Nonlinear continuous time modeling of a high pressure mercury vapor discharge lamp using feed forward back-propagation neural networks

A continuous time nonlinear model is proposed in this paper for modeling a high pressure mercury vapor discharge lamp. Parametric continuous models of discharge lamps exist in the literature. However, the development of new nonlinear models of this discharge lamp is an appealing area. Hence, this paper focuses on the performance study of a non linear model namely Herrick´s conductance model to fit the input-output behavior of the discharge lamp. The model coefficients are identified using the synaptic coefficients of a feed forward back-propagation neural network.