Online tuning method for notch filter depth in industrial servo systems

Abstract

This paper presents an online tuning method for notch filter depths in industrial servo systems. Notch filters are commonly utilized to suppress resonances in servo systems. Using a damping ratio of notch-filter zeros is especially important when the resonance has a damped behavior. This paper develops an online tuning method to adjust the damping ratio of notch-filter zeros, by adjusting the depth parameter of a bi-quad notch filter. The developed tuning method uses the electrical current command of a servo system as an input to detect a resonance. When a resonance occurs, the tuning method increases the depth parameter to decrease the damping ratio of notch-filter zeros, and then the tuning method checks whether the resonance is suppressed. If the resonance is not suppressed and the depth parameter has not reached the upper limit, the tuning method repeats the process of increasing the depth parameter. After several iterations, an appropriate depth parameter is selected and the resonance can be suppressed. Experiments are performed using two servo systems which are coupled with a ball-screw-drive load and a belt-drive load, respectively. The resonance of the belt-drive load has a behavior that is more highly damped behavior than that of the ball-screw-drive load. Experimental results show that the resonances of both systems are successfully suppressed using the developed tuning method.