烟草烘丝过程是卷烟生产过程中最重要的加工工序，如何提高烘丝过程出口烟丝含水率的控制水平一直是卷烟生产企业面临的难题。本项目从烟草烘丝过程的工程实际问题出发，秉持为实现最终工业控制而建模，拟进行如下研究：1.设计可准确表征烟草烘丝过程各阶段非线性特性的、适用于后续控制的多采样率FNN-ARX模型及其参数估计方法；2.针对烘丝过程中普遍存在的“干头干尾”问题，设计基于头尾段FNN-ARX模型长期预测的优化设定控制方法，对头尾段输入工艺变量的最优设定曲线进行优化；3.充分利用FNN-ARX模型的结构特点和蕴含的参数变化速率信息，同时考虑烘丝中间过程有界不确定干扰的影响，设计基于FNN-ARX模型的、无需利用系统稳态平衡点信息的输出跟踪鲁棒预测控制算法，并以实现在烟草烘丝工业实际应用为目的，开发出该算法基于离线计算在线综合模式的在线快速实现方法。最终，达到实现对整个烘丝过程精准“智”控的目的。

The cut tobacco drying process is the most important processing process in the cigarette production process. How to improve the control performance of the drying process has always been a challenging issue for cigarette manufacturers. Based on the idea of modeling for the ultimate industrial control, the project is aimed at the actual engineering problems of the cut tobacco drying process, and the following research will be carried out: 1. We will design the multi-sampling rate FNN-ARX models and their parameter estimation methods, which can accurately characterize the nonlinear characteristics of each stage of the cut tobacco drying process and are suitable for subsequent controllers design. 2. In order to solve the common problem of the "over-dried" in the head and tail stages of the drying process, the optimal setting control methods based on the long-term prediction of the two stages' FNN-ARX models will be designed. 3. Based on the parameter variation rate information of the FNN-ARX model, a robust predictive control algorithm with considering the unknown and bounded disturbance will be designed for output-tracking control of the middle stage drying process without using the system's steady state information. Moreover, in order to achieve the practical application in the cut tobacco drying industry, an online fast implementation method of the robust predictive control algorithm based on offline calculation online integrated mode will be developed. Finally, through the research of this project, we want to achieve precise "intelligence" control of the whole cut tobacco drying process.