Parameter identification from frequency response of MEMS energy harvesters

In this study, we present theoretical analysis and numerical results on a simple technique for extracting unknown model parameters for MEMS electrostatic energy harvesters. We show that the frequency response can be utilized in a least-squares minimization scheme to estimate the damping coefficient, mechanical stiffness and transducer/load parasitic capacitances. The accuracy of the method is tested by application to simulated cases of linear and non-linear harvesters. A single data sweep from such a pseudo-experiment suffices to determine the unknown parameters of the electromechanical model with accuracy. The method is shown to work satisfactorily for both linear and nonlinear devices.