A New Parabolic Sliding Mode Filter Augmented by a Linear Low-Pass Filter and Its Application to Position Control

This paper proposes a new sliding mode filter augmented by a linear low-pass filter (LPF) for mitigating the effect of high-frequency noise. It is based on the derivation of three new variants of Jin et al.'s (2012, "Real-Time Quadratic Sliding Mode Filter for Removing Noise," Adv. Rob., 26(8-9), pp. 877-896) parabolic sliding mode filter (J-PSMF) and investigation on their frequency-response characteristics. The new filter is developed by augmenting one of the variants of J-PSMF by a second-order linear LPF. It has better balance between the noise attenuation and signal preservation than both linear LPFs and J-PSMF. The effectiveness of the new filter is experimentally evaluated on a direct current (DC) servomotor equipped with an optical encoder. This paper also shows the application of the proposed filter to a positioning system under PDD 2 (proportional, derivative, and second derivative) control, which successfully realizes the noise attenuation and the nonovershooting response simultaneously.