Enhancing Grid Stability and Efficiency: Cost-Effective Hardware Implementation for Advanced Control of Grid-Connected PV Systems

Electricity has become a vital product for humanity. It is produced by large power plants that transform primary energy into electrical energy, which is then transported and distributed to consumers. Like fire in prehistoric times, electricity has changed human life. The vital impact of electricity on our daily lives is particularly noticeable during power supply interruptions [1]. Moreover, uncontrolled power outages can lead to significant societal and economic threats. There are notable examples of outages that have occurred in recent years around the world and their impacts in terms of the number of people affected: in September 2003, a power outage affected all of Italy (except two of its islands) for twelve hours; in the same year, the power outage in the northeastern United States and southeastern Canada left forty million people in the dark; in September 2005, the outage in Java-Bali, Indonesia, affected over one hundred million people; while in November 2009, the power outage in Brazil and Paraguay affected about ninety million people.

For these reasons, over the past few decades, the demand for economical, clean, and renewable energy has increased considerably and become a crucial necessity for sustainable development [2]. According to the International Energy Agency (IEA), global renewable energy capacity reached 3,064 gigawatts (GW) in 2021, with an annual increase of 9% from the previous year, highlighting the growing commitment to renewable sources [3]. This surge in renewable energy adoption is partly due to the depletion of industrial fuels, which has prompted a shift towards sustainable alternatives. Additionally, the urgency to reduce greenhouse gas emissions has become more critical; the Global Carbon Project reported that carbon dioxide emissions from fossil fuels rose to 36.4 billion metric tons in 2021 [4], emphasizing the need for cleaner energy sources [5].

1.1 Paper motivation

Among the various renewable energy sources (RES) available, solar energy stands out as a major contender due to its abundance and pollution-free conversion into electricity [6], and its relatively decreasing cost [7]. According to the International Renewable Energy Agency (IRENA), the global average cost of utility-scale solar photovoltaics (PV) fell by 89% between 2010 and 2021, making it one of the most competitive energy sources [8]. In recent years, the demand for solar energy sources has increased steadily due to several factors, namely the growing efficiency of solar cells and improvements in manufacturing technology. In 2021, the total installed solar capacity worldwide reached 1,000 GW, with a record 168 GW of new capacity added that year alone, highlighting the ongoing trend of connecting PV modules to the distribution grid in many countries [9]. This trend is expected to continue, with projections indicating that global solar capacity could exceed 5,000 GW by 2030. For example, Morocco has proposed action plans in which it is committed to taking full advantage of its triple strategy to adapt to climate change. The main axis of its strategy is that by 2030, the country should cover 50% of its electricity needs from RES [10]. This is a sign of the importance of implementing renewable energies.

PV solar systems are generally used in two main areas:

• Off-grid PV systems: in these systems, distant loads without grid connectivity are powered by solar panels.

• Grid-connected PV systems: with these systems, energy from solar panels is used to power both the electrical grid and nearby loads. This system can be put in parking lot shade structures, building facades, and rooftops, etc. [11].

Given the various GPV projects recently installed in different countries around the world, we can see that GPV systems currently dominate the PV market, particularly in the United States, Europe, and Japan. Despite the importance of this field, GPV systems have not been widely developed in due to technical, economic, and regulatory obstacles, namely:

• Maximum Power Point Tracking (MPPT): Maximizing the power output of PV modules requires efficient MPPT algorithms and technologies, which can be challenging to implement effectively [12-14].

• Voltage Dip Sensitivity: PV systems may be sensitive to voltage dips or fluctuations in the grid [15], which can affect their performance and reliability [16].

• Grid Synchronization: Ensuring that PV systems synchronize properly with the electrical grid is essential for safe and efficient operation. This requires advanced control and monitoring systems [17, 18].

• Cost and Component Density: The cost of PV components, such as inverters, batteries, and mounting structures, can impact the overall affordability and feasibility of PV installations. Additionally, the density of these components may pose challenges in terms of space requirements and installation logistics [19].

It is important to recognize that the increasing interest in PV systems necessitates expanded research and development in areas such as MPPT, grid synchronization, power quality, electronic interfaces, and control mechanisms.

Motivated by the significant demand for PV solar energy and the immediate need to improve GPV systems, the research presented in this paper focuses on the study, design, control, and implementation of a PV energy conversion system, operating at around 1.5 kW of nominal power.

1.2 Paper contribution

The integration of renewable energy sources (RES) into the electrical grid has seen significant growth, driven by global efforts to reduce dependency on fossil fuels and lower greenhouse gas emissions. However, many of these advancements remain at the simulation stage, with limited real-world implementation. For instance, Gao et al. [20] developed a simulation model to address common-mode leakage currents in non-isolated PV grid-connected inverters, using an improved double-vector model predictive current control (IDV-MPCC). While the model successfully demonstrated the suppression of common-mode voltage (CMV) and reduced harmonic distortion, its validation remains theoretical, without physical experimentation.

In another example, Rehan [21] utilized HOMER software to simulate a hybrid renewable energy system incorporating PV, biomass, and hydropower for grid-connected applications. Despite showing promising results in terms of economic feasibility and carbon emission reduction, the study is purely simulation-based and lacks real-world testing. Similarly, Dunna et al. [22] conducted a simulation using the HOMER Pro software to design a hybrid PV/Diesel/Generator system for Douala, Cameroon. The analysis suggested cost savings and improved system reliability, but challenges such as temperature variability and regulatory hurdles highlight the need for practical implementation to fully validate these findings

Most of these research efforts are heavily reliant on simulations using tools like MATLAB/Simulink [23, 24], which facilitate implementation in devices such as dSPACE (Digital Signal Processing and Control Engineering) [25], DSP (Digital Signal Processing) [26, 27], and FPGA (Field-Programmable Gate Array) [28, 29] much easier. However, these platforms are often costly, complex, and bulky, with some requiring a dedicated computer for the execution of control algorithms. This contributes to the high cost and logistical challenges of implementing grid-connected PV (GPV) systems, making them less accessible for widespread deployment.

The aim of this research is to develop a complete laboratory-scale test prototype of a low-power PV system connected to the low-voltage distribution grid, aiming to overcome the challenges mentioned above. This system could be beneficial for future researchers in the field of renewable energies to validate their research work. The proposed energy conversion system offers a promising and cost-effective approach, yielding significant and efficient results for use in GPV projects.

It is worth mentioning that designing and implementing a power system necessitates expertise in multiple areas of electrical and computer engineering. This includes systems theory, circuit analysis, board programming, power electronics, and software development, and feedback control systems.

1.3 Paper outline

This work is structured as follows: In Section 2, Several research studies were examined to uncover innovative contributions in the field. Section 3 outlines the theoretical concepts that must be taken into account and describes the methodology that facilitated the successful implementation of the GPV system. Section 4 is dedicated to the complete implementation of the grid-connected three-phase inverter, including the various control algorithms and components used. The experimental results are detailed in Section 5. Lastly, Section 6 provides the paper's conclusion and discusses future research directions.

The main requirements for GPV systems are to convert the DC voltage generated by solar panels into AC voltage suitable for grid connection. This involves using DC-DC converters to increase the DC voltage before converting it to AC with inverters. The output voltage and frequency of the inverter must match the grid's, a synchronization algorithm is used for this purpose. The control system must adapt to weather conditions, utilizing MPPT algorithm to maximize power extraction. The DC-DC converter focuses on voltage increase and MPPT, while the inverter ensures proper power injection and synchronization with the grid [40]. In addition to the three-phase inverter, the LCL filter is used to minimize the high THD value. The GPV system block diagram is depicted in Figure 1. Therefore, the proper selection of the components parameters is very important, including:

1.png

Figure 1. Grid-connected PV system diagram

3.1 Solar panel and MPPT control

Solar energy from photovoltaic (PV) panels is expected to be a major player in our future energy mix. The cost of PV modules has dropped significantly in the last decade, leading to the widespread establishment of large-scale PV power plants in many countries. Despite these advancements, the cost per unit of PV energy remains higher than that of fossil fuels. This has spurred intensive research efforts aimed at boosting the efficiency of PV systems, particularly by enhancing the performance of solar cells.

One effective strategy for improving PV system efficiency involves optimizing the components that make up the system, especially the power converter. A promising area for improvement is the development of more efficient MPPT algorithms. These algorithms, implemented through software in the power converter's firmware, can enhance performance without requiring additional hardware investment. They work by dynamically adjusting the operating voltage or current to maximize power output based on real-time environmental conditions like sunlight intensity and temperature fluctuations. However, this task is complex because environmental factors are constantly changing, and the relationship between PV panel output and environmental conditions is nonlinear. As a result, achieving optimal MPPT performance remains a significant challenge in the quest to maximize the efficiency of PV systems.

The proposal for an improved P&O MPPT algorithm aimed at eliminating power oscillations and ensuring that the operating point does not deviate from its tracking location was previously published [41, 42]. For more details, please refer to it.

3.2 Three-phase grid-connected inverter

3.2.1 Theory of the three-phase inverter

The Voltage Source Inverter (VSI) converts DC to AC voltage and is classified into single-phase and three-phase types. This paper focuses on three-phase VSIs, used for medium to high-power applications to provide a balanced source of three-phase voltages. Figure 2 shows that the inverter consists of three switching arms, each with two cells containing a diode and a transistor, operating in forced commutation.

In controllable mode, the inverter arm acts like a two-position switch, allowing two different voltage levels at the output. It's crucial that the switches in any inverter arm (S1 and S4, S2 and S5, or S3 and S6) are not activated at the same time, as this would cause a short circuit in the power supply. Similarly, both switches in an inverter arm should not be turned off at the same time. Table 1 shows the eight valid switching states for a standard three-phase VSI setup [43].

2.png

Figure 2. Three-phase inverter topology

3.2.2 Mathematical modeling of the three-phase inverter

The electrical behavior of the inverter in the stationary reference frame (abc) can be expressed as follows [36]:

$\left[\begin{array}{l}u_a \\ u_b \\ u_c\end{array}\right]=R\left[\begin{array}{l}I_a \\ I_b \\ I_c\end{array}\right]+L \frac{d}{d t}\left[\begin{array}{l}I_a \\ I_b \\ I_c\end{array}\right]+\left[\begin{array}{l}v_a \\ v_b \\ v_c\end{array}\right]$  (1)

where, $v_{a b c}$ are the output voltages of the inverter connected to the grid, $i_{a b c}$, are the three-phase grid currents, are the threephase grid voltages, $L$ and $R$ are respectively, the inductor and the resistance, between the grid and the inverter connected to the grid. Using the Park transformation, we can construct Eq. (1) to simplify the model of the three-phase inverter. Eq. (2) provides the inverter output voltages in the synchronous reference frame $(d q)$.

$\left[\begin{array}{l}v_d \\ v_q\end{array}\right]=L \frac{d}{d t}\left[\begin{array}{l}I_d \\ I_q\end{array}\right]+R\left[\begin{array}{l}I_d \\ I_q\end{array}\right]+\omega L\left[\begin{array}{c}-I_d \\ I_q\end{array}\right]+\left[\begin{array}{c}u \\ u_q\end{array}\right]$    (2)

3.2.3 Three-phase grid-connected inverter control strategy

The operation of a three-phase inverter relies on Pulse Width Modulation (PWM) algorithms to control switch activation, thereby generating the desired output voltage. Over the years, extensive research has led to the development of various PWM techniques, each tailored to specific machine types, power levels, and semiconductor components in power converters. While sinusoidal PWM has been historically prominent, the emergence of microprocessors has popularized Space Vector Modulation (SVM) as a fundamental approach in modern three-phase power conversion systems [44].

SVM offers numerous advantages over other techniques. This approach is based on approximating an overall control vector throughout a modulation period TPWM instead of doing separate calculations for each arm of the inverter. For a two-level three-phase inverter, there are eight distinct states, each determining a space vector. The collection of all vectors forms a hexagonal axis and divides the entire space into six sectors, numbered from 1 to 6, as depicted in Figure 3. A reference voltage vector rotates within the space vector diagram. For any given position, the reference vector can be approximated by three nearest stationary vectors, allowing the generation of gating signals for the inverter switches. Consequently, as the reference vector completes one rotation, the inverter output voltage undergoes one fundamental cycle. The subsequent sections will explain the general principles and the method for calculating the various parameters of the SVPWM [45]:

3.png

Figure 3. Space vector hexagon

a. General principle

Any two consecutive non-zero vectors have an angle of $60^{\circ}$. SVM is a method of digital modulation aimed at finding a combination of active and zero vectors to determinate a given reference voltage. In SVM, the three-phase voltages are mapped into the orthogonal two-phase complex reference frame $\alpha \beta$. This is known as the Clarke Transformation, as presented in Eq. (3). Then, the reference voltage (Vref) is used to calculate the angular frequency $\theta$ by the following equations.

$\left[\begin{array}{l}V_\alpha \\ V_\beta\end{array}\right]=\left[\begin{array}{ccc}1 & \frac{-1}{2} & \frac{-1}{2} \\ 0 & \frac{\sqrt{3}}{2} & \frac{-\sqrt{3}}{2}\end{array}\right]\left[\begin{array}{l}v_a \\ v_b \\ v_c\end{array}\right]$     (3)

hence,

$V_{r e f}=\sqrt{V_\alpha^2+V_\beta^2}$  （4）

$\theta=\tan ^{-1} \frac{V_\beta}{V_\alpha}$    (5)

b. Reference space-vector locating

The second part of the SVPWM calculation involves identifying the sectors, and region to which the reference vector belongs. The sector number is identified according to the flowchart presented in Figure 4.

4.png

Figure 4. Sector identification

c. Switching time calculation

In this step, the calculation of the modulation index (m) as well as the dwell times (where n is the sector number ranging from 1 to 6) is performed according to Figure 4 and the following equations:

$m=\frac{V_{r e f}}{U_{D C / 2}}$    (6)

$\begin{gathered}T_1=\frac{\sqrt{3}}{2} T_{p w m} m \sin \left(\frac{\pi}{3}-\theta\right) \\ T_2=\frac{\sqrt{3}}{2} T_{p w m} m \sin (\theta) \\ T_0=T_{p w m}-T_1-T_2\end{gathered}$     (7)

where, m is the modulation index, UDC is the available DC voltage, T_pwm is the sampling period, and θ the grid phase angle extracted by the PLL algorithm.

d. Switching sequence generation

Once the dwell times are computed, the SVPWM pulses are generated. The sequence of switch activations is carefully arranged to minimize transitions between vectors, which effectively lowers the switching frequency and decreases harmonic distortions. In the initial sector, the sequence 01277210 is employed for symmetry considerations, as illustrated in Figure 5.

5.png

Figure 5. Switching signals identification

3.2.4 Synchronization technique

The output voltage of the inverter must be in sync with the grid voltage for PV systems that are linked to it. Finding the grid voltage's phase angle at the point of common coupling is the aim of the synchronization procedure. This helps in setting current or voltage references for active/reactive power control and harmonic current compensation. Detecting the grid phase angle is vital for controlling grid-connected inverters, and synchronization algorithms need to respond swiftly to changes in the electrical grid. They also need to filter out noise and harmonic components, especially when currents absorbed by the load are unbalanced, which can create a reverse voltage system.

Several synchronization algorithms are designed to extract the phase angle of the grid voltage, with the most common methods being zero-crossing detection, grid voltage filtering, and the phase-locked loop (PLL) technique. Zero-crossing detection, the simplest method, identifies when the grid voltage crosses zero, providing phase and frequency information. However, it has limitations such as a low dynamic response and vulnerability to noise and harmonics, making it unsuitable for applications requiring precise and continuous phase angle detection.

Grid voltage filtering is another method for detecting the phase angle but requires additional filtering to achieve accurate phase angle and frequency detection, especially with distorted network voltage. However, the use of filters can introduce a phase shift, rendering this technique unsuitable for grid-connected inverter applications.

Our research adopts the phase-locked loop method for synchronization. Widely applied in communication technologies, PLL ensures synchronized output signals with reference inputs in both frequency and phase. Its primary aim is to adjust inverter current frequency to maintain temporal alignment with grid voltage, even amidst noise or harmonic components. Compared to other methods, PLL offers reliable frequency detection. Its fundamental functional diagram, presented in Figure 6, comprises a phase detector (PD), loop filter (LF), and phase angle detector (PD), also known as a voltage-controlled oscillator (VCO) [46].

6.png

Figure 6. Phase locked loop diagram

7.png

Figure 7. DSOGI-PLL block diagram

There are several PLL synchronization algorithms in the literature. It is important to note that a comparison of the most commonly used PLL synchronization algorithms has already been conducted and published in a previous work. This published study evaluated several algorithms based on criteria such as the accuracy of phase angle detection, robustness against harmonic disturbances and noise, and the speed of dynamic response. For more details on the comparison methodology and the results obtained, please refer to this published work. This information provides a solid basis for selecting the most efficient algorithm, which is used in our GPV system prototype.

Research has shown that the most suitable algorithm for our prototype is the DSOGI-PLL. Its diagram is depicted in Figure 7. Its implementation using a programmable board will be presented in the following section.

3.3 Inverter control

Two primary techniques are used to control three-phase VSI inverters: voltage control and current control. To regulate power flow, a voltage-controlled inverter modifies the phase angle between the grid voltage and the inverter's output voltage. Conversely, a current-controlled VSI utilizes pulse width modulation (PWM) techniques to regulate the active and reactive components of the current delivered to the grid. Current control methods provide faster response times and are less influenced by voltage phase shifts and grid voltage distortions. In contrast, voltage control methods are more sensitive to minor phase errors, which can lead to significant harmonic currents. Therefore, current control is preferred for grid-connected inverter management [47].

The PI (Proportional-Integral) controller is the most popular technique for current control in inverters, effectively minimizing the error between the measured output and desired grid current. It employs proportional (Kp) and integral (Ki) constants to ensure an excellent steady-state response, stable switching frequency, and accurate harmonic control. PI controllers are robust and do not require a system model. They can operate in both stationary $(\alpha \beta)$ and synchronous (dq) reference frames, with the synchronous frame preferred for its ability to eliminate steady-state error, making it ideal for grid-connected three-phase inverters.

According to the detailed mathematical model of the inverter presented earlier, the inverter output voltages in the dq frame are given by Eq. (2) [48]. By rearranging the equation, we can find that:

$\left\{\begin{array}{c}L \frac{d i_d}{d t}+R i_d=v_d-u_d+\omega L i_q \\ L \frac{d i_q}{d t}+R i_q=v_q-u_q+\omega L i_d\end{array}\right\}$   (8)

therefore,

$\begin{aligned} \hat{V}_d & =v_d-u_d+\omega L i_q \\ \hat{V}_q & =v_q-u_q+\omega L i_d\end{aligned}$   (9)

PI-type current controllers can be used to guarantee that the output currents follow the reference currents. In this instance, the following formula yields the output voltages of the current controllers:

$\left\{\begin{array}{l}\hat{V}_d=k_p\left(i_d^*-i_d\right)+k_i \int\left(i_d^*-i_d\right) d t \\ \hat{V}_q=k_p\left(i_q^*-i_q\right)+k_i \int\left(i_q^*-i_q\right) d t\end{array}\right.$   (10)

The reference voltage of the d-axis and the q-axis are expressed as follows:

$\left\{\begin{array}{c}v_d^*=u_d-\omega L i_q+\widehat{V}_d \\ v_q^*=u_q+\omega L i_d+\widehat{\nabla}_q\end{array}\right\}$   (11)

As illustrated in Figure 8, the block diagram of the synchronous regulator for the grid-connected inverter is derived from Eq. (11).

8.png

It should mention that PI (Proportional-Integral) controllers are a standard approach in current control strategies due to their simplicity and effectiveness. However, they also have limitations, particularly in dealing with non-linearities and varying system dynamics, which can lead to performance issues such as overshoot and prolonged settling times. To enhance the control performance, future improvements could explore adaptive control techniques that adjust parameters in real-time, or intelligent control strategies, like fuzzy logic or model predictive control [49], which can better handle uncertainties and nonlinear behaviors in renewable energy systems.

3.4 LCL filter

To choose an optimal filter topology suitable for the inverter in our GPV conversion system, parameters such as efficiency, weight, and volume must be considered. The LCL filter was chosen for our design due to its superior attenuation of inverter switching harmonics compared to other topologies. The main advantages of the LCL filter, as highlighted [50]:

·Minimizing grid distortion and reactive power generation;

·Providing -60 dB/decade attenuation for frequencies above the resonance point;

·Allowing operation at a relatively low switching frequency while maintaining effective harmonic attenuation;

Its performance is not affected by variations in grid im·pedance.

·The single-phase electrical diagram of this filter, depicted in Figure 9, is considered for its modeling. The circuit and components are identical for the other two phases.

9.png

Figure 9. LCL single-phase electrical diagram

Using Kirchhoff's law, the equations of the filter can be written as follows:

$\begin{gathered}i_a-i_b-i_c=0 \\ v_i-v_c=i_a\left(s L_i+R_i\right) \\ v_c-v_g=i_b\left(s L_g+R_g\right) \\ v_c=i_c\left(\frac{1}{s C_f}\right)\end{gathered}$      (12)

According to research, the LCL filter produces a resonance peak at its cutoff frequency, potentially causing system instability. To address this, passive damping is added using a virtual resistance R_fv, which reduces the voltage across the capacitor proportionally to the current through it. This improves the filter's simplicity, reliability, and efficiency. Among the possible placements for the resistance (series/parallel with the inverter-side inductance or filter capacitor), the series connection with the filter capacitor was chosen.

The equations of the filter then become as following:

$\begin{gathered}i_a-i_b-i_c=0 \\ v_i-v_c=i_a\left(s L_i+R_i\right) \\ v_c-v_g=i_b\left(s L_g+R_g\right) \\ v_c=i_c\left(\frac{1}{s C_f}+R_{f v}\right)\end{gathered}$    (13)

The experimental version of the GPV system proposed in this research was implemented in our laboratory "Advanced Systems Engineering, National School of Applied Sciences Kenitra" (Figure 14 and Figure 15).

14.png

Figure 14. Experimental version of the proposed GPV system

15.png

Figure 15. Power section prototype of the studied system

Due to the constraints of the laboratory facilities, the experimental model has been scaled down but still meets the intended objectives. The hardware design is validated by testing the various project components, including the three-phase power circuit, the three-phase load, the control board, the measurement board, the interface board, and the filtering board. After verifying the proper functioning of the various components through testing, this section concludes with the presentation of experimental data for grid integration.

5.1 Three-phase voltage measurement

A measurement board is required to monitor the voltage of the three-phase grid and the current fed into the grid for control purposes. The signals shown in Figure16 are the grid voltages outputted from the voltage sensor. By observing the average and maximum voltage and frequency measurements taken by the oscilloscope, it is noted that they correspond to what is expected: a sine wave in the range of 0 to 3.3 V, indicating an average voltage of nearly 1.5 V, a maximum voltage of 3.3 V (for a voltage of 250 V), and a frequency of 50 Hz.

The implemented algorithm must use the actual instantaneous values of the grid parameters. Therefore, in the developed code, voltages ranging from 0 to 3.3 V are mathematically converted to the range of $-220 \sqrt{ } 2$ to $220 \sqrt{ } 2$ V, and calculations are performed with these values. If a certain signal needs to be visualized, it must be shifted to the range of 0 to 3.3 V within the code itself. Then, apply the desired signal to one of the output pins of the digital-to-analog converter (DAC) and connect this pin to a channel of the oscilloscope.

16.png

Figure 16. Grid voltages outputted from the voltage sensors

17.png

Figure 17. Temporal evolution of the PLL phase angle $\theta P L L$ and the positive sequence component $v \alpha+$

5.2 Synchronization

Figure17 depicts the temporal evolution of the PLL phase angle $\theta P L L$ and the positive sequence component $v \alpha+$. It is noteworthy that the frequency of both signals is approximately 50 Hz , as expected. It is also observed that the detected phase angle is synchronized with the voltage of $v \alpha+$ since the cycles of both signals start and end simultaneously. This proves that the phase angle calculated and generated by the proposed algorithm is appropriate with the phase of the input voltage; consequently, the estimated frequency is equal to the grid frequency.

5.3 SVPWM

Figure 18 provides an oscilloscope snapshot showing the temporal evolution of the phase angle and the calculated sector number.

18.png

Figure 18. Temporal evolution of the phase angle and the calculated sector number

Figure 19 and Figure 20 show a typical oscilloscope snapshot for the pulses generated after the IR2111 control circuit for driving the low-side transistors of the inverter. The pulses generated by the IR2111 control circuit are the same as the pulses generated by the Arduino board. However, their amplitude is higher.

19.png

Figure 19. Low-side transistors control signals

20.png

Figure 20. Transistors control signals

A 700 ns dead time has been incorporated between the switching pulses of each pair of complementary transistors. This interval, which is the period between turning off one transistor and turning on the other in the pair, is necessary to prevent short circuits at the DC bus level.

An example oscilloscope overview of the composite voltages of the experimental grid-connected inverter, captured just prior to the filter, is shown in Figure 21. The related simulated waveforms and the experimental waveforms are strikingly comparable. Figure 22 shows the grid voltage and the usual output voltage of the grid-connected inverter, which is obtained after the LCL filter. The output voltages of the inverter and the grid voltage are identical in frequency and phase, as the illustration makes abundantly evident.

21.png

Figure 21. Experimental waveforms of the line voltages at the output of the three-phase inverter

22.png

Figure 22. Grid-connected inverter voltage at the LCL filter output synchronized with grid voltage

While significant progress has been made in the field of renewable energy systems, much of the research to date remains theoretical, relying heavily on simulation-based studies, these simulations, though insightful, lack validation through real-world implementations. In contrast, our research focuses on the practical implementation of an advanced grid-connected PV system. The prototype developed in this study moves beyond theoretical analysis by offering a cost-effective, low-power hardware solution that can be tested under real conditions. Unlike simulation tools such as MATLAB/Simulink or platforms like dSPACE, which are commonly used in simulation-based studies, our hardware design relies on affordable components, making it a more viable option for large-scale deployment. By addressing the cost, size, and complexity challenges of conventional devices, this prototype paves the way for wider adoption of grid-connected PV systems.