A method for constructing a digital twin system for zinc hydrometallurgy based on multi-process collaborative simulation
By constructing a digital twin system for multi-process collaborative simulation of hydrometallurgical zinc smelting, and utilizing a mechanism- and data-driven architecture and multi-agent reinforcement learning, the problem of inter-process coupling and optimization in hydrometallurgical zinc smelting production was solved. This system achieved dynamic simulation and global optimization of the entire process, thereby improving production efficiency and control level.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HEFEI UNIV OF TECH
- Filing Date
- 2026-03-27
- Publication Date
- 2026-07-17
AI Technical Summary
In the process of hydrometallurgical zinc production, the strong coupling and nonlinear correlation between processes make it difficult to monitor the concentration of key ions in real time and accurately, and the reaction process and optimization control are difficult. In addition, the industrial field data is multi-source, heterogeneous and high-noise, which increases the difficulty of accurate modeling and optimization decision-making.
A digital twin system for hydrometallurgical zinc refining based on multi-process collaborative simulation is constructed. Through a mechanism- and data-driven architecture, a basic model framework for each process is established. Online learning is performed using a Long Short-Term Memory (LSTM) network, combined with an intelligent decision engine based on multi-agent reinforcement learning, to achieve collaborative optimization between processes.
It realizes dynamic simulation and global optimization of the entire process of hydrometallurgical zinc smelting, supports autonomous collaborative decision-making of process operation parameters, improves production efficiency and control level, and reduces overall consumption.
Smart Images
Figure CN122405993A_ABST