About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Energy Technologies and CO2 Management: Resource Efficient Processes
|
| Presentation Title |
A Review on Trends and Challenges of Electricity Demand Forecasting and Scheduling in Steel Enterprises |
| Author(s) |
Minghui Chi, Yingqin Wang, Zhongheng Chen, Xiancong Zhao, Yafei Wu, Wei Li, Hao Bai |
| On-Site Speaker (Planned) |
Yingqin Wang |
| Abstract Scope |
The iron and steel production process involves complex procedures, with energy consumption accounting for 20%-40% of total production costs. As the primary and auxiliary energy source of enterprises, electricity constitutes approximately 10% of total energy consumption and runs through the entire production process. Iron and steel enterprises have diverse power supply and consumption units, and their electricity demand is affected by multiple factors on both the supply and demand sides. Imbalances between supply and demand may trigger power surges, equipment damage, and increased costs. However, research on electricity forecasting and dispatching largely focuses on single models: mechanism-based models struggle to adapt to fluctuations in operating conditions, while data-driven models lack mechanistic support and are prone to distortion. Additionally, existing literature on the "forecasting-dispatching" synergy logic is fragmented, failing to form a systematic understanding. Such practical challenges in industrial operations and gaps in academic synthesis underscore the necessity of in-depth exploration of this field. To this end, this paper focuses on the electricity forecasting and dispatching system of iron and steel enterprises, reviews its development status, sorts out the logical frameworks of mainstream models, and analyzes the core bottlenecks of existing methods in adapting to complex operating conditions and handling multi-factor coupling. It provides literature support and directional guidance for subsequent research on hybrid forecasting models such as the "mechanism-data" approach. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Machine Learning, Iron and Steel, Modeling and Simulation |