Complementary Detection About Workpiece Temperature Of Reheating Furnace Based On BP Neural Network

Steel rolling reheating furnace is one of the most important equipment in the production line, and its main effect is put inside the furnace temperature required by certain billet heating to follow-up rolling process which requires a certain range, so as to ensure the normal slab rolling. Therefore, the billet temperature distribution in the furnace of steel mouth out especially in surface and the center for realizing the furnace temperature closed loop optimal control and predict billet rolling effect has the extremely vital significance, and it is main quality indicators in the running of the billet reheating furnace heating. However, in many of the industrial production of heating furnace, the heating condition and billet heating quality judgment rely mainly on the thermocouples measuring points which are distributing in the furnace up and down. To solve the billet reheating furnace heating quality can not directly within the detection problem, usually we need establish reasonable mathematical model of heating furnace billet temperature to on-line estimate and predict billet temperature distribution in the furnace. But most of the billet heating model, on dealing with of the key boundary conditions is based on the furnace temperature which is measured by thermocouple, through empirical formula getting the billet model boundary conditions. Apparently empirical formula of boundary conditions has a certain error compared with the practical boundary conditions, so that it will bring on the complex process in the solution of equation.In order to solve the difficulty of billet temperature detection, this paper using CCD Thermal Imager temperature measurement technology, directly measure the billet temperature distribution, use the billet surface temperature to establish the billet temperature prediction model, and research the billet surface temperature and billet temperature distribution field.This article mainly as follows:1 The radiation temperature measurement method and its existing problems of the system and technology, then point out the significance and methods of using CCD Thermal Imager to measure temperature.2 Analyzed the billet surface temperature prediction model. Article selects the BP neural network to establish the temperature prediction Model and to analyze the simulation model.3 Analyzed the billet temperature model. Use finite difference method to calculate the internal billet temperature distribution field, and to analyze simulated condition of the temperature field. 4. Analyzed the influence on billet temperature by the thermal physical parameter of the billet itself. And put forward increasing the calculation precision though decreasing influence by the thermal physical parameters of billet.