基于RBFNN的内模控制在气垫式流浆箱中的应用

针对气垫式流浆箱常规控制中总压和浆位之间存在严重耦合的问题，在分析传统解耦控制方案的基础上，提出了一种基于RBF（径向基）神经网络的内模控制系统（IMC），采用径向基神经网络控制器（RBFNNC）在流浆箱控制中既充当对象模型，又充当控制器，实现了总压和浆位的解耦控制。仿真结果表明，该控制方法具有实现简单、控制效果好的优点。

Application of internal model control based on RBFNN in the closed air-cushioned headbox

The internal model control（IMC） sytem based on radial basis function neural network （RBFNN） is presented aiming at serious decoupling existing between total pressure and level in the headbox and on the basis of analysis of the traditional decoupling control system. The radial basis function neural network controller （RBFNNC） both as a object model and as a controller in the headbox control, is adopted to decouple the control for level and total pressure. The simulation result indicates that this system has the benefits of simple realization and good control effect.