A collaborative optimization strategy for energy reduction in ironmaking digital twin

Abstract

The expenses of pig iron manufacturing stay in a relatively high level due to the large consumption of coke and ore. There are a variety of methods to optimize the ironmaking process, including multi-indexes method, production-oriented approach, and cost control strategy. In order to reduce the production cost, we optimize the raw materials supply schedule of an ironmaking plant with five sintering machines and seven blast furnaces. The ironmaking plant used to fix the feeding system, leading to an inefficient operation and mismatched connection between these two procedures. At first, the physical characteristics of each blast furnace is captured by linear regression model. Then, a self-adaptive genetic algorithm with variable population size is constructed under distribution system. Finally, the improved genetic algorithm is applied to optimize the total production cost represented by the aggregation of seven coke ratios on a cloud computing platform. With the application of this cloud collaborative optimization framework, the mean coke ratio of the ironmaking factory has decreased 13.96kg/t Fe.