Abstract: The neural networks are used to estimate the parameters of linear systems with or without delay. Single layer neural networks can be used for system identification of linear time invariant systems either in time domain or in frequency domain. A three layer neural network which consists of one hidden layer is presented to estimate the parameters of a special kind of linear systems with delay. The application results using the neural networks have the same precision as that obtained using the conventional identification approaches. On the other hand, the neural network technique has better performance than the conventional techniques. For example, in the neural network approach, the parameters can be adjusted dynamically based on a learning algorithm. The neural networks are less sensitive to noise and easier to implement.

Key words: Frequency domain, Neural networks, Parameter estimation, Time domain