Abstract: Rough surface topography influences a certain number of tribological phenomena like contact pressure, adherence, friction, wear and lubrication. Therefore it is very important to describe the character of rough surface profile. The cepstrum analysis is a new signal processing technique in recent years, and is widely applied to analyse the signal and its echo in the areas of earthquake, speech recognition, acoustics et al. A cepstrum-based analysis method is presented for the auto-correlation of rough surface profile. Some experiments of cepstrum analysis on a few kinds of representative machining specimens are made. The results show that the method based on cepstrum can be used to analyse the auto-correlation of random surface profile effectively, and its analysis sensitivity is superior to the conventional auto-correlation function. Furthermore this method has the advantage of identifying the correlation length of the profile exactly and reasonably.

Key words: Auto-correlation function, Cepstrum analysis, Roughness