Measurement Data Classification in Information and Measuring Systems

To develop effective data compression devices for information-measuring systems, with adaptive data compression, it is proposed to modify the well-known functional classification of telemechanical data using an estimate of the data stationarity degree. The proposed approach facilitates the adaptation of the compression algorithm to the properties of the data, but remains insufficiently flexible for practical use. As a solution for this problem, a new classification of data based on the analysis of statistical properties of difference data series is proposed, which makes it possible to more accurately estimate the prospects of their compression. As the basis of the classification, it is proposed to use the range of difference values, which include 95 percent of all values of the data series under study. Also, a more exact classification is proposed to be based on the analysis of the second-order difference series. The application of the proposed approach to the classification problem of telemechanical data received from several power stations and substations showed its good practical applicability.