A Two-Stage Adjustment Method For The Balance Of Oxygen System In Steel Industry

Steel industry, which plays a vital role in the economic development and social construction of our country, has become one of pillar industries in our country. As an important part of iron and steel enterprise, the balance of oxygen system can not only ensure the safety in production of the iron and steel enterprises, but also can reduce energy consumption and the amount of oxygen diffusion. Therefore, it is important for the iron and steel enterprises to save production cost, promote the economic benefits and environmental protection. Most iron and steel enterprises are based on the artificial experience of the field staff to ensure the balance of oxygen system. However, the structure of the oxygen system is very complex, which is involved many users, and it is not ideal to balance the oxygen system with artificial experience alone.As to the imbalances in supply and demand for the oxygen system in steel and iron enterprise, a two-stage real time adjustment method is proposed to determine the adjustment programme online for the balance of the oxygen system in this paper. For the first stage, to categorize the data by the oxygen system conditions:a classifier based on Gaussian process is designed to determine whether the present time is an adjustment time point or not, and then the corresponding operating conditions are analyzed based on the statistical method. The model in different conditions is built in the second stage based on T-S model with parameters optimization:Firstly, fuzzy c-means clustering algorithm is used to divide the input/output space and extract the antecedent membership function parameters of the T-S model, then the improved bacterial foraging optimization algorithm is presented to optimize the parameters of the antecedent membership functions. First of all, to determine whether or not the present time is the adjustment moment, if it is, then analysis its corresponding conditions. Finally, according to the model in different conditions to determine the adjustment value of the adjustment user.To verify the effectiveness and superiority of the proposed method, the field data coming from the oxygen system of a steel enterprise in our country are chosen to conduct the validation experiments. Firstly, a classifier based on Gaussian Process is designed to separate all the adjustment point. At the same time, compared with other classifiers, the Gauss process classifier has the advantage of the oxygen system data. Then according to the model in different conditions to determine the adjustment value of the adjustment user. Compared the improved bacterial foraging algorithm with genetic algorithm and particle swarm optimization algorithm, the results show that the proposed method in this paper is more accurate than other methods.