Application of wavelet analysis for thermal signals

Analysis of thermal process signals by FFT-spectrum and wavelet multiresolution analysia (MRA). Describing the different characteristic of white and color noise in power-spectrum graph. Depending on the MRA feature of white noise, a threshold method suitable for thermal signal denoise was proposed. Accuracy and robust of identification algorithm that based on minimized deviation of output improved when using wavelet de-noise. Satisfying results were got with using the proposed method when processed operation data coming from thermal process.