Implementation of solar PV system unified ZSI-based dynamic voltage restorer with U-SOGI control scheme for power quality improvement

The main challenge in today's power system is to supply continuous, reliable power and satisfy the high demand. The incorporation of renewable energy sources into the utility grid system can be accomplished. However, the renewable sources are intermittent in nature and the loads work dynamically and cause imbalances to the system voltage within an immediate time. Intermittent renewable sources affect the voltage of the power grid system. Photovoltaic (PV) power generation with Z-source inverter (ZSI)-based dynamic voltage restorer (DVR) is used to avoid that. For step-up low DC voltage to required AC voltage for the compensation of the voltage disturbance, ZSI with the energy storage impedance network is used. DC-DC converters connect the PV cell and the battery storage to the impedance source network. This article also incorporates an upgraded second-order generalized integrator (U-SOGI) control system for the generation of reference voltage signals. The U-SOGI control reference voltage generation approach greatly improves system performance and decreases the harmonic voltage. The voltage-related problems in the system connected to the utility grid are mitigated with DVR. In different load and source conditions, the PV generation with DVR performance is verified by the digital simulation and experimental prototype.


Introduction
The electricity demand for loads has increased every day in India, due to technological advancement, population growth, and industry growth. The incorporation of renewable sources into the existing electrical energy system will meet demand. With the introduction of renewable sources incorporated into the present energy system, the deficit in demand is decreasing every year. Both isolated and grid-connected electrical networks, wind and solar power sources play a significant role. In contrast to other sources of renewable energy, PV incorporation plays a vital role in the electrical distribution system. Nevertheless, solar energy has an intermittent nature and affects both the grid and the reactive load [1]. Due to its intermittent existence, PV power does not provide a continuous power supply [2]. With the introduction of battery storage systems, the intermediation issue is avoided. The combination of these energy sources provides versatile energy and satisfies the necessary long-term peak demand. In the distribution system, this hybrid generation plays the main role in meeting the demand for load [3]. In the power distribution system, this PV power generation plays the main role in meeting the demand for load [4].
The PV system is fed by the power converter to the distribution grid to provide a continuous power supply [5][6][7]. Nevertheless, the simultaneous switching of non-linear loads disrupts the connected grid system. This can cause an imbalance of the line voltage, vibration, voltage falls, voltage swell, spike, grid failure and harmonics that can damage sensitive loads. Customized equipment, including DVR, Distribution Static Synchronous Compensator (DSTATCOM) and Unified Power Quality Conditioner (UPQC), used in the distribution system protects the critical loads [8]. DVR is the perfect solution to solving voltage-related problems in customer power devices as it is cost-effective. The cost-effective DVR is the most attractive among customer power devices and provides the solution for solving voltage-related problems [9,10]. The DVR is a static converter and is realized from the converter of the voltage source. A DVR is built in series through the interconnected series transformer with a power distribution line. The DVR restores voltage in unbalanced or voltage loss conditions in series with a supply line. In the application of Doubly Fed Induction Generator (DFIG)-based wind turbine in a fault ride-through capability, DVR with a storage system is implemented for mitigation of voltage sag and manages the DC link voltage [11]. To protect the sensitive load, DVR with a self-tuned fuzzy-PI controller is provided to maximize the power flow with improved power factor [12]. Voltage sag is compensated by using a DVR with second degrees of freedom of the resonant control scheme [13]. To secure the sensitive load, hybrid energy storage system based DVR is introduced with a sensitive load [14]. Transformerless DVR (TDVR) with a predictive based control algorithm is used through the buck-boost AC-AC converter to compensate for the voltage sag and swell [15]. DVR with fault current limiter is utilized for optimizing the power flow with power factor improvement to protect the sensitive load [16]. Using dual p-q theory, optimized power flow is done via DVR and a passive filter [17].
The DVR alone cannot sustain the load voltage continuously during peak demand. A primary supply of energy is required to meet the highest demand and to sustain the load voltage in the distribution system. With the incorporation of renewable resources, continuous demand for the load is met. The DVR is fed with PV systems connected to the grid, to prevent such power-conditioning issues in distribution networks. The DVR does not need a separate source to restore system voltage in the proposed PV generating system with a battery storage unit. The DVR input is integrated directly with the DC-link of the system being proposed.
The DVR with PV power generation holds the load voltage, DC-link voltage, and grid voltage stable. Compared to existing structures, the new design offers the following advantages. The DC-link is integrated with the load point and dynamic voltage restorer. To stabilize the DC-link voltage, the DC-link is integrated with the PV power generation. The PV system with battery supports the DVR and the load point via the DC-link for the load profile maintenance. DVR with the PV power generating system will provide the required active and reactive energy to the electric power distribution system. The DVR recovers the device voltage that is connected through the line in series. The DC power or external battery-powered DVR offers an efficient injection of the unbalanced voltage but the cost is high. The DVR should be supplied with sufficient DClink power, to protect the sensitive load, rather than with a separate source via a voltage source converter. The DC-link of DVR in this setup again relies upon the voltage of the power supply and cannot be stable to compensate for the line voltage. DVR may not protect the sensitive load under abnormal conditions. DC-link line interference can be prevented by integrating renewable energy with an energy storage system. Solar PV power with battery is used to supply the required DClink power and to supply a continuous load power with power quality enhancement.
In this article, the second section describes the ZSIbased DVR topology and DC-DC converter with solar photovoltaic. The next section deals with the control scheme for ZSI-based solar PV interconnected DVR. The final section discusses the simulation and experimental results of the proposed ZSI-based solar PV powered DVR.

Proposed ZSI-based DVR topology
DVR is a power electronic circuitry that is used to transform voltages from the DC battery/PV system to the AC system as a voltage source inverter (VSI). The boost transformer unit transmits the voltage to the distribution system from the DVR for the compensation of voltage based disturbances.
The primary winding is connected to the power system in series and the secondary winding with the DVR is connected. The harmonic filter consists of an LC circuit to minimize the waveform harmonic distortion generated by VSI. Figure 1 displays the proposed solar PV power generation with a dynamic voltage restorer. The solar PV system's main module comprises solar PV with boost converters, ZSI converter, battery storage unit with DC-DC converter, AC loads, and the power grid. The utility grid is fed to the load through the DVR. The DC link is interfaced with solar photovoltaic with DC-DC converter and battery storage unit. The critical loads are protected by DVR and by the solar PV generation connected with the grid. The voltage is also regulated at the DC-link and the power transmission efficiency in the distribution line is increased. The battery storage system stabilizes power during the intermittent nature of the solar photovoltaic. The solar photovoltaic system with the battery supports the DC-link when the maximum energy is needed in the daytime.

Solar photovoltaic with DC-DC converter
Solar energy is abundant, tremendously accessible in nature and less inertia. The solar PV system is based on the photoelectric effect theory, which transforms solar energy into electricity. Solar PV power is fed to the DC-link by boosting the DC voltage through the  DC-DC converter for the constant voltage of the DClink. Figure 2 shows solar PV with a Z-source network structure. A single diode model is the best approach: a current source, as seen in Figure 2, in parallel with a diode, where the current source's output is directly proportional to light on the PV cell. This model includes only three parameters, namely short circuit current (I sc ), open-circuit voltage (I oc ) and diode ideality factor, which can completely define the I-V curve. The incorporation of series resistance allows an enhancement of the single diode model. This model is generally referred to as the R s model. The R s model is by far the most common in PV system simulation because of its simplicity and computational efficiency. The comparative findings lead to the assessment of the best single-diode photovoltaic model design that is more suited for different implementations of the power systems, regardless of the varying atmospheric conditions at various times of the day. The solar PV array is modelled according to the fundamental equations [18].
Let I Ph represents photocurrent (A), I PV and V PV represent the solar current (A) and PV voltage (V), K represents the Boltzmann constant, I D represents the diode saturation current (A), T n represents the PV cell reference temperature ( O C), n represents the ideality factor of the diode, Ns represents the number of seriesconnected PV cells, q is electron charge (coulombs) and R S is the series resistance ( ).

Operating modes of solar PV supported ZSI-DVR
The ZSI-DVR system supported by PV and battery worked in three possible operating modes [19]. Because of the multi-mode capability of the topology proposed, it manages the various situations effectively (voltage interruption, night time, maintenance, etc.) and provides the continuous clean power for the load.

Mode: 1
Due to the availability of solar irradiation, the output power of the PV generating system is high. The voltage sag/swell and reactive power in the grid supply is compensated by the PV-DVR system in this operating mode. Simultaneously, the excess PV power is provided to meet the demand for load and charges the battery storage unit. Mode 1 operation schematic circuit diagram is shown in Figure 3.

Mode: 2
The grid is disconnected from the load by semiconductor switches S 1 , S 2 , and S 3 during the voltage interruption period. The PV system /battery is responsible for meeting a critical load requirement during this period. Mode 2 operation schematic circuit diagram is shown in Figure 4.

Mode: 3
The PV generation system is shut down for maintenance purposes during the maintenance period. PV-DVR is separated using a bypass semiconductor switches S 6 , S 7 , and S 8 from series injection transformers. Mode 3 operation schematic circuit diagram is shown in Figure 5. The status of the switches for the different operating modes of PV supported ZSI-DVR is shown in Table 1.

Control scheme of solar PV interconnected ZSI-based DVR
The control algorithm based on U-SOGI is shown in Figure 6 is used to filter off harmonic load current components and extract the basic component for DVR switching pulse generation. This algorithm extracts both in-phase and quadrature elements from sensed load currents and measures the harmonics of load currents. For dynamic changes, a damping factor has been added to rapidly damp oscillations. Sensitive load currents are also used in their extraction of basic power components.
The Quadrature Signal Generator (QSG) is used as a building block for the SOGI algorithm as well as for a feedback circuit to find and transmit the load current to the controller in an error-less phase and frequency [14]. An additional integrator is included in the QSG block in conjunction with traditional SOGI to dampen the constant DC offset portion from the load current.
The transfer function of the feedback loop for the traditional SOGI algorithm as the orthogonal signal generator (OSG) is derived by the following equation: To decrease the error between sensed and real value, a damping factor is added to this OSG block. To reduce system oscillations during dynamic load changes, the damping factor is applied. A higher value increases the system's oscillation [20,21]. For U-SOGI algorithm, modified from a traditional SOGI algorithm, the OSG block transfer function is indicated as The traditional SOGI algorithm cannot eliminate DC offset and inter harmonic elements present in the input load currents. For the DC offset attenuation, a low-pass filter can be inserted into the system, but it again disrupts traditional SOGI algorithm filter efficiency [22]. In the derived essential active and reactive amplitude, the traditional SOGI has more oscillations than the U-SOGI and DC offsets. In comparison, there is no DC offset component and no oscillations in the fundamental amplitude obtained using the USOGI algorithm.  formed as follows by line voltages.
Let V sab , V sbc and V sac are estimated phase voltages from line voltages and w pa , w pb and w pc are active in-phase voltage templates.
The quadrature unit templates of the phase voltage components (w qa , w qb and w qc ) are determined by the following equation: The voltage level of the terminal is determined as The three-phase reference terminal voltage is estimated by the following equation: It is then transformed from abc -dq0 components using abc -dq0 transformation unit. The dq0 transformation of the reference terminal voltage is derived as ⎡ ⎣ cos(ωt) cos(ωt − 120) cos(ωt + 120) − sin(ωt) − sin(ωt − 120) − sin(ωt + 120) The error signal V Tedq0 has been determined by comparison of the reference voltage and load voltage dq0 component.
An error-driven Proportional Integer Derivative (PID) controller is used to control the proposed PV-DVR function. The controller's role is to change the error signal on the power distribution grid to be minimized. The compensation signal is also taken from the PID controller in the dq0 reference frame for the generation of the gating pulses to DVR switches.
where K p , K i, and K d represent the proportional, integral and derivative gains of the PID controller respectively.
The PID controller output is returned to the threephase abc reference frame to produce gating pulses on the DVR by using a hysteresis band comparator. The DVR compensation signals are generated and controlled by the feedback switching control by their actual values. The two control functions are configured, as they retain the grid voltage and load voltage, respectively under dynamic sources and load conditions. ⎡ Let V * Cd , V * Cq and V * C0 are estimated reference compensation voltage in dq0 reference frame and V * Ca , V * Cb and V * Cc are estimated reference compensation voltage in abc reference frame.

Results and discussions
MATLAB/Simulink has been used to create the SPV-DVR with the U-SOGI control scheme. To test the robustness of the control scheme, analyses of the balanced, unbalanced supply, dynamic unbalanced load, and various radiation systems for solar photovoltaic are conducted. A real-time experimental prototype built in the laboratory has been tested with the proposed control algorithm. Part of the hardware prototype image of ZSI-DVR is shown in Figure 7. The device is working at the reduction of the supply voltage through an autotransformer and is used as a non-linear load by the R-L diode bridge rectifier. Voltage sensors are used for detecting Point of Common Coupling (PCC) voltages, and solar PV array tension and Hall Effect current sensors are used to monitor grid currents and load currents. A Field-Programmable-Gate-Array (FPGA) controller is used to implement the proposed control algorithm. The required isolation between the ZSI-DVR and the FPGA signals can be provided by an      Optocoupler. The detailed parameters and ratings of the proposed system are given in Table 2.

Case 1: balanced supply voltage with balanced non-linear load
Assessment of the control technique suggested under the specific load condition has been carried out. A resistive-inductive (R-L) load with a diode bridge rectifier is used as a non-linear load. Figure 8 illustrates the simulation findings for voltage and current compensation under balanced voltage supply and non-linear load. An experimental investigation is performed with a balanced supply voltage and non-linear load for voltage and current compensation. Voltage and current  compensation outcomes of the experimental study under balanced voltage supply and non-linear load are shown in Figure 9.
The experimental findings for SPV-DVR in balanced voltage and non-linear load during the voltage compensation process are shown in Figure 10. The source current is maintained clean sinusoidal after compensation. The load voltage harmonic spectra with THD of 1.2%, whereas without SPV-DVR the load voltage THD of 25.5%. The THD of the load voltage under load unbalancing is also within the IEEE-519 standard [25] as shown in Figure 10(d). PV supply is maintained constant throughout and thus a stable DC-link voltage can be observed from Figure 11. Figure 12 shows the supply voltage and load voltage after compensation of voltage interruption. Figure 12(a) shows a similar set of waveforms with a sudden voltage reduction to zero. At that time, solar PV supported ZSI-DVR starts to compensate voltage interruption and injects the voltage to compensate for the voltage interruption present in the electric power distribution system.

Case 2: balanced supply voltage with unbalanced load
The desired performance of the PV-DVR system may poorly be affected due to unbalancing the load current. To verify the efficiency of the proposed control method under such conditions, simulation and experimentations have been carried out. The U-SOGI based PV-DVR behaviour under balanced supply voltage and unbalanced non-linear loaded conditions are shown in Figures 13 and 14. Figure 13 shows the source voltage, injected voltage, compensated load voltage and load current. At the beginning of steady-state voltage compensation analysis, the rated voltage (230V in RMS) is applied after that voltage sag and voltage interruption is applied. The solar PV ZSI-DVR identifies the voltage sag and voltage interruption and injects an appropriate compensating voltage in series with the supply voltage to eliminate the voltage sag. Figure 14 shows the experimental outcomes of load current, compensated source current and load and source current in the phase-A and inductor current and capacitor voltage of the LC filter.
Due to the connection of non-linear and unbalanced load in the system, the load currents are distorted by the harmonics. After connecting the solar PV supported ZSI-DVR, the source current waveforms are compensated by injecting an appropriate compensation current. The experimental outcomes for SPV-DVR in balanced voltage and unbalanced non-linear load during voltage compensation progression are shown in Figure 15. The PV array output voltage and the voltage across the DC -link of ZSI-DVR is shown in Figures 16 and 17. The PV array output voltage magnitude is stepping up into 600 V by using the Z-source network. Figure 18 shows the dynamic characteristics of the DC-link voltage control of the solar PV interfaced ZSI-DVR. The PV power generating system is maintained a stable DC-link voltage of the DVR can be observed from Figure 18(a, b).
The THD level of the load voltage without solar PV supported ZSI-DVR is 24.2(%) in phase-A and THD level of the load voltage with solar PV supported ZSI-DVR is 1.2(%) in phase-A shown in Figure 19. FFT analysis of source and load current has been carried out  for the stable source voltage and unstable load current condition. The THD level of the load voltage without solar PV supported ZSI-DVR is 24.2%, 23.7%, and 24.10% and with solar PV supported ZSI-DVR THD level of the load voltage is 1.2%, 1.3%, and 1.2%.

Case 3: unbalanced supply voltage with dynamic load condition
The desired performance of the PV-DVR system may poorly be affected due to unbalancing in the grid voltage. To verify the effectiveness of the proposed   Figure 20. Figure 21 shows the digital simulation results of PV array output voltage, DC-link voltage of the ZSI-DVR, THD level of load voltage without ZSI-DVR and THD level of load voltage with ZSI-DVR under the conditions of unbalanced voltages and load currents.
Experimental test results of unbalanced distorted load currents, unbalanced source voltages and harmonic distortions in a load voltage before connecting the solar PV ZSI-DVR under the unbalanced supply voltage and load conditions are shown in Figure 22. Figure 23 shows the series transformer injected voltage, load voltage with the SPV-DVR, load current with solar PV ZSI-DVR and load voltage THD level in phase-A for case-3 with solar PV ZSI-DVR. THD (%) level of the load voltage without the use of SPV-DVR is 23.2, 24.6, and 24.1; they are diminished to 1.3, 1.2, and 1.3 consistently with the use of SPV-DVR. The output of solar PV-DVR is tested with regard to the THD level of load voltage. Load voltages THD level with the use of solar PV ZSI-DVR and without the use of solar PV ZSI-DVR is shown Table 3. The THD measurement thresholds are very small and suitable for grid-connected applications based on the simulation and experiment outcomes. The performance of the solar PV supported ZSI-DVR is examined in terms of THD of load voltage. From the results of both simulation and experimental study, it is noticeable that load voltage THD is found to be well within 5% of the boundary indicated by IEEE Std. 519-1992 [25].
For non-linear load conditions, the voltage harmonics THD level is very less in the proposed method, approximately 1.2% compared to the existing methods      with a load voltage THD of 4.54% as presented by Pradhan and Mishra (17) [17].

Conclusion
In this paper, the benefits of distributed generation integration in a distribution network with an emphasis on power quality improvement have been addressed by employing the proposed U-SOGI based PV-DVR system. Voltage sag/swell and harmonic distortion has been considered in this paper as a power quality problem. Applying the proposed algorithm, the THD of source current has been maintained much below 5%, satisfying the IEEE-519 standard. The performance of the novel U-SOGI based control technique for voltage compensation reference signal generation is enhanced by proper tuning of parameters. A definitive analysis shows that maximum PV power is fed to the grid while effectively compensating the load harmonics using the proposed U-SOGI control technique. The experimental verification under different test cases has been presented to verify the validity of the proposed algorithm with a scaled-down laboratory prototype. Results obtained in all test cases have shown the improved performance of the system using the proposed control technique under steady-state as well as different dynamic conditions. The PV system has been ensured by using the reduced senor-based MPPT controller, which also helps to maintain the stable DC-link voltage in all conditions. The PV-DVR system suitably maintains the clean sinusoidal grid current under load and source unbalanced conditions as well as varying solar irradiation conditions.