Establishment and Optimization of Gas Flow Prediction Model for Annealing Furnace Based on GA-SVM

The temperature accuracy of the steel coil is directly affected the production quality of the steel plate. In the annealing furnace temperature control system, strip temperature control is mainly decided by various furnace annealing furnace gas flow in the area, so setting up a precise model between gas flow and temperature has become an important problem of the control system of annealing furnace. In this paper, SVM is applied to the prediction model of the gas flow in the annealing furnace since it's suitable for nonlinear and high dimensional problems as a machine learning method. In view of that the choice of C, G and the model parameters of the SVM directly affect the prediction precision of the model, this paper uses GA to optimize these two parameters. Finally, we use the data of field production as training and testing samples to establish and verify the model. The experimental results showed that the established model well achieved the requirements of the accuracy of the target.