Reducing the fluctuation of oxygen demand in a steel plant through optimal production scheduling

摘要

Sufficient and timely oxygen supply was essential for the steelmaking process to ensure product quality, production continuity and energy saving. The intermittent production mode and the simultaneous operation of several converters in steelmaking process will lead to large fluctuations of real-time oxygen demand. However, the contradiction between the limited guarantee capacity of oxygen supply and the oxygen demand fluctuation exists, which will affect production stability and lead to oxygen emissions. Therefore, the oxygen demand optimization problem of the steelmaking stage (ODOPSS) with multiple converters in a steel plant was concerned and an optimal production scheduling model was built for minimizing the oxygen demand fluctuation and the penalty of casts' starting time deviation. To solve the model, a hybrid genetic algorithm combining variable neighborhood search (HGAVNS) was used. Results of computational experiments indicated that HGAVNS outperformed the genetic algorithm (GA) and the variable neighborhood search algorithm (VNS). The sensitivity analysis showed that the deviation penalty decreased with increasing model coefficient at the expense of increasing the oxygen demand fluctuation. Results also indicated that the more frequent supply of hot metal and increasing percentage of dephosphorized charges would in general increase the oxygen demand fluctuation. In addition, the analysis of two industrial cases showed that the oxygen demand fluctuation could be reduced by 49.32% and 51.32% respectively at most, indicating that the proposed model could be conducive to ensuring the production stability and reducing the potential oxygen emissions, so as to realize cleaner and sustainable steel production. (C) 2020 Elsevier Ltd. All rights reserved.