Industrial internet of things data analysis method and system and related equipment

Abstract

Embodiments of the present application provide an industrial internet of things data analysis method and system, and related equipment for reducing dimensions of multivariate time series analysis andrealizing intelligent prediction of parameter indices of industrial equipment based on a multivariate time series model. The method according to an embodiment of that present application includes construct a first matrix according to a time series corresponding to at least two parameter indices, and calculating a correlation coefficient matrix corresponding to the first matrix; calculating eigenvalues of the correlation coefficient matrix and eigenvectors associated with each eigenvalue; calculating a comprehensive weight value of each parameter index according to an eigenvalue of the correlation coefficient matrix and an eigenvector associated with each eigenvalue; according to the comprehensive weight value of each parameter index, selecting the time series corresponding to the preset number of parameter indexes in front of the ranking as the target series; constructing a multivariate time series analysis model according to each of the target sequences; According to the multivariatetime series analysis model, the prediction value of the future period of the parameter index corresponding to each target series is calculated.

Description

technical field [0001] The present application relates to the technical field of the Internet of Things, and in particular to an industrial Internet of Things data analysis method, system and related equipment. Background technique [0002] Because the data of the Industrial Internet of Things has a very strong time attribute, and has the characteristics of sequential occurrence and sequential accumulation. That is to say, all the data reported by an indicator can be regarded as a function with time as an important parameter. Historical data has a certain impact on future development, and the value of the next time node is what can be achieved under a certain significance level. predicted. Considering the overall operating status of the equipment, it is one-sided to study only one or two parameters. Adding as many parameters as possible can increase the control of the equipment. [0003] However, in actual IoT devices, there are a large number of parameter indicators, and ...

AI Technical Summary

Problems solved by technology

[0003] However, in actual IoT devices, the number of parameter indicators is very large, and it is not necessary to study all parameter indicators from the perspective of calculation amount or reflection of the whole

In the existing schemes, the necessary parameter indicators are often selected for monitoring based on manual experience, which often has greater subjectivity

If the selection of parameter indicators is unreasonable, for example, too many parameter indicators will increase the amount of calculation and waste resources. If the parameter indicators are too few, the equipment operation status information cannot be included.

Images