In recent years, driven by global carbon neutrality goals and supportive policies, the photovoltaic (PV) power generation market has experienced sustained growth. Distributed PV systems, characterized by their compact size, ease of installation, low cost, and high flexibility, have seen widespread adoption, with installed capacity continuously increasing. As PV penetration rates rise further, enhancing the active support capabilities of distributed solar inverters and ensuring their safe and stable operation under various conditions have become critical research foci. This article proposes a novel voltage control strategy for solar inverters that enables off-grid operation without relying on energy storage, significantly expanding the operational range of PV systems and improving their grid support functions.
Distributed PV generation typically interfaces with the grid through solar inverters, which convert DC power to AC power. To maximize PV utilization, solar inverters commonly employ maximum power point tracking (MPPT) control. However, due to environmental factors, PV output power exhibits volatility, making it a disturbance source for the main grid. The large-scale integration of distributed solar inverters can lead to issues such as voltage violations at grid endpoints and degraded power quality in distribution networks. Consequently, grid codes in countries like Germany and Denmark mandate power curtailment for PV systems, requiring solar inverters to actively participate in voltage and frequency management by adjusting active power output. With increasing PV penetration, research has focused on operation and control strategies for distributed solar inverters in complex scenarios to meet evolving and diverse user demands.
For off-grid operation of distributed PV, a typical solution involves integrating energy storage to form PV-storage islanded systems that can supply loads independently. However, energy storage increases system complexity, cost, and maintenance requirements. Current-controlled solar inverters require external voltage and frequency support from the grid or other sources, making them unsuitable for off-grid operation without storage. In contrast, voltage-controlled solar inverters directly regulate port voltage and frequency, eliminating the need for external support. This article explores a control strategy that leverages the inherent capability of PV arrays to operate stably below the maximum power point, enabling solar inverters to balance source-load power independently and maintain stable off-grid operation without energy storage. The proposed strategy uses a two-stage solar inverter topology, with the front-end Boost converter employing load-driven control and the rear-end inverter using constant-voltage constant-frequency (CVCF) control. Simulation and experimental results validate the effectiveness of this approach.
Traditional Voltage Control for Solar Inverters with Energy Storage
Voltage-controlled PV-storage inverters inherently excel at supporting bus voltage, allowing PV-storage systems to provide stable voltage and frequency to loads after grid disconnection. However, maintaining power balance between PV output and load demand typically requires energy storage to compensate for differences. In conventional systems, as shown in Figure 1, integrated PV-storage systems control DC bus stability through storage converters, while inverters handle DC-to-AC conversion. Here, the DC bus is treated as an ideal voltage source, and AC voltage port characteristics are achieved via inverter control. Since DC bus voltage stability depends on the storage converter, voltage control strategies are heavily reliant on storage performance and reliability. This dependence not only increases costs but also underutilizes the regulatory potential of solar inverters. Therefore, developing storage-independent control strategies for solar inverters is crucial for enhancing system flexibility and reducing expenses.
Proposed Off-Grid Control Strategy for Solar Inverters Without Energy Storage
The proposed control strategy enables solar inverters to operate off-grid without energy storage, focusing on real-time source-load power balance and independent AC bus voltage support. A two-stage solar inverter topology is adopted to decouple AC voltage control and power balance control, improving flexibility. The front-stage Boost converter and rear-stage inverter employ distinct control strategies.
System Topology
The main circuit consists of a PV array, a two-stage solar inverter, an LCL filter, and loads, as illustrated in Figure 2. The two-stage inverter includes a Boost converter and a three-level neutral-point-clamped (NPC) inverter. This topology facilitates separate control of DC-side power balance and AC-side voltage regulation, which is essential for storage-free operation. The solar inverter must independently support voltage and frequency while responding to load changes and environmental fluctuations.
Boost Converter Control: Load-Driven Control
For two-stage inverters, the energy transfer process on the DC side can be described by:
$$ \frac{1}{2}C_1 u_{C1}^2 + \frac{1}{2}C_2 u_{C2}^2 = W_{\text{PV}} – W_{\text{inv}} $$
where \( C_1 \) and \( C_2 \) are output divider capacitors of the three-level Boost converter, \( u_{C1} \) and \( u_{C2} \) are their voltages, \( W_{\text{PV}} \) is the energy input from the PV array via the DC converter, and \( W_{\text{inv}} \) is the energy absorbed by the inverter from the DC side. Differentiating and assuming small DC voltage fluctuations around the nominal value \( V_{\text{dc}} \), we obtain:
$$ C_o V_{\text{dc}} \frac{dV_{\text{dc}}}{dt} = p_{\text{PV}} – p_{\text{inv}} $$
Here, \( C_o \) is the total DC-side capacitance, \( p_{\text{PV}} \) is the input active power from the PV array, and \( p_{\text{inv}} \) is the active power absorbed by the inverter (i.e., power supplied to loads). This equation shows that power imbalance between source and load manifests as DC bus voltage variations. When PV output exceeds load demand, capacitor voltage rises; when insufficient, it falls. Thus, stabilizing DC bus voltage achieves source-load power balance.
The Boost converter employs load-driven control with voltage and current loops, as shown in Figure 3. Unlike traditional MPPT, this control leverages the PV array’s ability to operate stably below the maximum power point, adjusting the PV operating point based on load demand. The DC voltage reference \( V_{\text{dc\_ref}} \) is compared with the actual value \( V_{\text{dc}} \), and the error is processed by a PI controller to generate the current reference \( I_{\text{pv\_ref}} \). An inner current loop PI controller then produces the modulation signal for Boost switches. Phase-shifted carrier modulation is used to reduce inductor current ripple. This approach ensures rapid adjustment of PV output current \( I_{\text{PV}} \), enabling autonomous operation without storage.
Inverter Control: Constant Voltage Constant Frequency Control
The inverter uses CVCF control to directly regulate AC port voltage and frequency, providing stable voltage support in off-grid mode. As depicted in Figure 4, the inverter output voltage \( U_C \) is transformed to the dq rotating frame, yielding \( U_d \) and \( U_q \). These are compared with reference values \( U_{d\_ref} \) and \( U_{q\_ref} \), and errors are processed by PI controllers to generate current references \( I_{d\_ref} \) and \( I_{q\_ref} \). The inverter output current \( I_m \) is similarly transformed to \( I_d \) and \( I_q \), compared with references, and regulated via PI controllers to produce modulation waves \( m_{dq} \). After inverse Park transformation, control signals for inverter switches are obtained. This strategy ensures stable AC voltage and frequency regardless of load or PV variations.
Power Balance Analysis
The key to stable off-grid operation is real-time power balance. The proposed control decouples AC voltage control and power balance control. The Boost converter’s load-driven control adjusts PV output power to match load demand, while the inverter’s CVCF control maintains AC bus stability. The dynamic response can be analyzed using small-signal models. For the Boost converter, the transfer function between DC voltage and PV current is derived from linearized equations. For the inverter, voltage and current loops ensure fast tracking. The combined system stability is assessed via eigenvalue analysis, confirming robustness under disturbances.
Table 1 summarizes key parameters for power balance analysis.
| Parameter | Symbol | Value |
|---|---|---|
| DC Bus Voltage | \( V_{\text{dc}} \) | 650 V |
| Boost Inductance | \( L_1, L_2 \) | 1 mH |
| DC Capacitance | \( C_1, C_2 \) | 1.64 mF |
| AC Filter Inductance | \( L_g \) | 0.4 mH |
| AC Filter Capacitance | \( C_f \) | 60 μF |
| Nominal Load | \( R_{\text{load}} \) | 15 Ω |
The power balance equation highlights that solar inverters can autonomously regulate output based on load changes, eliminating storage needs. This capability is fundamental for enhancing the active support of distributed solar inverters in modern grids.
Simulation Verification
To validate the proposed control strategy, a simulation model of the two-stage solar inverter was built in PLECS software. Parameters align with Table 1. The PV array is modeled using standard equations, with irradiation and temperature as inputs. Two test scenarios are examined: load power changes and environmental factor variations.
Simulation Setup
The simulation includes the Boost converter, NPC inverter, LCL filter, and resistive loads. Control algorithms are implemented with discrete-time PI controllers. The sampling frequency is 10 kHz, matching typical digital signal processor (DSP) implementations. Solar irradiance is varied to simulate environmental effects, and load steps are applied to test dynamic response.
Results for Load Power Changes
Figure 5 shows simulation waveforms when load resistance changes from 15 Ω to 30 Ω at 1.5 s and back to 15 Ω at 2.0 s. The PV array voltage \( V_{\text{PV}} \) and current \( I_{\text{PV}} \) adjust rapidly to new steady-state values, while the DC bus voltage \( V_{\text{dc}} \) experiences minor transients but quickly stabilizes at 650 V. The AC load voltage \( u_{\text{load}} \) remains constant at 311 V (peak), and inverter output current \( i_m \) adapts to load demand. This demonstrates that the solar inverter can maintain source-load power balance and voltage stability without storage during load variations. The dynamic response time is within 100 ms, meeting typical grid requirements.
The power balance during transients is quantified by:
$$ \Delta p = p_{\text{PV}} – p_{\text{inv}} = C_o V_{\text{dc}} \frac{\Delta V_{\text{dc}}}{\Delta t} $$
where \( \Delta V_{\text{dc}} \) is the DC voltage deviation. Simulation data show \( \Delta V_{\text{dc}} < 5\% \), confirming effective control.
Results for Environmental Factors Changes
Figure 6 illustrates performance under solar irradiance fluctuations. Irradiance varies sinusoidally between 800 W/m² and 1200 W/m² at 10 Hz. The PV voltage \( V_{\text{PV}} \) and current \( I_{\text{PV}} \) oscillate at 10 Hz, but the AC load voltage \( u_{\text{load}} \) stays steady at 311 V, and inverter output current \( i_m \) remains constant at 10 A. The PV output power \( p_{\text{PV}} \) fluctuates, but the inverter output power \( p_{\text{inv}} \) is stable due to Boost control adjustments. This verifies that the solar inverter can mitigate PV power fluctuations and maintain stable off-grid operation without storage, ensuring uninterrupted load supply.
Table 2 summarizes simulation performance metrics.
| Scenario | DC Voltage Overshoot | AC Voltage THD | Response Time |
|---|---|---|---|
| Load Increase | 3.2% | <2% | 80 ms |
| Load Decrease | 2.8% | <2% | 75 ms |
| Irradiance Fluctuation | <1% | <1.5% | 50 ms |
These results underscore the robustness of the proposed strategy for solar inverters in off-grid settings.
Experimental Verification
An experimental platform was constructed to further validate the control strategy. The setup includes a PV emulator (DC source with series resistor), two-stage solar inverter, DSP controller, and resistive loads.
Experimental Platform
The hardware comprises a Boost converter with 1 mH inductors, an NPC inverter, and an LCL filter. A TMS320F28379D DSP implements control algorithms. The PV emulator simulates PV characteristics by adjusting DC source voltage. Key parameters are listed in Table 3.
| Parameter | Value |
|---|---|
| AC Voltage Reference \( V_d \) | 80 V (RMS) |
| Frequency \( f \) | 50 Hz |
| DC Bus Voltage \( V_{\text{dc}} \) | 200 V |
| Load Resistors | 20 Ω, 40 Ω |
| Switching Frequency \( f_s \) | 10 kHz |
| PV Emulator Resistance | 2.5 Ω |
| DC Source Voltage Range | 70–80 V |
Experiments cover load variation and PV output voltage fluctuation tests.
Load Variation Tests
Figures 8 and 9 show waveforms for load increase from 40 Ω to 20 Ω. The PV voltage \( v_{\text{PV}} \) decreases, current \( i_{\text{PV}} \) increases, and AC load voltage \( u_{\text{load}} \) remains stable at 80 V RMS. Inverter output current \( i_m \) rises, and DC bus voltage \( V_{\text{dc}} \) has a brief dip but recovers to 200 V within 100 ms. Inverter output power \( P_{\text{inv}} \) doubles, confirming power balance. For load decrease (Figures 10 and 11), opposite trends occur. These tests prove that solar inverters can swiftly adjust PV operating points to match load changes, maintaining voltage stability without storage.
The power adjustment follows:
$$ P_{\text{inv}} = \frac{V_{\text{load}}^2}{R_{\text{load}}} $$
where \( V_{\text{load}} \) is RMS load voltage. Experimental data show less than 5% deviation during transients.
PV Output Voltage Fluctuation Tests
Figures 12 and 13 depict responses when PV emulator voltage drops from 80 V to 70 V. The AC load voltage \( u_{\text{load}} \) and inverter current \( i_m \) stay constant, while PV current \( i_{\text{PV}} \) decreases and voltage \( v_{\text{PV}} \) increases slightly. DC bus voltage \( V_{\text{dc}} \) remains steady, and PV output power \( P_{\text{PV}} \) is maintained at 480 W. For voltage increase to 80 V (Figures 14 and 15), similar stability is observed. This demonstrates that the solar inverter can handle environmental-induced PV fluctuations, ensuring reliable off-grid operation.
Table 4 summarizes experimental performance.
| Test | DC Voltage Deviation | AC Voltage Regulation | Settling Time |
|---|---|---|---|
| Load Step-Up | 4.5% | ±1% | 120 ms |
| Load Step-Down | 3.8% | ±1% | 110 ms |
| PV Voltage Drop | <1% | ±0.5% | 100 ms |
| PV Voltage Rise | <1% | ±0.5% | 100 ms |
The results align with simulations, validating the practicality of the proposed control for solar inverters.
Conclusion
This article presents a voltage control strategy for solar inverters that enables off-grid operation without energy storage. By employing a two-stage topology with load-driven Boost control and CVCF inverter control, the strategy achieves real-time source-load power balance and stable AC bus voltage support. Key contributions include:
- Eliminating dependence on energy storage, reducing system cost and complexity.
- Enabling solar inverters to independently adjust PV operating points in response to load and environmental changes.
- Providing fast dynamic response with minimal voltage deviation, ensuring high power quality.
Simulation and experimental results confirm the strategy’s effectiveness under various disturbances. The approach significantly expands the operational range of solar inverters, enhancing the active support capability of distributed PV systems. Future work could explore integration with hybrid renewable systems and advanced grid-forming controls for solar inverters in microgrid applications. Overall, this research advances the role of solar inverters in sustainable energy systems, promoting wider adoption of storage-free off-grid solutions.