Abstract:
Currently, artificial neural networks are being used to solve problems related to control. To determine reliability of the neuro-control technique is to test it on a variety of realistic problems with the aim of seeing whether it works well and where it needs further refinement.
In the presented work, a multilayered back propagation ANN is first trained, to learn the inverse dynamic model of a temperature' control system and then configured as a direct controller to the process. The ability of ANN to learn the inverse model of the process plant is based on input vectors with no a-priori knowledge regarding its dynamics.
Three artificial neural networks with different structures are trained for other first order plant characteristics. Finally, a generalised ANN is obtained, which can be used as a direct controller for any of the first order plant characteristics subject to variations of any or all plant parameters with a range of ± 25 %.
