Hybrid intelligent pre-processing system of parameters in heating furnace control

Heating furnace has played a very important role in the iron and steel enterprises. Due to the complex environmental interference, it is difficult to measure parameters, and this has influenced the accuracy and stability of the control system deeply, at the same time, there may be some potential security risks. In this paper, a hybrid intelligent pre-processing system of process parameters is proposed for furnace control. The system consists of several steps: the prediction of some parameters by self-adaption or fuzzy neural network, the parameter filtering, the parameter estimation in the way of production rule, the abnormal parameters modification with case-based reasoning, etc. This system could make the furnace control become accuracy and stability, and eliminate potential security risks. This proposed intelligent pre-processing system has been successfully applied by a steel enterprise in Gansu Province, China, and achieved good results. The system has a good prospect of application and extension in other similar industrial processes.