Modeling Of Reheating Furnace And Its Online Calibration Based On Process Monitoring Feedback

Reheating furnace is an important equipment in the production line of steel rolling. Its main purpose is to heat slabs to a certain temperature which is determined by the requirement of steel rolling. The heating quality of the slab directly influence the steel quality, steel production, energy consumption and rolling reheating furnace. Thus, it is essential and beneficial to research on modeling of reheating furnace to improve heating efficiency and reduce energy consumption.This thesis does the research on regenerative reheating furnace system. Firstly, establishing a three-dimensional heat conduction differential equation by analyzing the heat transfer mechanism of the slab. Then, discussing the initial conditions and boundary conditions of the differential equation. Secondly, combining with the engineering practice and the total heat exchange factors, a two-dimensional temperature prediction model of the slab is established. Because of the different work conditions of the four heating zones of the reheating furnace, the total heat exchange factors must be different. According to the fours zones of the furnace, the total heat exchange factors are divided into four parts, then a two-dimensional temperature prediction model based on four total heat exchange factor is proposed in this thesis.The total heat exchange factor is vital to improve the precision of the model. Several parameters identification methods are discussed at first, then based on the production practice, particle swarm optimization algorithm is applied to this thesis to obtain the optimal total heat exchange factor. The algorithm provides the initial total heat exchange factors for the model on-line operation.The total heat exchange factor is changed with work conditions, so an on-line parameter self-calibration part is necessary in an exact mathematic model. In this thesis, fuzzy inference system and process monitoring is combined to calibrate the total heat exchange factor. Fuzzy inference system provides the total calibration value, and the process monitoring provides the proportion of the value. With the on-line parameter self-calibration part, the model has better accuracy and adaptability than before.Simulation results with actual data indicate that the model established in this paper can calculate the slab temperature quite accurately. The self-calibration strategy based on fuzzy inference system and process monitoring is effective to reduce the error between the calculated temperature and the real temperature.