Research And Application Of Data Compression Technology In Industrial Process Monitoring

The data in complex industrial process is useful for system modeling, control and optimization. The storage of industrial data has a great significant for the parameter identification, system modeling, failure detection and controller design. So it brings a great challenge to save large quantities of process data for the complex industrial process monitoring system. Industrial data contains a large number of redundant and relevant information, which makes using data compression to improve data storage rate possible.This paper analyzes the features of various methods based on summarizing the common industrial data compression methods. For the large amount and the noise of industry data, wavelet transform which is based on multi-resolution analysis principle and the Mallat fast decomposition algorithm can be applied in industrial data compression. In this paper, the basic principles of the wavelet transform and wavelet-based data compression method are described, and the performance of common mother wavelet is analyzed.The compression method based on wavelet transform breaks down the industrial data into low frequency sub-band and lots of high frequency sub-bands, applies adaptive threshold strategy to process the high-frequency coefficients which contains the vast majority of data information of the quantification threshold, and stores the low frequency coefficient, some of the high-frequency coefficients and their position coefficient to achieve the purpose of data compression. So it satisfies the performance requirements of industrial data compression. Meanwhile, the paper takes advantage of the variable characteristics of mother wavelet, achieves the function of selecting different compression parameters according to the industrial signal characteristics and enables the ability to handle all types of industrial signals. In the process of volume data compression based on wavelet-transform, the choosing of mother-wavelet, threshold, decomposition and storage solutions are completed.Industrial data compression method based on wavelet transform, not only has good compression performance and noise reduction capability, but also has the feature of selecting compression parameters according to the characteristics of the signal which makes the data compression method flexible. The test results demonstrate the effectiveness of this method through the application in refining furnace monitoring system.