基于传感器融合的连续蒸煮制浆动态系统的控制模型研究

The continuous cooking is the most advanced chemical pulping technology in the world, attributed to the advantages of stable quality of pulp, low energy consumption, and less pollution. Because of the big variations in the pulping raw materials in China, it is very difficult to provide a proper control by the current H-factor modeling method for the continuous cooking process, resulting in the ineffectiveness in the utilization of raw pulp materials and energy. Therefore, it is necessary to find a good solution to solve the problem in controlling continuous pulping process. The key in this research is based on a chemical sensor that is able to monitor the dissolved lignin and total solids content in the cooking liquors. These parameters are directly related to the quality index (i.e., pulp kappa and yield) for the pulping process control, from which a sensor fusion control model for continuous cooking dynamic system can be established. The study will focus on: the detection models of the dissolved lignin concentration and solid content in the fluid with high temperature high consistency is obtained by the attenuated total reflection-UV spectroscopic sensor; the mass transfer of the lignin in two phases of solid-liquid and the movement of lignin in the two-dimensional space in the continuous digester system solved by differential equations, as well as the development of their coupling model; based on the testing-points on the digester selected, the dynamic cooking control model is established though fusing the information collected by the ATR sensors in the different position of the digester. The results from this project will provide an optimized the process control in the continuous digester, aiming at efficiently utilizing the raw materials, saving energy and reducing pollution in the pulping production.

具有成浆质量稳定、能耗低、污染少等优点的连续蒸煮工艺是目前世界上最先进的一种化学制浆技术。基于我国制浆原材料（树龄，树种）波动大的特点，由现行H因子法对连续蒸煮过程控制不当而造成原材料、能源的浪费严重。为了解决该系统控制的瓶颈问题，本项目基于传感器融合技术实时获取与蒸煮质量（由纸浆卡伯值和得率共同表征）直接相关的溶出木素量和总固形物量的信息，在此基础上创建连续蒸煮动态系统控制的新模型。主要内容有：引入衰减全反射光谱（ATR）传感检测技术，研究动态循环高温高浓过程流体中木素浓度和固含量的在线检测模型；基于微分的思想对蒸煮器进行模拟分割，研究蒸煮过程中木素在固液两相传质与在二维空间运动的耦合模型；研究蒸煮器上传感器检测点的布局，融合处于不同位置的多个ATR传感器所采集的信息，建立蒸煮动态控制模型。本项目的成果将为优化连续蒸煮生产的有效控制，进而实现原材料高效利用、节能减排、清洁生产，提供指导。

作为一种先进的化学制浆技术，连续蒸煮是一个原料和反应溶液连续动态输入、同时纸浆连续动态输出的多相化学反应过程，其原料在蒸煮器内垂直运动过程中发生脱木素反应最终形成纸浆，期间伴有不同流向的复杂的动态的过程液循环。对于这样一个复杂的动态系统，建立其实用性强的控制模型一直是国内外连续蒸煮制浆生产的一个难点。为此，本项目基于传感器融合技术实时获取与蒸煮质量（由纸浆卡伯值表征）直接相关的溶出木素量，在此基础上创建连续蒸煮动态系统控制的新模型。本项目按照原计划开展工作，取得以下重要结果：（1）针对连续蒸煮工艺系统的复杂性，采用了连续搅拌反应器逼近蒸煮器，将蒸煮器看作一系列具有均匀成分和温度的连续搅拌反应器串联，研究了单个连续搅拌反应釜内多相流、多介质宏观上的扩散平衡动力学；（2）基于微分的思想，以单位体积的木片为研究对象，基于木素在蒸煮过程中遵守物质守恒定律，建立了纸浆卡伯值与蒸煮过程溶解木素总量之间的关系；（3）针对蒸煮过程液中溶解木素浓度难以检测的问题，采用衰减全反射紫外可见光谱（ATR-UV）技术开发了蒸煮过程流体中溶解木素浓度的在线检测方法；（4）针对传感器信号数据量大难以分析的问题，采用了支持向量机、人工神经网络等一些非线性智能算法提取数据中的关键信息，精确求解了过程液中溶解木素浓度；（5）为求解单位体积木片蒸煮过程实时溶解木素总量，在宏观剖析蒸煮器药液循环规律的基础上，设计了连续蒸煮器的物理布局；（6）基于以上研究，融合多源信息，建立了以蒸煮过程液中溶解木素浓度为主导变量的实用性强的连续蒸煮过程控制新模型；（7）实验室模拟实验验证了连续蒸煮过程控制模型的正确性和合理性，将这种建模方法推广试用于纸浆厂的间歇蒸煮工艺，验证了连续蒸煮过程控制模型建模方法的实用性和有效性。本项目的研究成果将为提升我国连续蒸煮制浆自动控制水平、提高制浆原料的利用率和纸浆的品质提供技术支持。