On this page

Research on Adaptive Regulation System and Energy Consumption Optimization Control of Steelmaking Process Parameters in the Steel Industry Based on Reinforcement Learning

By: Wen Qiu 1, Zhenlong Zhao 2, Jian Huang 1, Zhangman Liu 1, Yanlong Zhang 1, Hongyan Luo 1, Weiyi Xing 1, Baolong Li 1
1Benxi Beiyingshan Iron and Steel (Group) Co., LTD. Steel Plant, Benxi, Liaoning, 117017, China
2Nanjing Dongchuang Xintong Internet of Things Research Institute Co., LTD., Nanjing, Jiangsu, 210000, China

Abstract

The traditional steelmaking process is difficult to meet the quality and efficiency of the current users of steel products, for the problem, the steelmaking process parameter adaptive adjustment system and energy consumption optimization control strategy based on the PPO algorithm for the iron and steel industry is proposed. First of all, the steelmaking process parameters are defined, and at the same time, the main problem of this research is determined, and the problem is transformed into a Markov decision-making process, and the PPO algorithm is used to optimize the steelmaking process parameters, and ultimately, the optimal adaptive regulation scheme of the steelmaking process parameters is generated. On this basis, with the help of microcontroller and programming technology, the design task of steelmaking process parameter adaptive adjustment system was completed. It is found that the optimal control of energy consumption of the steelmaking process parameters adaptive adjustment system belongs to multi-objective problem, the maximum completion time of the product and the total energy consumption as the objective function, in addition to the corresponding constraints are given, and the PPO algorithm is used to solve the objective function and get the optimal energy consumption control strategy. Integrate the above theories, the research program of this paper to carry out empirical investigation and analysis. Under the condition of 40mm scrap input, the PPO algorithm is more effective than the traditional PID algorithm in the adaptive adjustment of steelmaking process parameters, which confirms the effectiveness of adaptive adjustment of steelmaking process parameters in the steel industry based on reinforcement learning. In addition, under the same furnace size conditions, compared with the DDPG algorithm and SAC algorithm, the algorithm in this paper is easier to achieve the optimal control scheme of energy consumption, and its maximum completion time and total energy consumption values are 2.581s and 136.615KJ.