Research On Optimal Scheduling Of Byproduct Gas System In Iron And Steel Industry

As one of the fundamental pillar industries in China, the iron and steel industry has developed rapidly in recent years. Meanwhile, it is an energy-intensive industry, whose energy cost accounts for 15% of the total energy cost of China. Therefore, efficient use of energy is crucial for reducing total operation cost. Byproduct gases, which are produced during the production of iron and steel without additional cost, are important energy sources to meet energy needs of the process of making iron and steel. Thus, optimization of byproduct gases will play a great role in energy saving. However,few research works have been reported on this field. As a result, it is extremely valuable to give a comprehensive research on this topic.Based on this background, this dissertation studies on byproduct gas system in the iron and steel industry, and makes a deeply research on optimization scheduling of byproduct gases. The optimal model is applied in an iron and steel plant in China, and an optimal result in terms of total cost reduction is obtained. The main contributions in this dissertation are as follows:(1) Based on a deeply and comprehensive analysis of byproduct gas system in the iron and steel industry, a"three systems, two aspects"theory is proposed. Three systems"means the whole byproduct gas system is delimited into three relative systems: produce-and-consume system, store system and transformation system. In produce-and-consume system, byproduct gas consuming users are divided into two kinds, the first kind of which is single-consuming users, and the second kind is mixed-consuming users. Byproduct gas consuming users belonging to the first kind of produce-and-consume system, store system and transformation system compose the objective, which is to be optimized."Two aspects"means both explicit cost and inexplicit cost should be considered when an optimal model is established.(2) According to different affecting factors of three kinds of byproduct gases, different ARMA models are established separately to predict the supply of byproduct gases. The results obtaining from the forecasting models are used as the input of the system to be optimized. (3) The consumption forecasting models of byproduct gases are also established. Consuming users of byproduct gases are divided into four kinds according to their different consuming characters. Time series prediction model, Levenberg-Marquardt BP (LM-BP) neural network prediction model, linear regression prediction model, and exponential smoothing prediction model are used to predict the demands of four kinds of byproduct gas consuming users. The results obtained from the forecasting models are used as the output of the system to be optimized.(4) An Mixed Integer Linear Programming (MILP) model of byproduct gases in the iron and steel industry is established. The objective function is to reduce total operation cost in the entire energy system for multi-period operation through optimizing byproduct gas distribution. The total operation cost includes penalty cost for emission or shortage of byproduct gases, penalty cost for deviation from normal gasholder levels, penalty cost for burners switching and fuel load changing in the boilers, oil purchasing cost. Besides, certain relative constraints are given and an optimal model is presented. The model is applied in K iron and steel plant and 30% of the total operation cost is reduced.(5) The"environmental cost"is considered in the optimization of byproduct gas system in the iron and steel industry, and a green scheduling model including environmental cost of byproduct gas system is proposed for the first time. The penalty costs for the emission of byproduct gases, and waste gases produced from burned byproduct gases and purchased oil, are all calculated in the objective function to make the model more reasonable. Compared with the previous model proposed in Chapter 6, the green optimal model could save 1.3% total cost.Finally, a summary of the research referred above is concluded and some interesting and challenging topics, which deserve further investigation in the future, are discussed.