| Summary: | This paper presents an EMD (empirical mode decomposition)-based extraction approach to decouple the electric field fringe component impacts on the nonlinearity errors of a micro encoder-like capacitive displacement sensor. A calculated capacitance model built under Maxwell’s equations against the impacts of electric field fringe component indicates that signals of this sensor are all in periodic co-sinusoidal waveforms. Applying the proposed EMD scheme, signals are decomposed into sets of intrinsic mode functions (IMFs) and a residual, in which a fundamental component represents all the features of sensor signal and reserves the local information. Interpreting sensor information from the extracted component instead of the original signal drastically diminishes the impacts of electric field fringe effect. Results from a test bench shows that after applying the EMD-based extraction approach, waveform errors were decreased from over 4.18% to less than 0.89%. Nonlinearity of the interpreted displacement was decreased from 1.54% to 0.29% for 8 mm stroke.
|
|---|
