The Study Of Predicting Improtant Parameters In Glass Production Process

Glass Industry consumes extensive energy, and cause serious air pollution. With energy prices rising and glass stringent production environmental policies introduced, the glass production cost keeps increasing. Moreover, glass corporations are struggling to survive because of overcapacity. Thus, the technology upgrades and the thermal efficiency improving are strongly demanded by glass corporations. By now the studies of furnace energy saving mainly aiming at waste heat power generation and Oxy-Fuel combustion, etc.. However the spread of these related technologies still takes time. Exploit potentialities from the existing production process is a fast and feasible method.The glass melting furnace is the key equipment in the production process; its working status will directly affects the corporate profitability. In condition that the glass production process have characteristics like multiple time delays, strong coupling between parameters and the characteristics of random perturbations, the research in this article are mainly about investigating mutual influences among various factors in the production process, explore intelligent methods which meet the technical characteristics of glass production process and predict important parameters with the explored methods. Main contents are:1. Investigate the glass production process in detail, analyze and summarize the impact caused by each aspect of the production process. Identify the key parameters which can help to improve production efficiency, and explore prediction methods that fit the characteristics of glass production.2. Propose a bottom glass temperature prediction method. Analyze the main influencing factors and select the input parameters of the prediction model. Prepare data appropriately to express the characteristics of the data. After the data is organized, predict the temperature of the bottom molten glass with the model of low calculation complexity.3. Analyze the factors affecting the quality of glass products; complete the pre-processing of the existing data about production quality. Predict the number of bubbles in products with the proposed model which meets the technical conditions, and provide the predicting outcomes as reference for helping the enterprise to improve energy efficiency and management level.