【摘要】 工业过程控制系统中的控制对象往往具有多变量强耦合、不确定、非线性等特征,解决这类控制问题的有效方法是采用先进控制技术,或在常规控制的基础上叠加先进控制的补偿分量。以对角递归神经网络控制器为研究对象,分析了该控制器的结构特点和学习算法,探讨对角递归神经网络技术的应用现状与工程实现中的相关问题,并提出了先进控制与常规控制同构组态的解决方法。更多还原

【Abstract】 As controlled objects in industrial process control systems tend to have such characteristics as multi-variable strong coupling,uncertainty and nonlinearity,the most effective way to solve such control problems is to take advantage of advanced control technology or superimpose advanced control compensation component on conventional control. Taking diagonal recurrent neural network( DRNN) controller as research object,this paper analyzes the structural features and learning algorithm of the controller,discusses the application status of DRNN technology and other related issues with respect to engineering realization,and finally presents the solution of isomorphic configuration of advanced control and conventional control.更多还原