The transition toward sustainable energy systems necessitates a profound evolution in power grid infrastructure, with grid connected inverters serving as the critical interface between distributed renewable generation and the main electrical network. Their performance directly dictates power quality, stability, and the overall security of the grid, especially under non-ideal conditions. One of the most prevalent and challenging grid anomalies is voltage imbalance, typically arising from asymmetrical faults or unequal single-phase loads, which can severely degrade the synchronization and control performance of conventional inverter systems.
The Phase-Locked Loop (PLL) is the cornerstone of synchronization for a grid connected inverter, enabling it to track the grid voltage phase angle and frequency accurately. Under balanced grid conditions, the standard Synchronous Reference Frame PLL (SRF-PLL) performs impeccably. However, its performance falters during voltage imbalances. An unbalanced grid voltage contains both positive-sequence and negative-sequence fundamental components. When transformed into the SRF-PLL’s rotating dq frame, which rotates at the positive-sequence frequency, the negative-sequence component manifests as an oscillating disturbance at twice the fundamental frequency (2ω). These oscillations corrupt the estimated phase angle, leading to distorted current references, increased harmonic injection, and potential instability, particularly in weak grids with high impedance.
To address this fundamental limitation, this analysis focuses on the Decoupled Double Synchronous Reference Frame PLL (DDSRF-PLL) and presents an enhanced control strategy. The core improvement involves replacing the traditional first-order Low-Pass Filter (LPF) within the DDSRF structure with a third-order LPF and integrating a novel, fast-acting current limiter. This combination significantly improves harmonic immunity, dynamic response, and system protection for a grid connected inverter operating in harsh grid environments.
Mathematical Foundation of the DDSRF-PLL
An unbalanced three-phase voltage can be decomposed into its symmetrical sequence components. For the purpose of PLL design targeting fundamental frequency tracking, we consider the positive-sequence (fundamental) and negative-sequence (fundamental) components, ignoring zero-sequence for three-wire systems. In the stationary αβ-frame, obtained via Clarke transformation, this voltage can be represented as:
$$ \mathbf{V}_{\alpha\beta} = V^+ \begin{bmatrix} \cos(\omega t + \phi^+) \\ \sin(\omega t + \phi^+) \end{bmatrix} + V^- \begin{bmatrix} \cos(-\omega t + \phi^-) \\ \sin(-\omega t + \phi^-) \end{bmatrix} $$
Where $V^+$, $\phi^+$ and $V^-$, $\phi^-$ are the amplitude and phase of the positive- and negative-sequence components, respectively.
The ingenious concept of the DDSRF-PLL is to employ two rotating reference frames simultaneously: one ($dq^+$) rotating at an estimated positive angular frequency $\hat{\omega}$, and the other ($dq^-$) rotating at $-\hat{\omega}$. When the PLL is locked ($\hat{\theta} = \omega t$), the positive-sequence component appears as a DC quantity in the $dq^+$ frame, while the negative-sequence component appears as a DC quantity in the $dq^-$ frame. The fundamental challenge is that these components are cross-coupled; the negative-sequence component causes a 2ω ripple in the $dq^+$ frame and vice-versa. The DDSRF structure introduces a decoupling network to cancel this effect.
The Park transformations for the two frames are defined by the rotation matrices $T^+(\hat{\theta})$ and $T^-(\hat{\theta})$:
$$ T^+(\hat{\theta}) = \begin{bmatrix} \cos\hat{\theta} & \sin\hat{\theta} \\ -\sin\hat{\theta} & \cos\hat{\theta} \end{bmatrix}, \quad T^-(\hat{\theta}) = \begin{bmatrix} \cos\hat{\theta} & -\sin\hat{\theta} \\ \sin\hat{\theta} & \cos\hat{\theta} \end{bmatrix} $$
Applying these transforms yields the coupled voltages in each frame:
$$ \begin{bmatrix} v_d^+ \\ v_q^+ \end{bmatrix} = T^+(\hat{\theta}) \mathbf{V}_{\alpha\beta}, \quad \begin{bmatrix} v_d^- \\ v_q^- \end{bmatrix} = T^-(\hat{\theta}) \mathbf{V}_{\alpha\beta} $$
Under lock conditions ($\hat{\theta}=\omega t$), the ideal decoupled outputs, denoted $v_d^{**}$ and $v_q^{**}$, should contain only the DC values corresponding to the sequence amplitudes. The decoupling network calculates these as:
$$ \begin{aligned}
v_d^{**+} &= v_d^+ – \bar{v}_d^{-} \cos(2\hat{\theta}) – \bar{v}_q^{-} \sin(2\hat{\theta}) \\
v_q^{**+} &= v_q^+ + \bar{v}_d^{-} \sin(2\hat{\theta}) – \bar{v}_q^{-} \cos(2\hat{\theta}) \\
v_d^{*-} &= v_d^- – \bar{v}_d^{+} \cos(2\hat{\theta}) + \bar{v}_q^{+} \sin(2\hat{\theta}) \\
v_q^{*-} &= v_q^- + \bar{v}_d^{+} \sin(2\hat{\theta}) + \bar{v}_q^{+} \cos(2\hat{\theta})
\end{aligned} $$
Here, $\bar{v}_d^{+}, \bar{v}_q^{+}$ and $\bar{v}_d^{-}, \bar{v}_q^{-}$ are the low-pass filtered (averaged) versions of the $dq$ components from each frame. This network effectively subtracts the estimated oscillatory disturbance caused by the opposite sequence.
Complex Variable Model and LPF Performance Analysis
A powerful method to analyze the dynamic performance and harmonic rejection of the DDSRF-PLL is through complex variable modeling. Defining complex signals $v_{dq}^+ = v_d^+ + jv_q^+$ and $v_{dq}^- = v_d^- + jv_q^-$, the system can be elegantly described in the complex frequency domain. The core relationship between the decoupled outputs $\bar{v}_{dq}^{**}(s)$ and the inputs depends critically on the Low-Pass Filter (LPF) transfer function $G_{LPF}(s)$.
The closed-loop transfer functions from the actual sequence components to their estimated DC values are derived as:
$$ \frac{\bar{V}_{dq}^{**+}(s)}{V_{dq}^+(s)} = G_{LPF}(s) \frac{1 – G_{LPF}(s+j2\omega)}{1 – G_{LPF}(s)G_{LPF}(s+j2\omega)} $$
This transfer function reveals the system’s ability to extract the positive-sequence DC component while rejecting the negative-sequence disturbance. The effectiveness of this rejection at the disturbance frequency ($s = j2\omega$) is paramount. The order and characteristics of $G_{LPF}(s)$ are therefore crucial design parameters for any grid connected inverter synchronization unit.
We analyze three candidate filters:
- First-Order LPF: $G_{LPF}^1(s) = \frac{k}{s+k}$, where $k$ is the cutoff frequency.
- Second-Order LPF: $G_{LPF}^2(s) = \frac{\omega_n^2}{s^2 + 2\zeta\omega_n s + \omega_n^2}$.
- Third-Order LPF: A cascade of a second-order and a first-order filter, $G_{LPF}^3(s) = G_{LPF}^2(s) \cdot \frac{k}{s+k}$.
Substituting these into the transfer function and evaluating the frequency response provides clear insights. While a first-order LPF provides basic decoupling, its harmonic attenuation roll-off is shallow (-20 dB/decade). This makes the extracted $v_d^{**+}$ signal susceptible to residual ripple if the grid voltage contains additional harmonics beyond the fundamental negative-sequence, a common scenario in real networks. The second-order filter offers steeper attenuation (-40 dB/decade), significantly improving harmonic rejection. The third-order LPF achieves an even steeper roll-off (-60 dB/decade), providing superior suppression of high-order harmonic interference that might couple into the PLL’s operation through the decoupling network.
| LPF Order | Transfer Function $G_{LPF}(s)$ | Attenuation Rate | Key Advantage for Grid Connected Inverter | Potential Drawback |
|---|---|---|---|---|
| 1st | $\frac{k}{s+k}$ | -20 dB/dec | Simple, minimal phase lag. | Poor harmonic rejection, high ripple in $v_d^{**+}$. |
| 2nd | $\frac{\omega_n^2}{s^2+2\zeta\omega_n s+\omega_n^2}$ | -40 dB/dec | Good balance between rejection and dynamic response. | May allow significant higher-order harmonic interference. |
| 3rd | $\frac{\omega_n^2}{s^2+2\zeta\omega_n s+\omega_n^2} \cdot \frac{k}{s+k}$ | -60 dB/dec | Excellent harmonic immunity, clean DC extraction. | Increased phase lag, slightly more complex tuning. |
The following analysis shows the response of $v_d^{**+}$ when the grid voltage contains multiple harmonic distortions (e.g., -5th, 7th, -11th). The first-order LPF fails to adequately suppress these, leaving substantial ripple. The second-order LPF performs much better, reducing ripple to acceptable levels. The third-order LPF, however, delivers an exceptionally clean DC signal, with residual ripple amplitude below 0.5%. This level of precision is vital for the inner current control loops of a high-performance grid connected inverter, ensuring accurate power delivery and low total harmonic distortion (THD) in the injected current.
Fast-Acting Current Limiter for Dynamic Protection
Beyond accurate synchronization, a grid connected inverter must adhere to strict grid codes regarding fault ride-through and current limiting. During voltage sags, the inverter may need to inject reactive current, which, combined with the remaining voltage, can cause the output current to exceed its safe maximum. Traditional current limiting methods often rely on calculating the Root Mean Square (RMS) value of the current. However, RMS calculation inherently involves averaging over a window (typically one cycle), introducing a delay of at least half a cycle (~10 ms for 50 Hz) before the limit can be applied. This delay is unacceptable for protecting semiconductor devices, which can be damaged within microseconds of an overcurrent condition.
A superior solution is a fast-acting current limiter based on instantaneous amplitude detection. Consider a sinusoidal current signal in the αβ-frame:
$$ i_\alpha(t) = I_m \cos(\omega t + \varphi), \quad i_\beta(t) = I_m \sin(\omega t + \varphi) $$
The instantaneous amplitude $I_m(t)$ can be computed directly without delay (aside from computational derivative delay) using the relation:
$$ I_m(t) = \sqrt{ i_\alpha^2(t) + i_\beta^2(t) } $$
For a pure sine wave in the αβ-frame, this calculation yields a constant DC value equal to the peak amplitude. In practice, with the positive and negative sequence currents properly separated by the enhanced DDSRF-PLL, this calculation can be applied to each sequence component independently. The limiter works by comparing this instantaneous amplitude $I_m(t)$ against a predefined threshold $I_{max}$. If the threshold is exceeded, the reference current magnitude generated by the inverter’s power/current controller is instantly saturated at $I_{max}$, while preserving its phase angle to meet grid support requirements. This entire process can be executed within a few control sample periods (e.g., < 100 µs), offering a dramatic speed improvement over RMS-based methods.
The comparison is stark: a traditional RMS-based limiter may take 10-20 ms to react to a sudden overcurrent, whereas the instantaneous amplitude-based limiter reacts within the sub-millisecond range. This rapid response is critical for the reliable operation of a grid connected inverter under stringent dynamic grid conditions, preventing nuisance trips and hardware damage.
Integrated Control Strategy and Simulation Validation
The proposed integrated control strategy for a grid connected inverter combines the third-order DDSRF-PLL for robust synchronization and a fast-acting current limiter for dynamic protection. The overall control block diagram for the inverter system incorporates these elements within a standard vector control structure. The PLL provides the accurate grid angle $\hat{\theta}$ and the separated positive-sequence voltage components ($v_d^{**+}, v_q^{**+}$). The outer power loop generates reference currents in the positive-sequence $dq^+$ frame. These references are then passed through the instantaneous amplitude limiter. Finally, the limited current references are regulated by high-bandwidth PI controllers in their respective synchronous frames, with feedforward and decoupling terms, to produce the voltage commands for the Pulse Width Modulation (PWM) stage.
To validate the effectiveness of this composite strategy, a detailed time-domain simulation model of a grid connected inverter system was developed. The test scenario was designed to stress both the synchronization and current limiting features:
- 0.0 – 0.25 s: Balanced grid, normal operation at 10 kW.
- 0.25 – 0.50 s: Unbalanced grid voltage (e.g., 30% sag in one phase). At t=0.3 s, the power command steps to 15 kW. At t=0.4 s, it steps back to 10 kW.
- 0.50 – 0.70 s: Grid returns to balanced condition.
The simulation results demonstrate the system’s robust performance. During the unbalanced and transient power periods, the enhanced DDSRF-PLL maintains a stable and accurate lock, with the extracted $v_d^{**+}$ showing minimal 2ω ripple. The three-phase grid currents remain sinusoidal and balanced despite the unbalanced voltage, with a THD consistently below 1.5%. The power response tracks its reference swiftly, with a settling time of under 20 ms for the 5 kW step change. The fast current limiter was also tested by artificially creating a severe sag condition that would demand a current beyond the rated limit. The limiter acted instantaneously, capping the current peak at the set threshold without introducing waveform distortion or instability, thereby validating its protective function for the grid connected inverter.
Conclusion
This analysis presents a comprehensive solution for enhancing the performance and resilience of grid connected inverters in unbalanced and distorted grid environments. The proposed integration of a third-order Decoupled Double Synchronous Reference Frame Phase-Locked Loop (DDSRF-PLL) with a fast-acting instantaneous current limiter addresses two critical challenges: precise synchronization under adverse grid conditions and rapid dynamic protection against overcurrent faults. The complex variable modeling confirms the superior harmonic rejection capability of the third-order LPF within the DDSRF structure, leading to cleaner sequence separation and more accurate phase tracking. The novel current limiter eliminates the inherent delay of traditional RMS-based methods, providing sub-cycle protection that is essential for meeting modern grid codes and ensuring hardware safety. Simulation results validate the combined strategy, showing excellent performance in maintaining power quality, dynamic response, and system stability. This approach provides a reliable and advanced technical framework for the next generation of high-penetration renewable energy systems, where the grid connected inverter must operate as a stable and intelligent grid-forming or grid-supporting asset.