Recognition Algorithm of Laser Echo Signal in Tuyere Raceway Based on Multiwavelet Denoising and ICA

Laser measurement of echo signal in Tuyere Raceway is characterized by significant randomness. To extract the edge signal in Tuyere Raceway, the paper started by the testing of laser echo signal on its chaotic nature, aiming to identify its nonrandomness. Multiwavelet random translation algorithm was employed for noise reduction processing of echo signals to improve the signal vs. noise ratio, which was followed by ICA blind source separation and the processing algorithm based on negative entropy. The test suggested that by the use of the recognition algorithm, the edge signal in the raceway could be better isolated. Being able to directly detect the edge of raceway, the method was proven to be of practical value.