In this research, the authors proposed a new sliding mode filter for effectively removing various noise. The filter employs a parabolic-shaped sliding surface, which is designed so that the output converges to the input value in finite-time when a constant input is provided, and thus it is referred to as parabolic sliding mode filter (PSMF). One of major advantages of PSMF is that it produces smaller phase lag than linear filters and a conventional parabolic sliding mode filter. The effectiveness of PSMF was validated through its application to a feedback, force control of a robot manipulator.
In addition, the authors presented an Adaptive Windowing PSMF (AW-PSMF), which is an improved version of PSMF, for obtaining smooth and reliable differential signal. In AW-PSMF, the window size is adaptively changed to find a window size that optimizes the trade-off between the output smoothness and the suppression of delay in numerical differentiation. The effectiveness of AW-PSMF in improving velocity feedback for position control was experimentally confirmed.
1th RSJ Advanced Robotics Best Paper Award in 2013.