基于神经网络逆系统的空气重介流化床多变量解耦控制  

Multivariable decoupling control based on neural network inverse system in an air-dense medium fluidized bed process

作　　者：高松[1] 王振翀[1] 

机构地区：[1]中国矿业大学(北京)机电与信息工程学院,北京100083

出　　处：《煤炭学报》2008年第2期225-227,共3页Journal of China Coal Society

基　　金：国家自然科学基金资助项目(50474030);国家高技术研究发展计划(863)资助项目(2007AA05Z315)

摘　　要：针对空气重介流化床干法分选过程的多变量非线性强耦合特点,在空气重介流化床干法分选过程的神经网络逆系统解耦方法的基础上,设计了线性闭环PI调节器,以实现空气重介流化床干法分选过程的高性能解耦控制.仿真结果表明,该方法能够较好地解决空气重介流化床干法分选过程的控制问题,具有较强的鲁棒性.Coal dry beneficiation with air-dense medium fluidized bed （ADMFB） is a nonlinear and multivariable coupling system, and strongly coupled system. A linear close-loop adjustor based on neural network inverse system which was existed for beneficiation process was designed to obtain the good decoupling control performance. The simulation results show that good control performance （high accuracy and good robust） can be obtained in dry beneficiation process based on this decoupling control method.

关 键 词：神经网络 逆系统 解耦控制 分选过程 PI调节器 

分 类 号：TD922.7[矿业工程—选矿]