An important aspect of control system design is system identification. Once the parameters of the systems are identified, a controller can be designed to meet a set of design specifications. However, when there is a large range of uncertainties in an existing system model, a controller added to a plant would not perform as expected. In this case, an adaptive control scheme may be used, but using a learning parameter identification scheme might yield a better solution. Neural network techniques seems to offer a powerful solution to this problem. Therefore, by using a neural network to learn what the parameters of a complicated system are, a controller can be designed to keep a good performance level. %%% The proposed research project would involve using a neural network to identify the parameters of a system so that a better controller can be designed. The objective of this research is to obtain a better estimate of the parameters of any given system. After a mathematical model is identified, a controller can be designed. The neural network would continually learn about the system so that the model obtained will be correct.
