Page 66 - Renewable Energy Devices and System with Simulations in MATLAB and ANSYS

Overview of Single-Phase Grid-Connected Photovoltaic Systems 53

power grid as shown in Figure 3.2. A conventional control structure for such a grid-connected PV
system thus consists of a two-cascaded loop in order to fulfill the demands/requirements [32, 70]—
the outer power/voltage control loop generates the current references and the inner control loop is
responsible for shaping the current, so the power quality is maintained, and also it might perform
various functionalities, as shown in Figure 3.18.
Figure 3.19 shows the general control structure of a single-phase single-stage grid-connected
PV system, where the PV inverter has to handle the fluctuating power (i.e., MPPT control) and
also to control the injected current according to the specifications shown in Figure 3.18. As it
can be observed in Figure 3.19, the control can be implemented in both stationary and rotating
reference frames in order to control the reactive power exchange with the grid, where the Park
transformation (dq → αβ) or inverse Park transformation (αβ → dq) is inevitable [70]. In terms
of simplicity, the control in the stationary reference frame (αβ-reference frame) is preferable, but
it requires an orthogonal system to generate a “virtual” system, which is in quadrature to the real
grid. In the dq-reference frame, the MPPT control gives the active power reference for the power
control loop based on proportional integral (PI) controllers, which then generate the current ref-
erences as shown in Figure 3.19b. The current controller (CC) in the dq-reference frame can be
PI controllers, but current decoupling is required in order to alleviate the interactive impact of
the d-axis and q-axis currents in the synchronous rotating reference frame (i.e., the dq-reference
*
frame). In contrast, enabled by the single-phase PQ theory [32, 71], the reference grid current i g
can be calculated using the power references and the in-quadrature voltage system. In that case,
the PI controller will give an error in the controlled grid current. The controller (CC) should be
designed in the αβ-reference frame. For example, a proportional resonant (PR) controller, a repeti-
tive controller, or a deadbeat controller [70, 72–75] can directly be adopted as the CC as shown
in Figure 3.19c.
Notably, since the CC is responsible for the current quality, it should be taken into account in the
controller design and the filter design (e.g., using high-order passive filter, LCL filter). By introduc-
ing harmonic compensators [26, 32, 70, 72] and adding appropriate damping for the high-order filter
[76, 77], an enhancement of the CC tracking performance can be achieved.
Similar control strategies can be applied to the double-stage system, as shown in Figure 3.20.
The difference lies in that the MPPT control is implemented on the DC–DC converter, while the
other functionalities are performed on the control of the PV inverter. There are other control solu-
tions available for single-phase grid-connected PV systems [78–80]. For example, the instantaneous
power is controlled in [79], where the synthesis of the power reference is a challenge; in [80], a
one-cycle control method has been applied to single-stage single-phase grid-connected PV inverters
for low power applications.

3.4.2 Grid Synchronization

It should be noted that the injected grid current is demanded to be synchronized with the grid volt-
age as required by the standards in this field [70]. As a result, grid synchronization is an essential
grid monitoring task that will strongly contribute to the dynamic performance and the stability of
the entire control system. The grid synchronization is even challenged in single-phase systems, as
there is only one variable (i.e., the grid voltage) that can be used for synchronization. Nevertheless,
different methods to extract the grid voltage information have been developed in recent studies
[26, 70, 81–84] like the zero-crossing method, the filtering of grid voltage method, and the phase-
locked loop (PLL) techniques, which are important solutions.
Figure 3.21 shows the structure of the PLL-based synchronization system. It can be observed that
the PLL system contains a phase detector (PD), namely, to detect the phase difference, a PI-based
loop filter (PI-LF) to smooth the frequency output, and finally a voltage-controlled oscillator (VCO).
Accordingly, the transfer function of the PLL system [26, 84, 85] can be obtained as