Research On The Energy Consumption Abnormal Diagnosis System Of Aluminum Melting Furnace Based On Energy Consumption Pattern Recognition And Classification

Melting process is one of the most energy consuming processes in aluminum profile production.As the core equipment of aluminum profile casting process,aluminum melting furnace is the key to ensure the normal production of aluminum profile.The abnormality of aluminum melting furnace in the production process will not only affect the normal production of the enterprise,reduce the qualified rate of products,increase the energy consumption of the enterprise,but also increase the emission of various pollutants and exhaust gases,and affect the efficiency of the enterprise.But at present,most of the production sites of enterprises still rely on the traditional threshold alarm and manual alarm to achieve abnormal detection.The detected abnormal information is often subjective,and the detected information often lags behind,so it is easy to miss the key abnormal information.Therefore,the design of abnormal energy consumption diagnosis system for aluminum melting furnace equipment is of great significance to ensure the overall safety and efficiency of aluminum production.Aiming at the defects of traditional anomaly detection technology,this thesis studies the characteristics of energy consumption data of aluminum melting furnace,proposes a combination of quantitative analysis and qualitative analysis,and uses the combination of real-time calculation and offline calculation in the quantitative analysis model to complete the dual reduction of calculation scale of energy consumption data of aluminum melting furnace,and improves the detection efficiency of online system.The specific research contents are as follows:1.Analyze the system structure and production process characteristics of aluminum melting furnace,and establish the basic thermodynamic model of aluminum melting furnace.Complete the specific construction of aluminum melting furnace process model.According to the characteristics of energy consumption data of aluminum melting furnace,an abnormal energy consumption analysis framework is proposed.The combination of qualitative and quantitative analysis,real-time calculation and off-line calculation are used to improve the operation speed and the overall utilization of resources of the anomaly diagnosis system.2.The firefly algorithm is selected to cluster the energy consumption data of aluminum melting furnace,and the firefly algorithm is adjusted by grouping optimization,variable step optimization and other heuristic algorithm fusion optimization.Basing on the improved FA,a clustering model is proposed,which shows faster operation and better accuracy.The energy consumption pattern of aluminum melting furnace is analyzed by clustering model based on improved firefly algorithm.3.Combined with the characteristics and needs of the energy consumption data in this thesis,the basic cart decision tree is improved and analyzed,and the decision tree model is trained by using the labeled data after clustering.By combining the trained decision tree classification model and outlier analysis algorithm,the energy consumption outliers of the system are analyzed in real time.4.Based on the improved firefly algorithm and the improved decision tree algorithm,the abnormal energy consumption diagnosis model of aluminum melting furnace is constructed,and the model is implemented based on the distributed big data storage and analysis system,and the fault diagnosis of aluminum melting furnace equipment is completed.