Design And Implementation Of A Platform For Knowledge Management And Algorithms Openness For Industrial Intelligence

As the country vigorously promotes the integration of a new generation of information technology and manufacturing,the industrial Internet is developing rapidly.As a key technology of industrial intelligence,predictive maintenance is a typical application scenario of big data and AI-enabled manufacturing.The market is broad and is a field.At present,there are very few relevant talents who master predictive maintenance technology,and the talent gap is huge.The first reason is that the industry has relatively high barriers to entry,and the second reason is that there are fewer teaching platforms in the field of industrial intelligence on the market,and there are very few suitable for newcomers to get started,and there is currently a lack of corresponding training systems in universities and courses.In view of this situation,this paper proposes the design and implementation of a platform for knowledge management and publicizing algorithms for industrial intelligence,which provides a one-stop learning platform for knowledge learning,data management,algorithm experience and openness for beginners in the field of industrial intelligence,and provide an industrial APP Of developers provide a new end-to-end development model.Starting from the basic process of software development,this paper gives chapters such as background investigation,research status,demand analysis,related technology research,platform design and implementation,system testing and verification,and gradually introduces the complete development process of the platform.In terms of functional design,corresponding to the topic of the thesis,firstly,two large functional modules are designed for knowledge management:a tutorial and a project module,and a data set module.The tutorial and project modules are used to showcase related technologies,algorithms,cases,and Projects and other tutorials provide a platform for beginners to acquire knowledge.The data set module provides data set download,online graphical display,and data set upload functions to provide users with data management functions;then,an algorithm library module and an API module for algorithm opening is designed,and based on these two modules,a new end-to-end development model is proposed.Industrial APP developers upload the algorithm in the algorithm library module after debugging the algorithm locally and automatically generate the API,and then calls the corresponding algorithm to obtain the processing result based on the API document,so that based on the platform,it provides similar cloud functions,and publishes the developer's algorithm online.Combined with the WeChat applet,the developer can develop his own industry APP without building a back-end server.In terms of function implementation,the platform is developed based on Java language,and the framework adopts SSM(Spring+SpringMVC+Mybatis).In terms of database,MySQL is selected as the main database for storing data on the platform,and Redis is used to store file storage paths,training results and other information,and used as a cache.The algorithm calculation is based on the Weka in the Java environment and the scikit-learn library and the tensorflow library in the Python environment.The platform uses these algorithm libraries to build a set of algorithm engines to support all the algorithm calculation functions of the platform algorithm library module and API module.The RabbitMq message queue processes complex computing requests asynchronously to improve user experience.