Research On Key Technology Of Industrial Data Analysis Platform Based On Operator Configuration

With the rapid development of Internet technology,the field of data analysis has gained more and more attention,and data analysis technology has made considerable progress.There are numerous scenarios in the industrial field that require the use of data analysis techniques.In 2015,the State Council put forward the "Made in China 2025" plan,and the data analysis technology in the industrial field received the national attention and support.Compared with the flourishing of data analysis technology in the Internet industry,there is still room for improvement in the progress of the industrial field.The analysis of quality problems in the industrial field and the adjustment of processes and other processes are often handled in a way that uses reference experience.With the improvement of automation of industrial equipment,the degree of data collection of industrial equipment is getting higher and higher.Using Internet software technology and machine learning methods,the efficiency and accuracy of industrial data analysis can be improved to some extent.A streamlined software platform can automate tedious data reads,data preprocessing,data analysis,and data storage.Machine learning algorithms can make the data analysis process focus on the data itself,rather than the experience of the business staff.The data analysis platform in the industrial field has high requirements for the ease of data analysis and the accuracy of data analysis results.This project implements an industrial data analysis platform based on the operator configuration model.The platform realizes the unification of industrial data analysis processes,including data source management modules,operator management modules,task management modules,and matching human-computer interaction pages.In the implementation of the platform,the Spring Boot back-end Web framework was adopted,and the React front-end framework was used as the infrastructure for software project R&D.The data source management module helps the platform realize the abstraction of data sources such as MySQL and CSV and adapt the platform to more data sources.The data analysis work of different business scenarios requires different combinations of algorithms to complete.The operator management module realizes a combinable operator concept through the abstraction of the algorithm elements,and the operators can adapt to various types through different combinations.Application scenarios.This topic implements the hot load function of the operator through the class loading function of the Java programming language.The operator of the Jar package can be loaded into the platform without restarting the server.The Dataframe of Spark's big data processing framework was selected as the operator's versatility and algorithm basis.Based on the RocketMQ message queue,the platform synthesizes information such as data sources and operator lists into executable tasks,and task consumers complete task execution.The bolt data collected by automatic wrench tightening was used as a case of quality analysis to verify the practicality of the platform.