Simulation and result of
AFD
According to the previous description of AFD theoretical a 1 kW
single-phase PV power generation system is established in
Matlab/Simulink. This control was designed to monitor
the\(\text{\ \ F}_{\text{PCC}}\text{\ and\ }V_{\text{PCC}}\). The AFD
controller will produce a signal to the inverter to stop supplying to
the local load if voltage and frequency are out of the limits determined
by IEEE 1547 [18]. The simulation module includes the inverter
circuit connected to a utility grid control and AFD islanding Detection
section at the . The function of AFD controller module was achieved by
using the s-function in Matlab/Simulink. The \(V_{\text{PCC}}\ \)and\(I_{Pv-inv}\) are in phase at the initial setting. The grid supply
was set to be disconnected at t=0.1 s.
So the sequence of the AFD detection algorithm Firstly, the frequency
data was taken and tested against the UFP/OFP threshold, if it is not in
the threshold, the islanding occurred, else it is not islanding. If
there is no islanding, a source with frequency modified by the\(C_{f}\ \)is then injected into the inverter output current every half
cycle and every full cycle to produce a \(t_{z}\ \)on the output current
waveform. The value of the \(C_{\text{f\ }}\)will slightly increase
every half cycle to drift the current frequency from voltage frequency
until islanding is detected and a signal will be sent to stop the
inverter from operating.
Figures 11-13 are the cases result for AFD controller which these cases
were difficult to be detected by the passive IDM.
The results show that the AFD is capable of detecting islanding
effectively with very small NDZ and a detection time within 0.06 s. The
disturbance injection plays a significant role in performing islanding
detection to meet the PV grid interconnection standard. The simulation
results show the more disturbances injected the faster the islanding
detection time but the higher the harmonic distortion.