Streaming computing method and system supporting industrial data

Abstract

The invention provides a streaming processing calculation method and system for industrial data. The method comprises the steps: carrying out the data collection of equipment, obtaining a bit array generated by the equipment, recognizing an equipment field, converting the equipment field into a JSON character string, and transmitting the JSON character string to message-oriented middleware in realtime; analyzing the data information of the message-oriented middleware by using an Flink computing engine, and carrying out asynchronous matching on corresponding dimension information; performing windowing calculation on the data information according to different dimensions, writing a first calculation result into an analysis layer of the message-oriented middleware, performing index calculation, and packaging and storing a second calculation result into a column-type storage database. Data calculation is achieved through a big data flow calculation engine Flink, data storage is completedthrough a big data distributed file storage system Hadoop, and data access is safe through a kerberos network authorization protocol; the problems of data timeliness calculation, automatic analysis report and data storage safety of a current industrial enterprise are solved.

Description

technical field [0001] The present invention relates to the technical field of data parsing, in particular to a method and system for supporting industrial data to implement stream computing. Background technique [0002] As the country vigorously develops "new infrastructure", it is a new trend to use big data as the technical center to provide intelligent production for traditional industrial enterprises. In traditional enterprises such as industrial enterprises, there are data that have nowhere to be used, or where it is used, but the data is scattered in different enterprise equipment or systems. Industrial enterprises have great demands on production quality and cost, and production quality can often directly affect the production cost of industrial enterprises. [0003] Industrial enterprises are still in a "primitive" state, relying on a lot of manpower. In the past, most enterprises used the experience of old masters to preset or change the production equipment for...

AI Technical Summary

Problems solved by technology

In the past, most enterprises used the experience of master craftsmen to preset or change the production equipment for the quality control of industrial products. This method not only wastes a lot of manpower and time, but also requires the actual operator to have strong experience.

And this kind of experience is often irreproducible, and personnel can only have a certain amount of experience through long-term accumulation

[0004] Patent document CN110879820A provides an industrial data processing method and device, by collecting industrial data generated by at least a part of the production process; correlating each industrial data according to the process time of each industrial data collected; performing abnormal detection on each industrial data after time correlation, Through effective data processing technology, the discrete production process is stitched together, and the problem is quickly located through the analysis of the entire production link. However, the above patent documents only focus on detecting abnormalities after time correlation of data, and do not involve stream sorting of data.

Images