研究了递归网络模型在传感器动态建模中的应用,给出了递归网络模型的结构及相应的训练算法。该方法避免了传感器模型阶次的选择的困难。试验结果表明,应用递归网络对传感器进行动态建模是一种行之有效的方法。 关键词: 递归网络; 传感器; 动态建模; 递推预报误差算法 Abstract : Dynamic modeling of sensors is an important aspect in the fields of instrument technique. A new method hasbeen discussed for dynamic modeling of sensors using recurrent network model. A recursive prediction error algorithm isapplied to train the recurrent network. Experimental results show that the dynamic modelingmethod is effective. Key words : recurrent network ; sensor ; dynamic modeling ; recursive prediction error(RPE