors can be added to allow the use of control strategies that reduce the vibrations, this solution is often impractical in industrial environment. In this paper, we propose the use of the Acceleration Aided Kalman Filter (AAKF), which requires a low-cost load side MEMS accelerometer, to obtain accurate and robust estimates of load variables. Using load side information, two different control approaches are implemented, a PID with Active Damping and a Model Predictive Controller (MPC). The experimental results confirm that both control algorithms work as intended, considerably reducing load side vibration while using high dynamic movements, therefore reducing the cycle time of the positioning. Thus, the AAKF is validated as an effective alternative to the standard load position sensor in industrial environment.