Quality Prediction Of Loose Rewetting Process In Tobacco Production Line Based On Cloud-edge Collaboration

The tobacco production line is the key link of tobacco processing technology.,and the quality of the process determines the final processing quality and sensory quality of cigarette products.The tobacco production process has the characteristics of strong production continuity,complex coupling between variables,and frequent fluctuations in raw material composition.This also determines that the key to the optimal operation of the tobacco production process is how to perceive various changes in the process and product quality in the shortest time when the internal and external conditions such as raw material characteristics and production conditions change,and then coordinate the operation and process parameters of the production process to ensure the optimal operation of the entire production process and improve the predictability and autonomy of the tobacco production system.At present,most of the research is mainly carried out from the theoretical level,focusing on the accuracy improvement of prediction and optimization algorithms,while ignoring the high timeliness requirements of the actual production line for quality monitoring,and it is difficult to ensure the real-time performance of process quality prediction and early warning in the process manufacturing process.In this paper,the loose rewetting process of tobacco production line is taken as the research object.Aiming at the problems of difficult data coordination and poor real-time prediction effect of loose rewetting process,a cloud-edge collaborative quality prediction method for loose rewetting process of tobacco production line is proposed.Firstly,by analyzing the tobacco production process of the tobacco production workshop,the cloudedge collaboration framework of the tobacco production workshop is proposed,and the application of the cloud application center and the edge aware node of the tobacco production workshop is described according to the framework.Secondly,by designing the cloud edge interaction mechanism of data and prediction model,the periodic update of cloud-edge data communication and edge perception node prediction model is completed,which breaks through the inaccuracy of prediction results caused by factors such as moisture fluctuation of tobacco incoming material and temperature and humidity change of workshop environment,and then realizes the accurate prediction of process quality of loose rewetting process.On this basis,aiming at the characteristics of complex data association and strong time series relationship in tobacco-making process,a CNN-LSTM prediction model is constructed by combining Convolutional Neural Network and Long Short-term Memory Neural Network to predict the process quality of tobaccomaking production line and provide decision-making instructions for subsequent processing.Finally,the practicability and effectiveness of the proposed method are verified by the application of cloud edge interaction and the prediction of process quality through the real data of loose rewetting of tobacco production line.The research method in this paper provides a new idea for data interaction and process quality prediction in the tobacco production workshop,which has certain guiding significance for using cloud-edge efficient collaboration mechanism to achieve low latency processing of process monitoring and quality prediction tasks,and improves the stability of process quality prediction in the tobacco production line.