Demand Peak Forecasting of the Fused Magnesia Furnace Group with Model Prediction and Adaptive Deep Learning

During the fused magnesia production process (FMPP), there is a demand peak phenomenon that the demand rises first and then falls. Once the demand exceeds its limit value, the power will be cut off. To avoid mistaken power off caused by demand peak, demand peak needs to be forecast, so multistep demand forecasting is required. In this article, we develop a dynamic model of demand based on the closed-loop control system of smelting current in the FMPP. Using the model prediction method, we develop a multistep demand forecasting model consisting of a linear model and an unknown nonlinear dynamic system. Combining system identification with adaptive deep learning, an intelligent forecasting method for furnace group demand peak based on end-edge-cloud collaboration is proposed. It is verified that the proposed forecasting method can accurately forecast demand peak by utilizing industrial big data and end-edge-cloud collaboration technology.