AUTOMATIC EXTRACTION CONTROL METHOD OF REHEATING FURNACE FOR AUTOMATICALLY SETTING OPTIMUM WORK EXTRACTION PITCH ACCORDING TO HEATING CONDITIONS AND OPERATION CONDITIONS OF ROUGH ROLLING AND FINISH ROLLING

摘要

PURPOSE: To provide an automatic extraction control method of reheating furnace, the method capable of uniformly maintaining quality and output per steel types and working patterns by determining return schedule of the whole rolling line considering accelerating and decelerating patterns and return schedule, thereby accurately predicting return schedule of a succeeding extraction material. CONSTITUTION: An automatic extraction control method of reheating furnace comprises a step of checking whether extraction of a work is completed or not in the reheating furnace; a step of reading information on the relevant work and operation data including operation and finish rolling setup data from a hot rolling process facility if extraction of the work is completed as results of the checking; a step of drawing up a predicted return schedule by predicting a return schedule from extraction time point of the reheating furnace to coiling completion time point through a return scheduler based on the read operation data; a step of calculating the minimum extraction pitch between two extraction materials such that the predicted return schedule satisfies restriction conditions on a rolling line; a step of determining an extraction time of a next extraction material from the calculated minimum extraction pitch and reheating furnace restriction conditions delivered from the reheating furnace; a step of drawing up an actual result return schedule by collecting return actual results on a work of which extraction is completed in the checking step; a step of detecting an error on the predicted return schedule by comparing the actual result schedule with the predicted return schedule if drawing up of the actual result return schedule is completed; a step of applying the compensated extraction time to the reheating furnace by compensating the extraction time of the next extraction material as much as the detected error; and a step of learning the predicted return scheduler such that the detected error is compensated.