Neural networks for continuous-time systems modeling from input/output data

In this paper we have investigated the neural network techniques for estimating the delay and the other parameters of a continuous-time linear system. The algorithm has been tested on a number of examples, including actual noisy sensor data and propagation delay. We have used a simple form of back propagation or gradient descent algorithm for training the network. In terms of computer memory and computation time, our circuit is very simple in structure and requires less hardware for implementation.<
>