The Method For Condition Recognition Of Fused Magnesium Furnace Based On Probabilistic Rough Set

Fused magnesia with features such as a high melting point,good anti-oxidation,and strong insulating properties,are mainly used in the production of refractory materials,such as aerospace industry,cement and chemical industry.Fused magnesium furnace is the main equipment for the production of fused magnesia.Product quality and production efficiency of fused magnesia are determined by its operational safety and efficiency.At present,for abnormal condition identification of fused magnesium furnace,operators make decisions by observing the fluctuation law of three-phase electrode currents and the conditions of smelting equipment.However,operators can not identify the abnormal conditions in time and accurately,which will lead to the occurrence of abnormal conditions.Therefore,accurately identifying conditions of fused magnesium furnace is the key to ensuring safety production,which has important practical significance.Fused magnesium furnace is the research background of this paper.The method for working condition identification of fused magnesium furnace based on multi-heterogeneous information is proposed through in-depth analysis of the common abnormal conditions and related variables in the smelting process of fused magnesium furnace.By combining extracted feature values from fused magnesium furnace images with operational variables and expert knowledge,the recognition model for working condition of fused magnesium furnace based on probabilistic rough set is established.And the accurate identification of fused magnesia furnace conditions is realized.The main research contents of this paper are as follows.1.In order to dig out the information between the similar gray level and multi-scale characteristics of fused magnesium furnace images,a texture feature extraction method based on gray level co-occurrence matrix(GLCM)without background factors according to the working scene of fused magnesium furnace is proposed in this paper.The simulation results show that the texture feature extraction algorithm is effective.2.To solve the problem for classification errors and too many invalid points due to the distance threshold R,a clustering algorithm based on improved clustering using references and density(CURD)added inner circle representative points is proposed in this paper.The simulation results show that the proposed method has strong classification ability,which improves the classification accuracy compared with the original method,and has higher accuracy and stability than K-means algorithm.In order to make full use of the relationship among quantitative information,the proposed method is applied to the discretization of continuous attributes.3.Through the research on the identification method of fused magnesium furnace condition,the model for working condition recognition of fused magnesium furnace based on probabilistic rough set is established.The extracted clustering attribute and the expert knowledge is used as condition attributes and the model was trained and tested with data of fused magnesium furnace simulation platform.The results show that the proposed method for condition identification of fused magnesium furnace in this paper has high recognition rate and has high application value.