Research On Prediction And Optimization Of Leaf Conditioning Process Parameters

Leaf moistening is one of the most important processes in the threshing and redrying process.The quality of leaf moisturizing directly affects the effect of the subsequent threshing and air separation process,and indirectly affects the burning performance and sensory evaluation quality of cigarettes.At present,the threshing and redrying plant is gradually developing in the direction of intelligence,refinement and automation.While paying attention to process cost and process quality,it pays more attention to how to use production equipment efficiently and reasonably.Because the leaf conditioning process involves changes in the physical structure and chemical composition of the tobacco leaves,there is a problem that the tobacco leaves are difficult to observe in the hot air conditioning machine.The traditional method of adjusting process parameters is to constantly adjust the parameters to achieve a good leaf moisturizing effect through experiments.This method not only wastes a lot of working time of the tobacco factory,but also needs to readjust the parameters in response to the tobacco leaves of different origins.Therefore,how to establish a reasonable prediction model based on process data to adjust the parameters of tobacco leaves from different origins is the key to improving the quality of the leaf conditioning process,improving the technology of leaf conditioning equipment,and increasing the production efficiency of the redrying plant.This article takes the characteristic process of leaf moisturization as the research object.Starting from the physical structure and chemical composition of the tobacco sheet itself,the analysis shows that temperature and moisture are the two factors that have the greatest impact on it.The selected process parameters to be studied are: front steam nozzle pressure,front-end water flow rate,hot air temperature,return air temperature,feed blade temperature,and feed blade moisture.The indicators for prediction or optimization are: outlet leaf temperature and outlet leaf moisture.Three methods of multiple linear regression,BP neural network,and cyclic neural network are used to establish a prediction model for the effect of tobacco leaves after moisturization.The adjustable parameters are studied and analyzed.Mean square error,root mean square error,and average absolute error are used.Three indicators to describe the gap between the true value and the predicted value.According to the cyclic neural network model with the best prediction effect,reasonable process parameter levels were set,and the process plan was expanded to 10,000.An improved multi-objective optimization algorithm was proposed to optimize the process plan.Taking into account the two major factors of production quality and equipment service life,a strategy of cross-using different programs according to different time periods was proposed.The analytic hierarchy process was used to establish a comprehensive evaluation model,and it was determined that the moisture of the feeding leaves was the most important factor affecting the quality of the tobacco leaves after moistening,which provided a theoretical basis for a reasonable selection plan.The research in this paper reveals the specific relationship between the temperature and moisture changes of the tobacco slices and the process parameters in the secondary conditioning process,and raises the problems of uneven quality of the exported tobacco slices and serious equipment loss encountered in the conditioning process.A reasonable solution.It provides an important theoretical basis for improving the parameters and optimizing the process of the threshing and redrying plant,which is of practical significance for promoting the sustainable development of the redrying plant.