Efficient modelling of an RF MEMS capacitive shunt switch with artificial neural networks

In this paper an efficient way of modelling an RF MEMS switch will be demonstrated. On the base of several fullwave numerical simulations of a switch, artificial neural networks (ANNs) are established to relate different input and output parameters of a switch. The good coincidence between the simulated data and the model is shown at the example of a capacitive shunt switch in coplanar technology. S-parameters of the switch or the resonant frequency are computed with high accuracy for different geometrical parameters, with the same data also the inverse problem of determining the required geometry for a given resonance is solved.