Research On Optimization And Scheduling Algorithm Of Discrete Manufacturing Workshop Based On Digital Twin

In the time of rapid development of digital economy,people’s production and consumption methods have changed greatly,from the adaptability of products to the development of personalized needs,so the traditional assembly line of a single production model will soon not adapt to the actual production needs,but the discrete production method of the workshop has the characteristics of flexible production,through which the workshop production can be transformed to digital and intelligent.The product lifecycle management concept of digital twin provides a reliable direction to realize the change of manufacturing production method,and the 3D visualization,production simulation and intelligent optimization of workshop can be realized through digital modeling,simulation and artificial intelligence algorithms.Therefore,based on the digital twin technology chain,this paper makes the following research for discrete production workshops.First,this paper completes the mathematical modeling of the discrete shop production scheduling problem for the Job-shop Scheduling Problem(JSP),determines the objective function,and uses multi-level objective optimization to solve the production solution.Next,based on Digital Twins(DT)technology,the digital twin-based discrete shop data fusion scheme is proposed to build the information flow architecture of the twin shop,and the Multi-Information Warehouse(MIW)of the shop is constructed with real-time data,twin data and service system data,so as to build Digital Twin Workshop(DTW).After that,we proposed a Product Lifecycle Management(PLM)information fusion scheme to realize realtime interoperability of workshop information,and the constructed Digital Twin Workshop has good simulation accuracy through comparison tests.Next,based on the similarity characteristics of twin shop data and real-time shop data,a dynamic scheduling mechanism based on twin network is proposed.Because the twin neural network(Siamese Network,SN)has the similarity recognition feature,the interference factor is output by calculating the high-dimensional feature values of both through the twin network,and the dynamic scheduling mechanism of the workshop is designed by judging the threshold value of the interference factor,and the scheduling model is continuously iteratively optimized by updating the production plan in time,so that the workshop production tends to be optimal.Finally,based on the above results,the construction of a flexible scheduling workshop system is completed to realize the presentation of the results of management on the production line.In the system development of this digital workshop,functions will be divided according to the current actual production tasks of the workshop,and each functional module can complete a certain independent sub-task,or realize the collaborative operation from the virtual workshop to the physical workshop through manual intervention,or arrange a better production route through the data mining module of the virtual workshop.