造纸生产过程的分布式协同控制与智能优化

The study on optimization and control of pulp and papermaking producing processes plays an important role in enhancing product quality, reducing production cost, responding market rapidly, and boosting the development of pulp and papermaking industry in Zhejiang province. This project focuses on studying energy efficiency optimization problems in pulp and papermaking production. The operation of pulp and papermaking mills in Zhejiang has three features: the equipment is obsolete, the level of automation is low, and the raw material changes frequently. Aiming at these three features, this project will study the modeling, control and optimization problems of pulp and papermaking processes. A characteristic model with an inference procedure will be developed based on the mechanism model. To implement the real-time optimization, a two-level optimization method will be proposed for the global and the local optimization of the pulp and papermaking processes. With the paramemters from the optimization, the basis weight, moisture, ash content, and so on are controlled with an adaptive nonlinear control strategy in control layer, and relevant equipments run at the best working points. In this way, the energy consumption will be reduced in pulp and papermaking processes.

制浆造纸生产过程的优化控制研究对提高产品质量、降低生产成本、快速响应市场需求和促进浙江省造纸行业的发展具有重要意义。本项目针对浙江省中小制浆造纸企业装备陈旧、自动化水平低和原料品质不稳定三大特征，以提高制浆造纸生产过程运行能效为目标，研究生产过程的优化控制策略。项目将在制浆造纸生产过程机理模型的基础上开展特征建模和知识推理建模方法的研究，提出高效建立复杂过程混合模型的新方法；对制浆造纸生产过程进行分级优化，协调生产过程成纸品质指标和节能指标；针对定量、水分和灰分等关键对象的非线性多变量大时滞特点研究自适应控制方法。通过上述研究揭示制浆造纸生产过程能耗分布规律和成纸质量参数动态响应特性，提出大规模复杂系统分级优化方法，发展非线性系统自适应控制理论，在保证成纸质量的前提 下实现制浆造纸生产全过程的节能优化运行。

项目在制浆造纸生产过程机理模型的基础上开展特征建模和系统辨识建模方法的研究，提出高效建立复杂过程混合模型的新方法；针对非线性多变量大时滞系统模型提出多种新型自适应动态解耦、控制、观测算法；结合机器学习、强化学习方法挖掘纸浆造纸生产过程特征，提出制浆造纸生产过程分级节能智能优化设计方案，协调生产过程成纸品质指标和节能指标；针对关键单元的非线性多变量大时滞特点研究特征压缩自适应控制方法，完成了项目的研究目标。项目共发表论文69篇，其中 SCI论文42篇，申请中国发明专利17项，包括美国专利2项。在过程控制顶级刊物IECR发表1篇封面论文，荣获IEEE CCTA优秀论文奖。成果在实际生产过程中得到了应用。