Research On Key Technology Of Press State Prediction Based On Digtal Twin

With the continuous development of intelligent factory construction,new requirements are put forward for the real-time monitoring and performance prediction of production equipment.As a kind of complex industrial equipment,the research on state prediction of printing press in its production process is still blank,which restricts the improvement of intelligence level.Therefore,it is of great practical significance to establish the digital twin model based on the data acquisition in the production process of the printing press,to predict the key state of the printing press,and to realize the intellectualization of the working process of the printing press.The main contents of this paper are as follows:(1)The digital twin model of press state prediction was constructed.Combined with the structural characteristics of printing equipment and the theoretical research foundation,the general frame of printing machine state prediction is established by using improved polychromatic set theory and dynamic Bayesian network.(2)The method of data acquisition and processing of press status is proposed.The main classification of multi-source data of printing press is studied,and the method of data processing by sliding window technology is proposed,which provides technical support for equipment condition monitoring.(3)Establish a printing press state prediction method based on dynamic Bayesian network.A hierarchical framework of press state was proposed,a dynamic Bayesian network model for press state prediction was built,and the prior probability of state data was analyzed by kernel density estimation function.(4)Based on the measured data of the MK Cerutti R983 gravure printing press and taking the registration state as the research object,the state prediction of the gravure printing press was studied,the influencing factors of the prediction accuracy were analyzed,and the feasibility of the press state prediction model and scheme proposed in this paper was verified.This paper is based on the trend of intelligent printing equipment,through theoretical modeling,data processing,predictive analysis,test verification,combined with the specific cases of production equipment of well-known equipment manufacturing enterprises.The key technology of state prediction of printing press based on digital twin is explored,which provides a theory and method worthy of reference for improving the intelligent level of printing press and satisfying the intelligent production.