Industrial real-time data compression method with high compression ratio

TL;DR: In this article, an industrial real-time data compression method with a high compression ratio was proposed. But the method is characterized by reading in n data points and using a minimum square fitting method to fit a linear equation.

Abstract: The invention discloses an industrial real-time data compression method with a high compression ratio. The method is characterized by: reading in n data points and using a minimum square fitting method to fit a linear equation; if the n data points do not satisfy that linear equation deviation is less than a compression threshold, storing a starting point, then reading in the n data points after the starting point to fit the linear equation and restarting; if the n data points satisfies, reading in and determining whether the subsequent data point data value deviation is less than or equal to the compression threshold Delta; if the deviation is less than the Delta, abandoning, continuously reading in and determining the subsequent data points till the deviation is larger than the compression threshold Delta; if the deviation is larger than the compression threshold Delta, recording a fitted value of the previous data point, taking the previous data point as a start of reading in the data, re-reading in the n data points so as to carry out fitting determination and storing till all the industrial real-time data is compressed. In the invention, a direction of the linear equation accords with a variation trend of the subsequent data points. In a certain compression threshold scope, a data compression ratio is good and the ratio is larger than the ratio obtained by using the traditional compression methods.