PRODUCTION PLAN-BASED IRON AND STEEL PLANT OXYGEN LOAD PREDICTION METHOD

摘要

A production plan-based iron and steel plant oxygen load prediction method, belonging to the technical field of information, and relating to the technologies of influence factor extraction, neural network modeling, similar sequence matching and the like. The method comprises: employing industrial actual operation data, first extracting relevant data such as a production plan and actual production performance of converter steelmaking, performing influence factor analysis, and extracting a main influence variable of oxygen consumption; then establishing a neural network prediction model of the oxygen consumption of a single converter, and, using mean square error as an evaluation indicator, providing prediction results having time granularity for the converter during the blowing stage; and finally, by combining smelting time point and smelting duration information for each apparatus in a converter production plan, providing an oxygen load prediction value for a planned time period. The result obtained by the method is high in precision, can be adjusted in real time according to a change of a production plan, has guiding significance for actual production, and can also be generalized to an iron-making process to finally predict the total oxygen consumption of an iron and steel plant.