Trend automatic extraction method based on data block

Abstract

The invention aims to provide a trend automatic extraction method based on a data block. The method comprises the following steps: firstly, collecting axle temperature data of a certain train as original data, and then drawing an original data curve; searching for an optimal segmentation value of the original data, extracting a data envelope center line, and searching for a local extreme point ofthe data envelope center line; and finally, dividing the data into data blocks according to the obtained local extreme points, and performing trend extraction analysis in each data block to provide atheoretical basis for further predicting axle temperature change, judging train faults and the like. The method solves the problems that in the prior art, the trend is extracted after data processingmethods such as filtering and smoothing are adopted for the collected data, the processed data result directly influences the trend result, the data smoothness degree is difficult to determine, and data features are lost during excessive smoothing.

Description

technical field [0001] The invention belongs to the technical field of data processing, and in particular relates to an automatic trend extraction method based on data blocks. Background technique [0002] With the development of big data, the field of data processing and analysis technology is also constantly developing. The trend of data represents the changing state of parameters. On the basis of existing massive data, data trend extraction and analysis can provide guidance for subsequent behaviors. and decision making. How to accurately extract the trend of the data on the basis of respecting the original data has become a major problem to be solved. The data collected directly from the industry fluctuates greatly due to the influence of noise, and it is difficult to directly extract the trend. If the collected data is first processed by filtering and smoothing and then extracting the trend, the processed data results will directly affect Trend results, and the smoothn...

AI Technical Summary

Problems solved by technology

[0003] The purpose of the present invention is to provide a method for automatically extracting trends based on data blocks, which solves the problem in the prior art that data processing methods such as filtering and smoothing are used to extract the collected data, and the processed data results will be directly Affects the trend results, and the degree of data smoothing is difficult to determine. Excessive smoothing will lead to the loss of data characteristics

Images