Energy consumption simulation and diagnosis in the coking process: A mechanism and data-driven approach

With the enhancement of energy resources and environmental constraints, key energy-consuming industries, such as the steel industry, must urgently achieve an efficient transformation of energy conservation. The energy consumption is influenced by multiple factors, including operating parameters, energy utilization methods, and energy-saving technologies in steel enterprises, with its variation levels exhibiting complexity and regularity. However, the underlying mechanisms driving energy consumption fluctuations remain unclear, and there is a lack of real-time collaborative control strategies adapted to actual production scenarios. Moreover, the energy-saving potential of existing technologies and the energy-saving effects after technological retrofit remain uncertain. To address these challenges, this study develops an energy consumption simulation and diagnostic model based on both mechanism and data-driven approaches. The model simulates the fluctuation trends of energy consumption and identifies key influencing factors, thereby providing targeted operational optimization suggestions. Taking the coking process as a case study, a hierarchical diagnostic approach is applied to evaluate the energy consumption level by selecting the minimum energy consumption value for 30 days as the benchmark. The analysis identifies that the coal composition is the most effective controllable factor. Adjusting key coal parameters is expected to reduce energy consumption by 3.60 kgce/t. Furthermore, deploying new energy-saving technologies can further decrease energy consumption by 4.27 kgce/t. In summary, this research provides a technical framework and practical guidance for the benchmarking of energy efficiency of steel enterprises.