Approach of impulse interfere noise cancellation based on blind source separation

The paper analyzed the character, the origin and the danger of impulse interfere in the coal mine monitoring system. The impulse interfere in monitoring system has characters of sparse, shortness, and high amplitude. According to those characters, the adaptive denoising method based on FOCUSS was proposed to eliminate the impulse interfere. In the experiment, according to the procedures proposed by FOCUSS, the simulation was implemented based on real data which collected from coal mine practicality. Also, the wavelet algorithm was compared with the FOCUSS method in the experiment. By the wavelet method, the maximal absoluteness errors is 0.22, the minimal errors is 0.05; by the FOCUSS method, the maximal absoluteness errors is 0.13, the minimal errors is 0.0015. Also, the data form recovered by FOCUSS method is more similar to the real data from than the data form recovered by the wavelet method. The result of simulation testified the correctness of the adaptive denoising method based on FOCUSS. These instructions give you basic guidelines for preparing papers for conference proceedings.