Industrial scene working condition clustering method, system and device based on Markov random field model and storage medium

Abstract

The invention discloses an industrial scene working condition clustering method, system and device based on a Markov random field model and a storage medium, and belongs to the technical field of industrial data analysis. The method comprises the following steps: step S100, preliminarily screening out characteristics which can influence the change of the operating condition of the power plant, and performing preliminary treatment on the characteristics; step S200, performing noise reduction on the primarily processed characteristics affecting the operation condition change of the power plant; step S300, carrying out dimension reduction on the feature data influencing the change of the operating condition of the power plant after the noise reduction processing; and step S400, clustering the data after dimension reduction processing to obtain an operation condition. The system comprises a preliminary screening module, a noise suppression module, a feature dimension reduction module and a working condition clustering module. The efficiency of processing and analyzing the unit mass data is further improved, and high efficiency and economical efficiency which cannot be achieved when the unit mass data is manually processed are achieved.

Description

technical field [0001] The invention relates to a construction method, system, equipment and storage medium of working conditions in an industrial scene, and belongs to the technical field of industrial data analysis. Background technique [0002] In the field of industrial big data analysis, actual industrial signals often come from many production system sensors, and these data change over time. Working conditions are defined as several possible states that divide the high-dimensional time series signal. For example, the operating state of a thermal power plant can be roughly divided into start-up period, operation period, stop period, abnormal period and so on. Working condition mining and identification uses machine learning algorithms to mine hidden working condition information in high-dimensional data, which has certain guiding significance for actual production, and is also an important pre-processing method for many subsequent algorithms (prediction, classification...

AI Technical Summary

Problems solved by technology

A subsequence that is too long may contain more than one hidden working condition, while a subsequence that is too short may contain fewer data features, which will have an adverse effect on the results of the clustering algorithm, and artificially segmenting the sequence is not conducive to retaining Time Consistency of the Original Time Series

The hierarchical clustering method based on data distribution intuitively determines the segmentation points for clustering according to the distribution characteristics of data in certain dimensions. It is a category feature with a very concentrated distribution. Unstable or volatile features cannot be used in this type of method, and the use is relatively limited.

Images