A de-noising algorithm for PIND signals based on kalman filter

The Particle Impact Noise Detection (PIND) is a screening test that is required before aerospace relays delivery. The weak remainder signal is easily submerged by several kinds of noise, and hard to extract or identify. In this paper, the sources and characteristics of the noise are analyzed. A self-adaption de-noising algorithm is presented for ambient noise based on Kalman filter. The signal model and its parameters are received using Linear Prediction Coefficients and iterative algorithm. An attenuation coefficient is defined to distinguish if the Kalman input is remainder signal or the noise. The experiment demonstrates that the de-noising algorithm is effective to improve the ratio of signal to noise.