基于FNN解耦纸张定量水分控制策略的研究与应用

Research and Application of Basis Weight and Moisture Content Control Strategy Based on FNN Decoupling

关键词:  定量  水分  模糊控制  神经网络  FNN

Key Words:basis weight  moisture content  fuzzy control  neural network  FNN

摘要:针对纸张抄造过程中纸张定量与水分之间存在强耦合的问题，提出一种模糊神经网络（Fuzzy Neural Network，FNN）的解耦控制器，首先利用模糊控制对控制系统进行耦合补偿，然后利用神经网络的自学习、自调整能力不断在控制过程中优化模糊控制规则及解耦补偿参数，成功地将纸张抄造过程的多变量系统转变为单变量系统，实现纸张定量、水分之间的解耦。仿真结果表明，采用FNN解耦控制器具有较好的动态响应和较强的鲁棒性。将该策略应用于国内某造纸厂的纸板机控制系统，纸张定量控制精度为±3.9 g/m2左右，水分控制精度为±1.0%左右，满足该纸机定量水分高精度控制要求。

Abstract:Aiming at the problem of strong coupling between paper basis weight and moisture content in the paper making process, a decoupling controller based on fuzzy neural network （FNN） was proposed. Firstly, coupling compensation of control system was implemented using fuzzy control. Secondly, the neural network was used to optimize the fuzzy control rules and the decoupling compensation parameters in the control process. The multivariable system of paper making process was transformed into a single variable system, which could realize the decoupling between the paper basis weight and moisture content. The simulation results showed that the FNN decoupling controller had better dynamic response and stronger robustness. The strategy was successfully applied to the production of a paper mill in Zhejiang province by the Shaanxi XiWei Aotumation Control Engineering Limited. The basis weight control accuracy was about 3.9 g/m2, and the water content control precision was about 1%, which met the high precision control requirement of the paper machine.