A Noise Removal Method in Measurement of Particle Size Distribution by the Shifrin-Transform

Abstract

In actual measurement of particle size distribution by the Shifrin-transform, many noises are easily misread as distribution peaks, and bring serious difficulties to the measurement. The purpose of this paper is to find a method of removing these noises to improve existing technology. By analyzing the source of these noises, we found these noises are mainly caused by the serious large angle data loss, and then propose a compensation function in the Shifrin-transform to remove these noises. Simulations explain that the method has a good effect. The method is actually tested for the sample particles of the unimodal and bimodal distribution by experiments. The result shows that the noises disappear but the number and location of original distribution peaks aren’t affected after using the compensate function. So this method can remove effectively noises and restore accurately original distribution in the measurement.