Virtual impedance-based overclocking oscillation suppression method and device for grid-forming type photovoltaic grid-connected system

Abstract

The invention discloses a virtual impedance-based overclocking oscillation suppression method and device for a network-building type photovoltaic grid-connected system, and the method comprises the steps: building a network-building type photovoltaic grid-connected system model, and setting a virtual synchronous machine in the network-building type photovoltaic grid-connected system model; setting virtual impedance in the virtual synchronous machine so as to change output impedance of the virtual synchronous machine, and avoiding an oscillation frequency range caused by insufficient alternating-current potential margin of the grid-forming type photovoltaic grid-connected system and a power grid through the change of the output impedance; the broadband oscillation of the system can be well suppressed, the grid-connected stability of the virtual synchronous machine is improved, and the stable engineering suppression effect of overclocking oscillation is realized.

Description

Method and apparatus for suppressing overfrequency oscillation in grid-connected photovoltaic systems based on virtual impedance Technical Field

[0001] This invention relates to the field of overfrequency oscillation suppression, and in particular to a method and apparatus for overfrequency oscillation suppression in a grid-connected photovoltaic system based on virtual impedance. Background Technology

[0002] With the significant increase in the penetration rate of photovoltaic (PV) power generation in the power grid, its impact on the traditional power grid is also expanding, from simple local voltage fluctuations, output power fluctuations, and harmonic pollution to multiple aspects such as the safety and stability of the entire power grid and peak-shaving dispatch. The coexistence of the two modes of "low-voltage access, local consumption, and large-scale decentralized development" and "high-voltage long-distance transmission and consumption, medium-high voltage access, and large-scale centralized development" has a profound and significant impact on the power system, both in depth and breadth. Regarding the stability of PV grid-connected systems, existing literature indicates that there is a risk of overfrequency oscillations in PV grid-connected systems.

[0003] Overfrequency oscillation is generally analyzed using two main methods: time-domain analysis and frequency-domain analysis. Time-domain analysis analyzes the time-domain characteristics of the signal, including modal analysis and other methods. Frequency-domain analysis mainly studies the characteristics of the signal in the frequency domain, including impedance analysis and frequency scanning methods.

[0004] Impedance analysis was originally used to study the impact of filter parameters on the entire circuit. This includes methods for sequence impedance modeling. The basic principle of sequence impedance modeling is to superimpose a perturbation signal onto a sinusoidal signal, and then analyze the signal at each layer using methods such as wavelet transform. This yields an approximate response value of the entire system to the input signal. Dividing the response value by the input value gives the impedance at the perturbation frequency. Compared to modal analysis, impedance analysis has the advantage of strong engineering applicability and is convenient and quick to implement. Modal analysis, in turn, has the advantage of being able to calculate parameters containing oscillation mode information, such as participation factors, oscillation frequencies, and damping.

[0005] Oscillation suppression measures are generally divided into three categories. The first category is optimizing control parameters. This involves analyzing the parameters involved in overfrequency oscillation through characteristic structure analysis, and then controlling and optimizing these parameters to ultimately achieve the suppression goal. The second category is additional control methods. This involves introducing additional control structures without changing the original control structure. By changing the original characteristics of the system through these additional control structures, the oscillation can be suppressed. The third category is replacing the original control. This strategy replaces the entire control system with a new control method to solve the current oscillation problem, giving the entire system new dynamic characteristics and thus achieving the goal of oscillation suppression.

[0006] The first type of shortcoming in existing overfrequency oscillation research and mitigation solutions lies in the difficulty of online identification and analysis, and its strong dependence on engineering. Currently used methods such as time-domain analysis, frequency-domain analysis, modal eigenvalue analysis, impedance analysis, and frequency scanning are mostly based on relatively accurate equivalent modeling and linearization assumptions. They often require known control parameters, network topology, and operating points, and must be performed offline or under specific operating conditions. However, in high-penetration distributed photovoltaic grid-connected scenarios, the strength of the grid, the connection method, the inverter control mode, and the load status are highly time-varying, causing model parameters and oscillation frequencies to drift with operating conditions. Traditional methods are sensitive to uncertainty and noise, making it difficult to consistently and accurately provide oscillation types, dominant modes, and source-side location. Furthermore, obtaining impedance or frequency scan data often requires active perturbation or dedicated testing, which is costly to implement on-site, affects operational safety, and has limited characterization capabilities for oscillations caused by multiple inverters in parallel and wideband coupled oscillations. This easily leads to problems such as "detecting oscillations but failing to interpret them clearly, inaccurate location, and untimely warnings."

[0007] The second category of shortcomings focuses on "insufficient adaptability and synergy of suppression measures." Existing suppression methods typically fall into three categories: optimizing control parameters, adding auxiliary controls to the original control structure, and replacing the original control. While parameter optimization is relatively straightforward, it relies on system models and sensitivity and modal analysis results. Tuning often targets a single or a few operating conditions, making it difficult to meet the damping requirements across the entire bandwidth from low to high frequencies, and may sacrifice grid-connected dynamic performance or introduce new coupling modes. Adding auxiliary controls increases control links and parameter dimensions, leading to complex implementation, long debugging cycles, and uncertainties in multi-machine interactions. Without unified coordination, it may also induce oscillations in other frequency bands while suppressing localized issues. Replacing the original control strategy incurs high costs for retrofitting existing power plants, requires re-verification and compliance with grid connection specifications, and presents significant barriers to engineering implementation. Overall, existing solutions often employ a "single-machine thinking, offline tuning, and segmented governance" approach, lacking adaptive and closed-loop linkage capabilities for scenarios with high proportions of photovoltaic power generation, multi-source and multi-frequency coupling, and rapidly changing operating points. Therefore, they struggle to provide stable, low-cost, and scalable engineering-based suppression effects. Summary of the Invention

[0008] The technical problem to be solved by the present invention is to provide a method and device for suppressing overfrequency oscillations in a grid-connected photovoltaic system based on virtual impedance, so as to achieve an engineering-based suppression effect on the stability of overfrequency oscillations.

[0009] To address the aforementioned technical problems, the present invention provides a technical solution: a method for suppressing overfrequency oscillations in a grid-connected photovoltaic system based on virtual impedance, comprising: establishing a grid-connected photovoltaic system model; setting a virtual synchronizer in the grid-connected photovoltaic system model; setting a virtual impedance in the virtual synchronizer to change the output impedance of the virtual synchronizer; and avoiding the oscillation frequency range caused by insufficient phase margin between the grid-connected photovoltaic system and the power grid through the change of the output impedance.

[0010] To solve the above-mentioned technical problems, another technical solution adopted by the present invention is: an overfrequency oscillation suppression device for a grid-connected photovoltaic system based on virtual impedance, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the above-mentioned overfrequency oscillation suppression method for a grid-connected photovoltaic system based on virtual impedance.

[0011] The beneficial effects of this invention are as follows: by establishing a grid-connected photovoltaic system model, a virtual synchronous machine is set in the grid-connected photovoltaic system model; and a virtual impedance is set in the virtual synchronous machine to change the output impedance of the virtual synchronous machine. By changing the output impedance, the oscillation frequency range caused by insufficient phase margin between the grid-connected photovoltaic system and the grid is avoided, which can effectively suppress the wideband oscillation of the system, improve the grid connection stability of the virtual synchronous machine, and achieve an engineering-based suppression effect on overfrequency oscillation stability. Attached Figure Description

[0012] Figure 1 is a flowchart of the steps of an overfrequency oscillation suppression method for a grid-connected photovoltaic system based on virtual impedance according to an embodiment of the present invention; Figure 2 is a structural schematic diagram of a grid-connected photovoltaic system according to an embodiment of the present invention; Figure 3 is a structural schematic diagram of a virtual synchronous machine controller according to an embodiment of the present invention; Figure 4 is a structural schematic diagram of a virtual synchronous machine dual-loop controller according to an embodiment of the present invention; Figure 5 is a schematic diagram of the equivalent model after introducing virtual impedance according to an embodiment of the present invention; Figure 6 is a structural schematic diagram of a virtual synchronous machine based on virtual impedance according to an embodiment of the present invention; Figure 7 is a structural schematic diagram of an improved virtual synchronous machine based on virtual impedance according to an embodiment of the present invention; Figure 8 is a structural schematic diagram of a dual-loop controller after introducing virtual impedance according to an embodiment of the present invention; Figure 9 is a Bode plot of the system output impedance when keeping the virtual inductance constant and changing the virtual resistance according to an embodiment of the present invention; Figure 10 is a Bode plot of the system output impedance when keeping the virtual resistance constant and changing the virtual inductance according to an embodiment of the present invention; Figure 11 is a system eigenvalue distribution diagram according to an embodiment of the present invention; Figure 12 is an overall design flowchart according to an embodiment of the present invention; Figure 13 is a structural schematic diagram of an overfrequency oscillation suppression device for a grid-connected photovoltaic system based on virtual impedance according to an embodiment of the present invention. Detailed Implementation

[0013] To explain in detail the technical content, objectives, and effects of the present invention, the following description is provided in conjunction with the embodiments and accompanying drawings.

[0014] The above-mentioned method and device for suppressing overfrequency oscillations in a grid-connected photovoltaic system based on virtual impedance can be applied to the overfrequency oscillations of a grid-connected photovoltaic system. The following is a detailed description of the specific implementation: In one optional implementation, as shown in Figure 1, a method for suppressing overfrequency oscillations in a grid-connected photovoltaic system based on virtual impedance includes the following steps: S1, establishing a grid-connected photovoltaic system model, and setting a virtual synchronizer in the grid-connected photovoltaic system model; the grid-connected photovoltaic system model includes a photovoltaic power supply, a DC-DC boost circuit, an inverter circuit, and a filter circuit connected in sequence; the virtual synchronizer is set between the inverter circuit and the filter circuit; in specific implementation, the photovoltaic grid-connected system mainly consists of four parts, namely, a photovoltaic power supply, a DC-DC boost circuit, an inverter circuit, and a filter circuit. The photovoltaic power supply consists of a conversion circuit, an inverter circuit, and a filter circuit. While boosting the voltage, the circuit uses MPPT (Maximum Power Point Tracking) to track the maximum power output of the photovoltaic system, and then connects it to the inverter for grid-connected inverter operation, as shown in Figure 2. , The voltage and current generated by the photovoltaic cells. , , They are Circuit parameters, , It is controlled by MPPT. parameter, For photovoltaic DC voltage after circuit transformation This is the output current of the photovoltaic cell, which is related to the light intensity and the ambient temperature, and can be expressed as:

[0015] In the formula, This is the short-circuit current. Open circuit voltage, and Defined as:

[0016] In the formula, and These are the output voltage and current of a photovoltaic cell when it operates at its maximum power point, among which... , , and for:

[0017] In the formula, Let S be the air temperature, S be the light intensity, and k, a, b, and c be compensation factors. To ensure the photovoltaic array always operates at its maximum power point, an MPPT control strategy is adopted to guarantee maximum power operation of the system even when light intensity and temperature change. For reference irradiance (calibrated irradiance), 1000 W / m² is commonly used. 2 , For reference short-circuit current (in) , (short-circuit current below) For reference open circuit voltage (in) , (open circuit voltage below) For reference to the maximum power point current, For reference to the maximum power point voltage, The reference temperature (calibrated temperature) is usually 25°C (i.e., 298.15 K).

[0018] The duty cycle that enables the system to reach its maximum power point is calculated using the following formula:

[0019] in, It is a state variable of the PI controller.

[0020] The equation for the Boost circuit (i.e., the DC-DC boost circuit) is:

[0021] in, and These represent the output voltage and output current of the Boost circuit, respectively. pv i is the inductance value of the Boost circuit. a1 R is the current flowing through the inductor of the Boost circuit, and R is the DC-side equivalent load resistance.

[0022] S2. A virtual impedance is set in the virtual synchronizer to change its output impedance. This change in output impedance avoids the oscillation frequency range caused by insufficient phase margin between the grid-connected photovoltaic system and the grid. The virtual synchronizer includes a virtual synchronizer controller and a dual-loop controller connected in sequence. The virtual synchronizer controller is connected to the filter circuit. The dual-loop controller is connected to the inverter circuit. The virtual impedance is set between the virtual synchronizer controller and the dual-loop controller. Specifically, the instantaneous active power and reactive power output by the VSG (virtual synchronizer) are... and The inverter output voltage and output current After the dq transformation, the result is obtained using the following formula:

[0023] In the formula, u od Let u be the d-axis component of the inverter output voltage vector in the dq coordinate system. oq Let i be the q-axis component of the inverter output voltage vector in the dq coordinate system. od Let i be the d-axis component of the inverter output current vector in the dq coordinate system. oq The q-axis component of the inverter output current vector in the dq coordinate system.

[0024] The virtual synchronous machine comprises a virtual synchronous machine controller and a dual closed-loop controller. The control strategy of the virtual synchronous machine controller primarily simulates the active power frequency regulation and reactive power voltage regulation characteristics of the synchronous machine, while also addressing the inertial damping characteristics lacking in droop control, another form of grid-based control. The active power control loop consists of a virtual speed governor and a virtual rotor to simulate their characteristics; the reactive power control loop consists of a virtual exciter. The power outer loop control based on the virtual synchronous machine is shown in Figure 3. The virtual rotor characteristics in the virtual synchronous machine are represented by the second-order swing equation of the generator. Using the swing equation to simulate the characteristics of the synchronous generator, the virtual rotor model is shown in the following equation:

[0025] In the formula, , These are the mechanical torque and electromagnetic torque of the synchronous generator, respectively, in N·m; In a synchronous generator, it is defined as the rotor angular frequency; in virtual synchronous machine control, it is the system output angular frequency, with units of rad / s. ω is the angular frequency of the power grid, in rad / s; D is the damping coefficient, in N·m·s / rad; J is the moment of inertia, in kg·m·s / rad. ; This represents the phase angle of the synchronous generator's electromotive force.

[0026] The virtual speed governor simulates the active-frequency droop characteristic of a synchronous generator, and the output power of the virtual synchronous machine... Given power and the output of the regulator The virtual speed governor model is composed of the following formula:

[0027] In the formula, The active power setpoint, This is the active power-frequency droop factor.

[0028] The reactive power control element in VSG control is called a virtual exciter, which simulates the QU droop characteristic of a synchronous generator. The reactive power control model of VSG is obtained from the following formula:

[0029] in, The given value for reactive power. This is the voltage reference value for VSG; This is the QU droop coefficient.

[0030] The voltage and current dual-loop control serves as the inner loop control of the virtual synchronous machine. Both the voltage and current loops employ PI control. The three-phase signal synthesized from the outer loop of the virtual synchronous machine is transformed by dq and used as the setpoint. The voltage across the filter capacitor is used as feedback, and through cross-compensation and PI control, a reference value for the inner current loop is generated. The current across the filter inductor is then used as feedback, and through cross-compensation and PI control, a modulation signal is generated. The dual-loop control strategy is shown in Figure 4. The equations for the outer voltage loop and the inner current loop control are as follows:

[0031] in , , , For the integrator output of the voltage loop and current loop PI controllers; and These are the proportional and integral coefficients for the voltage outer loop control, respectively. , These represent the d-axis and q-axis components of the voltage loop reference value obtained from the outer loop output of the VSG and after dq transformation. , These are the d-axis and q-axis components of the reference value output from the voltage loop control to the current loop, respectively. , These are the d-axis component and q-axis component of the inverter output voltage vector in the dq coordinate system, respectively. and These are the proportional and integral coefficients for the inner current loop control, respectively. , These represent the d-axis and q-axis components of the inverter-side filter inductor current in the LCL filter in the dq coordinate system. , These represent the d-axis and q-axis components of the voltage modulation command output by the inner current loop controller in the dq coordinate system.

[0032] The voltage and current output from the VSG are filtered by an LCL filter before being connected to the power grid. The state equation of the filter can be expressed as follows:

[0033] in, , These represent the d-axis and q-axis components of the output voltage of the photovoltaic system after VSG control, in the dq coordinate system. , These represent the d-axis and q-axis components of the grid voltage in the dq coordinate system, respectively. , These are the d-axis and q-axis components of the inverter output current vector in the dq coordinate system, respectively. f L f C f These are the parasitic resistance, filter inductance, and filter capacitance, respectively. g R is the grid-connected inductance value of the LCL filter. e This is the equivalent series resistance of the grid-connected inductor of the LCL filter.

[0034] To suppress wideband oscillations in the system and improve the stability of the virtual synchronous machine (VSG) connected to the grid, the output impedance of the VSG needs to be reshaped to change its impedance and avoid the oscillation frequency range caused by insufficient phase margin between the system and the grid. Adding impedance to the actual circuit would consume electrical energy, reduce energy utilization, and cause impedance heating. Therefore, an impedance control module is added to the VSG controller. By monitoring the current, the current is fed back to the control circuit. The feedback current is multiplied by the virtual impedance to obtain the impedance voltage drop, which in turn changes the voltage reference value of the control circuit, thereby changing the output impedance of the VSG. The principle is: adding a virtual impedance to the control circuit. After introducing virtual impedance, the equivalent model is shown in Figure 5.

[0035] The virtual impedance is transformed using the dq axis transformation, and the specific expression is as follows:

[0036] In the formula: The given potential for a normal virtual synchronous machine, , These are the d-axis and q-axis components of the output current of the virtual synchronous machine, respectively, and are also the feedback quantities of the virtual impedance. , These are the d-axis and q-axis components of the new given voltage after virtual impedance compensation, respectively. This given voltage serves as the input to the voltage loop. and These are virtual resistance and inductance, respectively.

[0037] In another alternative implementation, the virtual impedance ignores the differential term.

[0038] As shown in Figure 6, the control of the virtual inductor in the virtual impedance strategy contains a differentiating term. This term can easily cause oscillations in high-frequency components, reducing the oscillation suppression effect. A low-pass filter can be added to address this issue, but eliminating high-frequency harmonics will increase system delay. Since the analysis mainly focuses on impedance in the frequency band below 1000Hz, the differentiating term of the virtual impedance can be ignored, simplifying the circuit and reducing system delay. At this point, the virtual impedance... The simplified VSG control block diagram based on virtual impedance is shown in Figure 7.

[0039] Figure 7 illustrates the improved VSG control block diagram based on virtual impedance control, in which... The improved virtual impedance control equation shown in Figure 7 can be described as follows:

[0040] In the formula, The given potential for a normal virtual synchronous machine, This represents the d-axis component of the output current of the virtual synchronous machine. This represents the q-axis component of the output current of the virtual synchronous machine. This represents the d-axis component of the new given voltage after virtual impedance compensation. This represents the q-axis component of the new given voltage after virtual impedance compensation. and These are virtual resistance and inductance, respectively.

[0041] As shown in Figure 7, the virtual impedance is located between the dual-loop control and the virtual synchronous machine control. After the virtual synchronous machine control, the output voltage minus the voltage drop of the product of the output current and the virtual impedance yields the given voltage value for the dual-loop control. Under the improved strategy, the output grid-connected voltage changes from the original... , Become , This, in turn, changes the output impedance characteristics of the VSG.

[0042] In another alternative implementation, to explore suitable virtual inductance and virtual resistance values, a dual-loop control block diagram with virtual impedance is introduced, as shown in Figure 8. The transfer function of the dual-loop controller of the virtual synchronous machine is:

[0043] in,

[0044] In the above formula, The transfer function representing the voltage loop through the PI controller. , These are the proportional and integral parameters of the voltage loop; , Represents the proportional parameter of the current loop; , Represents the filter inductor. Represents parasitic resistance; , Represents the filter capacitor; , These are the transfer functions of the grid-connected voltage and current after passing through the virtual synchronizer and filter, respectively. The transfer function of the output voltage; This is the internal equivalent given voltage reference value for the virtual synchronous machine before the introduction of virtual impedance compensation; The output impedance of the virtual synchronizer before considering virtual impedance, The transfer function of virtual impedance; This is the cutoff angular frequency of the low-pass filter; Inverter gain; For virtual resistance, This is a virtual inductance.

[0045] The equivalent impedance after adding the virtual impedance control strategy is: Therefore, from the perspective of the transfer function, we can derive... The resistance and inductance directly affect the output impedance. To more accurately determine the influence of resistance and inductance on the system, Bode plots are used to analyze the inductance and resistance. First, keeping the inductance L = 1mH, the Bode plot of the system transfer function is plotted after changing the virtual resistance, as shown in Figure 9.

[0046] Observing Figure 9, it can be seen that in the frequency band below 1000Hz, as the virtual resistance increases, the output resistance... The resistance value increases significantly, and the amplitude of the Bode plot analysis also increases with the increase of the virtual resistance. The phase plot reveals that in the frequency band above 1000Hz, when the virtual resistance is relatively small, the phase-frequency response curve... Above this point, the system exhibits inductive behavior, which is detrimental to stable system operation; by adding a virtual resistor, Under the phase-frequency characteristic, the phase gradually decreases. From the perspective of root locus, it can be considered that the conjugate poles gradually move towards the negative half-axis of the s-plane and gradually become stable.

[0047] From the above analysis, we can derive the virtual resistance. The value cannot be too small, but The virtual resistance cannot be increased indefinitely; an excessively large virtual resistance will increase the difference between the output voltage and the reference voltage of the virtual synchronizer. Keeping the virtual resistance value constant, the virtual inductance... Figure 10 shows the Bode plot of the output impedance from 2mH to 8mH. As can be seen from Figure 10, the size of the virtual inductance has almost no effect on the system output impedance below 1000Hz. Therefore, increasing the virtual inductance will not increase the equivalent inductance of the system in the mid-to-low frequency range. However, the virtual inductance is of great significance to the low-pass filter.

[0048] In another optional implementation, the method further includes: establishing a small-signal model of the grid-connected photovoltaic system; calculating the eigenvalues ​​of the small-signal model and obtaining the oscillation frequency; analyzing the distribution of the eigenvalues ​​in the complex plane and evaluating the stability of the grid-connected photovoltaic system under different operating conditions; wherein, establishing the small-signal model of the grid-connected photovoltaic system includes: performing local approximate linearization on the photovoltaic nonlinear system, decomposing the photovoltaic nonlinear system into a DC component and a higher-order small-signal component, where the DC component represents the system at a stable operating point, and the higher-order small-signal component represents the dynamic response at the operating point, and establishing the small-signal model of the grid-connected photovoltaic system based on the decomposed grid-connected photovoltaic system model.

[0049] The step of establishing the small-signal model of the grid-connected photovoltaic system based on the decomposed grid-connected photovoltaic system model includes: determining the small-signal model corresponding to the DC-DC boost circuit; linearizing the virtual synchronous machine to obtain the small-signal model corresponding to the virtual synchronous machine; linearizing the filter circuit to obtain the small-signal model corresponding to the filter circuit; and determining the small-signal model of the grid-connected photovoltaic system based on the above small-signal models.

[0050] In practice, the dynamic process of the entire photovoltaic grid-connected system can be decomposed into many differential algebraic equations. The photovoltaic nonlinear system is locally approximated and linearized, that is, the nonlinear system is decomposed into DC components and higher-order small-signal components. The DC component approximation can be regarded as the system being at a stable operating point, and the higher-order small-signal components can be regarded as the dynamic response at the operating point.

[0051] The voltage generated by the photovoltaic cell , Transformed DC voltage intermediate variables and the current through the inductor This can be represented as the sum of a DC quantity and a higher-order small-signal component:

[0052] By setting the AC and DC terms to be equal, the small-signal model can be obtained as shown in the following equation:

[0053] Solving the above equations simultaneously, we obtain the small-signal model of the Boost system:

[0054] In the formula For system state variables; and Let be the coefficient matrix, where and Represented as:

[0055]

[0056] ,

[0057] in, It is a small signal change that controls the PWM duty cycle of the Boost circuit. It is the only control input in the system, and closed-loop control of the photovoltaic system state is achieved by adjusting it.

[0058] Next, the control of the virtual synchronous machine is linearized, resulting in: ,in, The second-order swing equation and reactive power-voltage droop characteristic equation of the virtual synchronous machine control strategy, after linearization, are obtained as follows:

[0059]

[0060] After virtual synchronous machine control, voltage and current tracking is performed using a dual closed loop. The SPWM modulation signal is generated through the dual closed loop, and linearization yields:

[0061] in:

[0062]

[0063]

[0064] in:

[0065]

[0066] After the system is controlled by a virtual synchronous machine, the output voltage and current are connected to the power grid through an LCL filter for linearization, resulting in:

[0067] in:

[0068]

[0069] After modeling the transmission lines and the power grid, since the voltage and frequency of the power grid change very little and can be considered constant, while the output voltage and frequency of the VSG are unstable, there is a phase difference between the power grid voltage and the VSG voltage. It can be represented as:

[0070] in, The rate of change of phase angle, The VSG outputs the angular frequency. This is the angular frequency of the power grid.

[0071] Transforming the grid voltage to a common coordinate system yields:

[0072] in, , These represent the minute changes in grid voltage along the direct and quadrature axes in the dq coordinate system, respectively. The equivalent voltage amplitude of the power grid. The power angle of the power grid in steady state (initial power angle). This represents a small disturbance to the power angle of the power grid.

[0073] Step 5: The small-signal state-space equations for the entire system can be obtained;

[0074] in, , , ;in, It contains 18 state variables, as shown below:

[0075] In the above formula, with The subscript dq indicates the small signal component of the corresponding symbol, and the subscript dq indicates the value obtained after the corresponding symbol is transformed by dq.

[0076] After performing small-signal modeling on the entire inverter system, its eigenvalues ​​and oscillation frequencies need to be calculated. Small-signal linearization modeling is then performed on the grid-connected inverter system to construct its state-space expression. Eigenvalue analysis is then performed on its state matrix to obtain the system's oscillation modes and corresponding oscillation frequencies. By analyzing the distribution of eigenvalues ​​in the complex plane, the system's stability under different operating conditions is evaluated. Participation factor analysis is used to quantitatively describe the influence of system state variables on each oscillation mode, thereby identifying key control parameters that significantly affect broadband oscillation modes. Based on this, by perturbing the key parameters and plotting eigenvalue root locus curves, the impact mechanism of parameter changes on the oscillation modes and stability of the photovoltaic grid-connected system is analyzed. By changing these parameters and observing the root locus curves, the impact of different parameter changes on the stability of the photovoltaic system is obtained through root locus analysis.

[0077] Solving for the coefficient matrix of the entire photovoltaic coefficient yields the eigenvalues ​​and eigenvectors of the system:

[0078] In the formula, for the matrix Solve the matrix to find its eigenvalues. Next, find the left and right eigenvectors of the eigenvalues. and .like Then there will be set up ,but:

[0079] In the formula, if the transpose of V is equal to the inverse of U, then the product of the transpose of V and U is the identity matrix, and U and V are orthogonal matrices. In the calculated eigenvalues, each eigenvalue corresponds to a mode and a different eigenvalue root. There are different situations: when When the imaginary part of the characteristic root is 0, there is no oscillation; when When the eigenvalues ​​are conjugate eigenvalues, then... This indicates that the system has a suppressive effect on this oscillation; the system can dampen this oscillation when The time indicates that the system has an amplifying effect on this oscillation, and the system will diverge this oscillation.

[0080] A grid-connected photovoltaic inverter model was established based on the Matlab platform. The main parameters are shown in Table 1: Table 1 Main parameters of the grid-connected photovoltaic inverter model

[0081] The eigenvalues ​​of the above single-machine model are calculated and plotted on the complex plane, as shown in Figure 11.

[0082] Figure 12 shows the overall design flowchart of this embodiment.

[0083] In another alternative embodiment, as shown in FIG13, an overfrequency oscillation suppression device for a grid-connected photovoltaic system based on virtual impedance includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the overfrequency oscillation suppression method for a grid-connected photovoltaic system based on virtual impedance described in any of the above embodiments.

[0084] In summary, this invention provides a method and apparatus for suppressing overfrequency oscillations in grid-connected photovoltaic systems based on virtual impedance. This method achieves targeted suppression of overfrequency oscillations, addressing the insufficient frequency band coverage of existing methods. By introducing a frequency-selective virtual impedance design into the control stage of the grid-connected photovoltaic inverter, the suppression mechanism can be concentrated on specific oscillation modes within the overfrequency band. Unlike the fixed-parameter damping or passband-consistent virtual impedance commonly used in existing technologies, this invention constructs a frequency band selection function based on the oscillation frequency characteristics. This significantly enhances the virtual impedance within the target overfrequency range while minimizing its impact on normal operating frequencies such as power frequency and subsynchronous frequency. Through this differentiated impedance shaping, this method not only effectively improves the damping ratio of the overfrequency mode but also avoids negative interference with the system's dynamic performance, overcoming the problems of "limited bandwidth and easy mis-suppression of normal power oscillations" inherent in traditional methods. This significantly improves the stable operation capability in high-penetration photovoltaic scenarios.

[0085] This invention achieves adaptive impedance regulation, improving adaptability under various operating conditions and weak grid conditions. When grid strength, load changes, or multiple units operating in parallel cause oscillation frequency drift, traditional setpoint control often fails to maintain a consistently effective suppression effect. This method, however, adjusts the impedance amplitude and phase angle using detected frequency domain amplitude and phase shift indicators, ensuring the controller always matches the current oscillation mode. This mechanism significantly enhances the stability of grid-connected photovoltaic inverters under weak grid conditions or rapidly changing operating conditions, enabling the system to maintain wide-band steady-state suppression capabilities without relying on complex modeling and offline tuning. This fundamentally solves the core problems of existing technologies, namely "parameter sensitivity and poor adaptability."

[0086] This invention reduces system coupling interference and improves the collaborative stability of multi-inverter parallel scenarios. The proposed virtual impedance overfrequency suppression method can create an equivalent impedance reconstruction effect between grid-connected photovoltaic inverters, effectively reducing high-frequency circulating currents and mutual excitation problems caused by control coupling between multiple inverters. By directional dissipation of overfrequency oscillation energy, this method weakens the coupling modes between multiple inverters, thereby avoiding the "local suppression-system amplification" phenomenon commonly found in existing technologies. Furthermore, since the virtual impedance structure of this invention is located in the local control layer, it does not change the grid-connected protection logic and does not rely on communication links, enabling rapid deployment and collaborative gain in large-scale photovoltaic power plants. Compared with traditional centralized coordinated control, this method has advantages such as simple structure, easy engineering implementation, and low invasiveness to existing systems, comprehensively improving the system stability margin and operational reliability in multi-grid-connected scenarios.

[0087] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent modifications made based on the content of the present invention specification and drawings, or direct or indirect applications in related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for suppressing overfrequency oscillations in a grid-connected photovoltaic system based on virtual impedance, characterized in that, include: A grid-connected photovoltaic system model is established, and a virtual synchronous machine is set in the grid-connected photovoltaic system model. A virtual impedance is set in the virtual synchronous machine to change the output impedance of the virtual synchronous machine. The oscillation frequency range caused by insufficient phase margin between the grid-connected photovoltaic system and the grid is avoided by changing the output impedance.

2. The method for suppressing overfrequency oscillations in a grid-connected photovoltaic system based on virtual impedance according to claim 1, characterized in that, The virtual impedance ignores the differential term.

3. The method for suppressing overfrequency oscillations in a grid-connected photovoltaic system based on virtual impedance according to claim 2, characterized in that, The governing equation for the virtual impedance is: In the formula, The given potential for a normal virtual synchronous machine, This represents the d-axis component of the output current of the virtual synchronous machine. This represents the q-axis component of the output current of the virtual synchronous machine. This represents the d-axis component of the new given voltage after virtual impedance compensation. This represents the q-axis component of the new given voltage after virtual impedance compensation. and These are virtual resistance and inductance, respectively.

4. A method for suppressing overfrequency oscillations in a grid-connected photovoltaic system based on virtual impedance according to any one of claims 1 to 3, characterized in that, The grid-connected photovoltaic system model includes a photovoltaic power source, a DC-DC boost circuit, an inverter circuit, and a filter circuit connected in sequence; the virtual synchronizing machine is located between the inverter circuit and the filter circuit.

5. The method for suppressing overfrequency oscillations in a grid-connected photovoltaic system based on virtual impedance according to claim 4, characterized in that, The virtual synchronizer includes a virtual synchronizer controller and a dual closed-loop controller connected in sequence; the virtual synchronizer controller is connected to the filter circuit; the dual closed-loop controller is connected to the inverter circuit; and the virtual impedance is set between the virtual synchronizer controller and the dual closed-loop controller.

6. The method for suppressing overfrequency oscillations in a grid-connected photovoltaic system based on virtual impedance according to claim 5, characterized in that, The transfer function of the dual-loop controller of the virtual synchronizer is: in, In the above formula, The transfer function representing the voltage loop through the PI controller. 、 These are the proportional and integral parameters of the voltage loop; ， Represents the proportional parameter of the current loop; ， Represents the filter inductor. Represents parasitic resistance; ， Represents the filter capacitor; 、 These are the transfer functions of the grid-connected voltage and current after passing through the virtual synchronizer and filter, respectively. The transfer function of the output voltage; This is the internal equivalent given voltage reference value for the virtual synchronous machine before the introduction of virtual impedance compensation; The output impedance of the virtual synchronizer before considering virtual impedance, The transfer function of virtual impedance; This is the cutoff angular frequency of the low-pass filter; Inverter gain; For virtual resistance, This is a virtual inductance.

7. The method for suppressing overfrequency oscillations in a grid-connected photovoltaic system based on virtual impedance according to claim 4, characterized in that, It also includes: establishing a small-signal model of the grid-connected photovoltaic system; calculating the eigenvalues ​​of the small-signal model and obtaining the oscillation frequency; analyzing the distribution of the eigenvalues ​​in the complex plane and evaluating the stability of the grid-connected photovoltaic system under different operating conditions.

8. The method for suppressing overfrequency oscillations in a grid-connected photovoltaic system based on virtual impedance according to claim 7, characterized in that, The establishment of the small-signal model of the grid-connected photovoltaic system includes: performing local approximate linearization on the photovoltaic nonlinear system, decomposing the photovoltaic nonlinear system into DC components and higher-order small-signal components, where the DC component represents the system at a stable operating point and the higher-order small-signal components represent the dynamic response at the operating point, and establishing the small-signal model of the grid-connected photovoltaic system based on the decomposed grid-connected photovoltaic system model.

9. The method for suppressing overfrequency oscillations in a grid-connected photovoltaic system based on virtual impedance according to claim 8, characterized in that, The step of establishing the small-signal model of the grid-connected photovoltaic system based on the decomposed grid-connected photovoltaic system model includes: determining the small-signal model corresponding to the DC-DC boost circuit; linearizing the virtual synchronous machine to obtain the small-signal model corresponding to the virtual synchronous machine; linearizing the filter circuit to obtain the small-signal model corresponding to the filter circuit; and determining the small-signal model of the grid-connected photovoltaic system based on the above small-signal models.

10. A device for suppressing overfrequency oscillations in a grid-connected photovoltaic system based on virtual impedance, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the overfrequency oscillation suppression method for a grid-connected photovoltaic system based on virtual impedance as described in any one of claims 1 to 9.

Images