The Study on Paper Board Thickness Measurement by Using Data Fusion.

This paper presents a new method of using a multi-output data fusion from a laser displacement sensor (LDS) to reduce vibration error for online noncontact paper board thickness measurement. A data-in-data-out fusion model for the thickness measuring was proposed. And a revised clustering method for fusing the canner gap data was presented. An EMD decomposition algorithm was used to analyze the high frequency noise. In order to find a desired filtering algorithm, five different filter designs were compared. After the comparison, sliding filter and Kalman filter were selected for noise removal to the original signal data from the LDS. In the fourth section, the revised clustering method was compared to an average algorithm for calculating the offset curve. Based on the proposed fusion model and revised clustering method, the fusion steps for online thickness measurement have been defined. Experimental results show the effectiveness of this method, and the thickness online measuring relative average error was less than ±4% at the range of 150μm-1500μm. [ABSTRACT FROM AUTHOR]