现代有色冶金工业中常采用多个反应器串联形成级联反应器，它是决定有色冶炼生产关键性能指标的核心设备。然而，冶金反应器传热传质过程动力学复杂、多设备串级关联影响且存在目标冲突，特别是我国有色金属矿源成分波动、生产过程重要工艺参数难以实时准确测量等特点带来不确定性，严重影响有色冶炼生产过程优化运行。为此，本项目在研究单反应器模型的基础上进行动态关联分析并结合生产历史数据和经验知识，采用多值中智集模糊方法建立有色冶金级联反应器不确定动态优化模型；采用交互式多准则决策预先处理不确定性并提出基于主从式多智能体的分布式智能优化计算方法，协调各级反应器目标与整体目标之间的矛盾；充分利用实际生产运行数据进行工况评估并自适应校正整体优化模型不确定因子，实现运行反馈滚动优化；结合仿真和工程试验研究，验证所提方法的可行性与有效性。本项目研究将丰富有色冶金过程控制理论与方法，成果的应用将推动有色冶炼绿色高效生产。

In order to avoid short reaction time, high speed, and low transformation rate caused by excessive reaction, the modern nonferrous metallurgy industry often uses multiple reactors to form a reactor cascade in practice. It is the core equipment that can determine the key performance indicators for nonferrous metals production. However, the kinetics of heat and mass transfer processes in metallurgical reactors is complex, and there are cascade reactors with correlation effects and conflicting goals. In particular, the characteristics of nonferrous metal ores in China fluctuate, and important technological parameters in the production process are difficult to measure accurately in real time, which seriously affect the operational optimization of nonferrous metals production process. For this reason, firstly, this project carries out dynamic correlation analysis, combines both production history data and experience knowledge, adopts multi-valued neutrosophic set method to establish non-deterministic dynamic optimization model of cascade reactors in nonferrous metallurgy based on the mechanism of a single reactor. Then, the TODIM (an acronym in Portuguese of interactive and multiple attribute decision making ) method is used to deal with uncertainties in advance and a distributed intelligent optimization method based on multi-agent systems with master-slave mode is proposed to resolve conflicts among single reactor objectives and the overall objective. Furthermore, actual production operational data are fully utilized for performance assessment and to update the uncertain factors adaptively in the overall optimization model to realize operational optimization through feedback. Finally, the feasibility and effectiveness of the proposed method will be testified in both simulation and practical industrial applications. This project will enrich the theory and method of process control in nonferrous metallurgy, and would promote green and efficient production of nonferrous metals.