Black-Box Modelling of AC-DC Rectifiers for RFID Applications Using Support Vector Regression Machines

Abstract

In this paper the use of support vector regression (SVR) machines for modelling of nonlinear dynamic behaviour of an AC-DC rectifier is presented. The use of SVR machines yields a black-box model which significantly reduces the simulation time. The simulated AC-DC rectifier is commonly found in radio-frequency identification (RFID) circuits and it is an excellent test case as it involves two substantially different time constants, the first one related to the radio-frequency signal and the second related to the rectification process. Two different AC-DC rectifier models are proposed. The first model models the DC voltage at the output of the rectifier after the transient process, i.e. it models the stationary value of the DC voltage. The second model models the transient process, i.e. the change of the output voltage as a function of time. The accuracy and the simulation speed are compared to the transistor levels simulations and it can be concluded that the SVR models are accurate and at least an order of magnitude faster than the transistor level models.