Research On Energy Consumption Prediction Method Of Resistance Furnace Based On Data Driven

As one of the basic processes in manufacturing industry,heat treatment needs to consume a lot of energy and has a huge pollutant discharge every year.In the 14 th Five-Year Plan of China,it is clearly pointed out that the heat treatment device should be transformed into intelligent,digital and green.Adhering to green manufacturing,energy conservation and emission reduction are key tasks in the future.Industrial resistance furnace is one of the heat treatment devices,which has the advantages of accurate temperature control,high energy consumption and less pollutant discharge,but also has the disadvantages of slow temperature rise and large power consumption.The old-fashioned resistance furnace in C hina still has a huge amount of ownership,low intelligence and a lot of power waste.Through the prediction of the energy consumption of the resistance furnace,it can reflect the real-time working state of the resistance furnace,provide reliable dynamic energy consumption prediction data,and provide detailed data support for optimizing annealing process,overhaul of resistance furnace or adjusting production plan.Therefore,this paper analyzes the energy consumption data of the industrial resistance furnace processing process,establishes a data driven energy consumption prediction model,and establishes a monitoring system by studying the energy consumption characteristics of the resistance furnace.The main contents of this paper are as followsFirstly,the structure of the industrial resistance furnace is analyzed,and t he energy consumption model is made up of auxiliary system and heating system.The energy consumption of heating system is divided into two processes: heating and heat preservation.Through the analysis of the energy consumption characteristics of the resistance furnace in the working stage,the energy consumption model of the resistance furnace is built,and the model support is provided for the energy consumption monitoring system.Secondly,on the basis of collecting the current,power and temperature da ta in the process of industrial resistance furnace as input vector of prediction model,in order to improve the accuracy of the model,the off-line multi parameter energy consumption prediction model based on data-driven is established by extracting the characteristic parameters of current.The results of data driven models based on ANFIS,GPR,SVR and PSO SVR are compared to verify the effectiveness of the data driven off-line multi parameter energy consumption prediction method.Then,in view of the problem that offline model does not have online update samples and poor real-time performance,a data driven online energy consumption prediction model is proposed.The prediction accuracy,relative error and training time of the model are compared with the online SVR of incremental learning,SVR of sliding time window and dynamic NARX neural network.Finally,based on the energy consumption model of industrial resistance furnace,the energy consumption monitoring system is developed.The overall framework,function modules and database design of the energy consumption monitoring system are described.The system function interface based on QT Creator and VC++ 2008 is introduced.The effectiveness and practicability of the prediction algorithm in the system are verified by case operation.