An improved sliding-mode repetitive learning control scheme using wavelet transform

This paper presents an enhanced sliding-mode repetitive learning control (SMRLC) scheme using the wavelet transform. Distinct from previous wavelet transform-based repetitive control schemes, the proposed SMRLC learns from a switching signal that is equivalent to the compensation error of the SMRLC, thereby speeding up the learning process. The wavelet analysis is employed to decompose the switching signal, capture its low-frequency components effectively, and synthesize a de-noised, high-scale signal for the learning process. Experimental study on a directly driven Hoekens straight-line mechanism is conducted, and the proposed wavelet transform-based SMRLC is experimentally compared with a Fourier series-based SMRLC.