Model Predictive Control Based on System Identificationof Photovoltaic Grid Connected Inverter

Abstract

This paper proposes a model predictive control of photovoltaic grid-connected inverter based on system identification. The single phase inverter is experimented and its model is determined by using System identification approach with Hammerstein-Wiener model. The derived nonlinear voltage model has accuracy more around 97.34% and it is transformed to the state space model by linearization. A simulation of model based controller uses the discrete time model of inverter to predict the behavior of the output voltage for each possible switching state every sampling time. Then cost function is applied as a criterion for selecting the most suitable switching state for the next sampling interval. The model output is compared with the reference voltage sine wave and the error is feedback to the optimizer. Simulation results shown that the proposed control scheme can achieve the output target with 97% of accuracy.Index Terms-Model predictive control, system identification, hammersteinwiener model, grid connected inverter.

D‐Q frame predictive current control methods for inverter stage of solid state transformer

“…where V c is the amplitude of the control voltage, w = 2π f , φ is the shift angle of the signal. From (2) and (12), the control variable in the second orthogonal imaginary circuit v β is obtained as…”

“…It is well known that proportional integral (PI) controllers could not cope with system uncertainties [6]. Numerous methods have been proposed in the literature to overcome the drawbacks of conventional PI controller, including sliding mode control (SMC) [7,8], fuzzy logic control (FLC) [9,10], model predictive control (MPC) [11][12][13]. Despite SMC has its advantages of insensitive to parameter variations and external disturbances, it suffers from an undesirable chattering phenomenon and limited switching frequency when using software control.…”