Research And Application Of Key Technologies In Digital Twin Workshop For Intelligent Terminal Structural Parts Processing

In recent years,China has proposed the "Made in China 2025" plan,to promote the exploration of new modes of intelligent manufacturing and help the upgrading of manufacturing.Intelligent plant is a basic application of intelligent manufacturing,its “smart” feature are mainly reflected in scientific planning and continuous optimization.This paper takes a smart terminal structural parts processing workshop as the object,and carries out research and application of key technologies in digital twin workshop,to provide decision support for construction planning and production execution.First of all,based on the practical experience of many virtual workshop constructions,an overall framework of the digital twin workshop is given for the intelligent terminal structural parts processing workshop.The digital twin model is constructed to simulate the operation process of production lines,AGVS,and automatic tridimensional warehouse in virtual environment.Via comprehensive quantitative analysis of key performance indicators such as production capacity and resource utilization based on simulation,realizing the production system planning verification and optimization.Secondly,the digital twin model is further applied to the production plan executable analysis.This paper studies and builds a plan risk assessment model under the uncertainty system,and presents a hybrid intelligent algorithm combined with stochastic simulation,neural network and genetic algorithm,which is used to realize the production plan risk assessment and optimization of the parallel pipeline.Then,the real-time data-driven simulation operation mechanism is studied.By integrating the information system such as Supervisory Control And Data Acquisition System and Manufacturing Execution System,the digital twin model is used as a carrier to build an online simulation model,which can realize the real-time state display and transparent control of the workshop.Finally,the intelligent logistics scheduling model is constructed,which merges the task assignment and global path planning problems.Then,a combined rule scheduling method based on simulation and Support Vector Machine is designed for simple logistics network,a dynamic priority method based on neuroevolution is proposed for complex logistics networks.These two method makes the model flexible and adaptable to production logistics with different scales and states,and relevant examples are given to verify the effectiveness.