一种基于约束多目标深度强化学习的挥发窑温度场优化方法
By optimizing the temperature field of the volatilization kiln through constrained multi-objective deep reinforcement learning, the conflict between zinc recovery rate and carbon emissions and the complexity of temperature distribution patterns were resolved, achieving low-carbon optimization and improved time efficiency in the zinc smelting process.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CENT SOUTH UNIV
- Filing Date
- 2023-09-18
- Publication Date
- 2026-07-17
AI Technical Summary
The conflict between zinc recovery rate and carbon emissions in volatilization kilns, as well as the complex temperature distribution pattern and the constraints on temperature field controllability, make it difficult to efficiently obtain the optimal temperature field.
A constrained multi-objective deep reinforcement learning approach is adopted. By designing an evaluation index for uncontrollable factors and a dynamic penalty method, combined with a Chebyshev scalarization function, the temperature field of the volatilization kiln is optimized, and the optimization results are obtained by training a deep neural network.
While meeting the constraints of temperature field controllability, the zinc recovery rate was improved, carbon emissions were reduced, optimization time was shortened, and operational guidance for low-carbon operation was provided.
Smart Images
Figure CN117236174B_ABST