Model based control using artificial neural networks

An internal model control (IMC) using artificial neural networks is presented in this paper. IMC is significant because the stability and robustness properties of the structure can be analysed and manipulated in a transparent manner, even for nonlinear systems. Artificial neural networks are used for the construction of plant models and their inverse. Backpropagation algorithm is used to train the network and the effect of training parameters to network performance is investigated. The proposed control method is studied for real-time control on a heater PT326. The performance of the neural control method is compared with that of a conventional PID controller, which is tuned by Ziegler-Nichols´ ultimate cycle method. The control structure is shown to perform well in robust control