Inverse Modeling of Nonlinear Systems and Application

Inverse modeling is discussed based on neural networks. U-D factorization Kalman filter learning algorithm is introduced to train a feed forward multilayer neural network, which is used as inverse models. A scheme of Direct Current Electric Arc Furnace (DC EAF) control systems using Neural Networks (NN) is presented.