Distributed photovoltaic grid-connected converter low-voltage ride-through control method based on sum-difference coordinate transformation

Abstract

The invention provides a distributed photovoltaic grid-connected converter low-voltage ride-through control method based on sum-difference coordinate transformation, and solves the problems of insufficient synchronism, poor stability, insufficient reactive power support potential excavation and the like existing in collaborative response under power grid voltage drop during distributed photovoltaic grid connection. According to the method, impedance modeling and equivalent symmetry transformation are carried out on the distributed photovoltaic inverter, and a coupling multi-dimensional matrix equation is established; an original coupling system is decoupled into a sum current channel and a plurality of difference current channels, the control complexity is simplified, and an open-loop output impedance expression and a closed-loop output impedance expression are deduced; secondly, formulating a refined active / reactive current collaborative distribution strategy for different voltage drop depths, dynamically calculating the minimum reactive current during voltage drop, preferentially meeting the national standard, and maximizing active output or balancing loss under the limit of rated current; and finally, through sum-difference coordinate transformation, the total current demand and distribution are efficiently decoupled and issued to a current loop controller of each inverter, so that rapid, synchronous and stable cooperative operation of multiple inverters is realized.

Description

TECHNICAL FIELD

[0001] The application relates to a low-voltage ride-through control method for a distributed photovoltaic grid-connected converter based on sum-difference coordinate transformation, in particular to an asymmetric inverter sum-difference coordinate decoupling control method and a method for coordinated reactive power compensation of multiple inverters in a photovoltaic power station during grid voltage drop. BACKGROUND

[0002] With the continuous increase of renewable energy in the power system, countries have introduced new energy grid connection technical specifications to cope with the challenges of power grid stability caused by the uncertainty of renewable energy output. Especially in areas with high photovoltaic penetration, grid-connected inverters must have low-voltage ride-through (LVRT) function to meet the safety operation requirements of the power grid. When the grid voltage drops due to short-circuit faults and other faults, the photovoltaic system needs to maintain grid-connected operation within the specified time and voltage drop range, and achieve voltage support by dynamically adjusting the output power. This technical feature has become a core indicator for evaluating the grid performance of photovoltaic power stations.

[0003] In recent years, with the maturity of photovoltaic power generation technology, LVRT technology has gradually developed towards active support for the grid, coordinated control of software and hardware, and fine adaptation to multiple scenarios. Especially under the background of increasing new energy installed capacity and increasing requirements for power grid safety and stability, the reactive and active coordination capability, rapidity and robustness of fault ride-through are becoming increasingly important. However, the complex transient process caused by high proportion of new energy grid connection, the adaptability of weak grid environment, and the difficulty in balancing cost and performance are still the current technical breakthrough difficulties.

[0004] In addition, how to fully exploit and dynamically allocate the reactive power output margin of the inverter to achieve precise support for the grid voltage, and how to improve the synchronization and stability of low-voltage ride-through of multiple units in a photovoltaic power station through distributed coordinated control strategy, are important links in building a high-resilience new energy station. Although existing inverters have reactive and active regulation capabilities, there are still risks of delay and oscillation in the coordinated response of multiple units during sudden deep voltage drop. The current coordinated control of multiple inverters in a weak grid environment has the disadvantages of insufficient synchronization, poor stability, and insufficient potential for reactive power support.

[0005] Therefore, in-depth research and application of dynamic reactive power margin allocation technology and distributed coordinated control strategy are of great significance for significantly improving the voltage active support capability and fault ride-through transient synchronization of photovoltaic power stations and building a high-resilience new energy station. SUMMARY

[0006] The application aims to provide a low-voltage ride-through control method for a distributed photovoltaic grid-connected converter based on sum-difference coordinate transformation, which realizes the collaborative low-voltage ride-through of the distributed photovoltaic by reasonably distributing the active and reactive power according to the requirements of different working conditions when the low-voltage ride-through is performed through the centralized control of the distributed photovoltaic inverter.

[0007] The application provides a low-voltage ride-through control method for a distributed photovoltaic grid-connected converter based on sum-difference coordinate transformation, which comprises the following steps:

[0008] Step S1: impedance modeling and equivalent transformation are performed on the distributed photovoltaic inverter, and a multi-dimensional matrix equation is constructed to lay a foundation for subsequent decoupling processing;

[0009] Step S2: for the parameter-heterogeneous multi-inverter system, a sum-difference coordinate transformation matrix is constructed, the multi-dimensional matrix obtained in step S1 is simplified into a diagonal matrix through the matrix, and system decoupling is realized;

[0010] Step S3: the p-dimensional model decoupled in step S2 is remapped to the N-dimensional physical space to obtain an equation corresponding to the actual inverter and a sum-difference coordinate transformation matrix adapted to the actual inverter;

[0011] Step S4: the accurate expression of the open-loop and closed-loop output impedance is derived based on the result of step S3 in combination with the control loop;

[0012] Step S5: according to the low-voltage ride-through technical regulation, the current given value of the active and reactive power of the system under normal working conditions and when the grid voltage drops between 0.2pu and 0.85pu is calculated;

[0013] Step S6: the current given value of the active and reactive power of the system when the grid voltage drops below 0.2pu is calculated.

[0014] As preferred in the application, the implementation process of step S1 comprises:

[0015] Firstly, the n inverters are all inductively filtered, and are connected to the high-voltage grid through a step-up transformer;

[0016] Further, the filter inductance of the i-th converter is defined as L i =L / n i (i=1,2,…,N), wherein L is a reference inductance value, n i is the proportional coefficient of the i-th inverter;

[0017] On this basis, the expression of the voltage u pcc at the common coupling point is established through the root superposition theorem:

[0018]

[0019] where u i is the output voltage of the i-th inverter, Z L is the filter impedance of the first filter, u g is the grid voltage, Z g is the grid impedance, and s is the Laplace operator.

[0020] To construct the unified decoupled model, an equivalent symmetrization method for parameter heterogeneous systems is proposed: the i-th non-symmetrical inverter is equivalent to a parallel structure of n i symmetrical virtual inverters, and its rated current is reduced to 1 / n i of the original value to satisfy the power conservation, i.e.,

[0021]

[0022] where i i is the output current of the i-th inverter.

[0023] Let the total number of equivalent inverters be:

[0024]

[0025] After obtaining the above expression in the frequency domain, it is converted into a time-domain equation by inverse Laplace transform, and then a p-dimensional coordinate system is selected for modeling. Since the output current is controlled by the inverter output voltage, the control variables and output variables of the p-dimensional coordinate system are defined as u j and i j , respectively.

[0026] U j = [u1, u2, …, u p ] T

[0027] I j = [i1, i2, …, i p ] T

[0028] where j = 1, 2, …, p.

[0029] Finally, the system is modeled in the p-dimensional coordinate system, and the admittance model of the multiple parallel inverters is:

[0030] I(s) = Y(s)(U(s) - u g (s)E)

[0031]

[0032] Wherein, E is a unit vector, Y(s) is a transfer function matrix of p equivalent inverters, representing the dynamic coupling effect between virtual converters.

[0033] As preferred in the present application, the implementation process of step S2 specifically comprises:

[0034] First, for the decoupling requirement of the equivalent symmetrization system, an extended dimension and a difference coordinate transformation matrix T SD :

[0035]

[0036] The construction thereof needs to meet the orthogonality, modal separation and parameter adaptability to ensure the energy conservation in the transformed coordinate system, realize the frequency domain decoupling of the sum mode and the difference mode, and be compatible with the topology reconstruction rule of the virtual converter;

[0037] Further, after the matrix transformation, the output object is:

[0038]

[0039] Through the sum-difference coordinate transformation, the control coordinate model is converted into a diagonal matrix, which means that in the equivalent topology structure, the sum current is controlled only by the sum voltage, and the difference current is controlled only by the difference voltage; finally, through the coordinate transformation of T SD , the diagonal impedance matrix can be obtained:

[0040]

[0041] Wherein:

[0042]

[0043] As preferred in the present application, the implementation process of step S3 is specifically:

[0044] After decoupling by step S2, the original coupled system is decoupled into independent sum current channels and p-1 difference current channels; since there are only N physical converters in the actual system, it is necessary to remap the p-dimensional decoupled model to the N-dimensional physical space through inverse mapping, that is:

[0045] T′ SD I(s)=T′ SD Y(s)(U(s)-u g (s)E)

[0046]

[0047] Through the inverse mapping, the improved sum-difference coordinate transformation matrix T' SD is finally obtained, which is applicable to the non-symmetrical parameter system. ISD :

[0048]

[0049] As preferred in the present application, the implementation process of step S4 is specifically:

[0050] In combination with the control loop, assuming that each converter adopts a uniform delay model G d (s) and current loop controller G c (s), then based on the sum-difference coordinates obtained in step S3, the open-loop transfer functions of the decoupled sum-difference channels can be derived respectively as follows:

[0051]

[0052] On this basis, the closed-loop transfer functions of the sum current loop and the difference current loop are as follows:

[0053]

[0054] As preferred in the present application, the implementation process of step S5 includes:

[0055] During voltage sag, the minimum reactive current specified in the national standard is preferentially delivered, and the maximum active power is output as much as possible under the rated current limit of the inverter;

[0056] Under normal working conditions, the inverter works in the rated state, at which time the required power of the system is calculated to obtain the rated current of each inverter, which is input to the sum-difference coordinate transformation matrix to obtain the sum current and the difference current for each PI controller to provide reference input; wherein the sum current of the active reference current and the difference current are allocated by matrix transformation, while the sum current and the difference current of the reactive reference current and are set to zero.

[0057] When the grid voltage is between 0.2pu and 0.85pu, the sum current and the difference current of the reactive reference current are given by the following formula:

[0058]

[0059] Wherein, I G is the dynamic reactive current injected into the grid by the photovoltaic power generation system, U G is the voltage at the grid-connected point of the photovoltaic power generation system, I N is the rated current of the photovoltaic power generation system, and the unit is ampere (A);

[0060] In addition, the difference current of the reactive reference is allocated according to the different rated powers of each inverter, that is:

[0061]

[0062] in, Let i be the reactive reference current value that should be allocated to the i-th inverter. The reactive reference current value to be allocated to the i-th control current loop, where the value of i is not 1;

[0063] At this point, the system switches to low voltage ride-through mode. If the calculated active power and current reference values ​​are less than the maximum active current under rated current, then... If so, the photovoltaic array will continue to supply power to the grid at maximum active power:

[0064]

[0065] Among them, U gd This represents the per-unit value of the d-axis component of the grid voltage;

[0066] Conversely, the active power and current reference values ​​will switch to the rated current limit. Right now:

[0067]

[0068] Similarly, the formula for calculating the active power differential current reference value is as follows:

[0069]

[0070] in, Let i be the active reference current value that should be allocated to the i-th inverter. This is the active reference current value that should be allocated to the i-th control current loop, where the value of i is not 1.

[0071] As a preferred embodiment of this application, the implementation process of step S6 is as follows:

[0072] When the mains voltage is less than 0.2 pu, Given 0.975×I N , The differential current distribution method for reactive power reference is the same as the calculation method in step S5. and The given value is zero; at this time, the system operates in low voltage ride-through mode to balance system power loss.

[0073] The advantages of this application compared to existing technologies are:

[0074] 1. Decoupling control improves collaborative performance: Decoupling control of multi-inverter clusters is achieved through sum and difference coordinate transformation technology, which solves the coupling interference problem in traditional multi-inverter collaborative control, making the distribution of active and reactive power more accurate, significantly improving the synchronization and stability of multi-unit response during low voltage ride-through, and reducing the risk of oscillation.

[0075] 2. Enhanced parameter adaptability and weak grid compatibility: An innovative equivalent symmetric model for asymmetric inverter clusters is proposed, which transforms the parameter heterogeneous system into a symmetric system, improving its adaptability to weak grid environments and overcoming the shortcomings of existing technologies in terms of insufficient synchronization and poor stability under parameter difference scenarios.

[0076] 3. Dynamic regulation and optimization of power output: Based on the voltage range division (0.2pu to 0.85pu and below 0.2pu), a differentiated control strategy is designed to prioritize meeting reactive power demand while maximizing active power output, fully tapping the reactive power output margin of the inverter, achieving precise support for the grid voltage, and solving the problem of insufficient potential tapping of reactive power support in existing technologies.

[0077] 4. Improved transient performance with fast response: Through the optimized design of the sum and difference current loop PI controller, the tracking speed of the current loop to the reference value is enhanced, and the output can be quickly adjusted during low voltage surges, improving the dynamic response performance of the system, reducing the response delay during sudden deep voltage drops, and enhancing the speed and robustness of fault ride-through.

[0078] 5. Enhanced System Resilience through Global Collaboration: By decoupling the physical space mapping of the model, the virtual model and physical equipment are accurately mapped, ensuring that theoretical control strategies are directly applied to the actual system. This enhances the global collaboration capability of distributed photovoltaic power stations during fault ride-through and provides technical support for building highly resilient new energy power stations. Attached Figure Description

[0079] Figure 1 The diagram shows the topology and control structure of a low-voltage ride-through system for a distributed photovoltaic grid-connected converter based on sum and difference coordinate transformation.

[0080] Figure 2 Topology diagram for n asymmetric inverters;

[0081] Figure 3 The equivalent topology diagram for n asymmetric inverters;

[0082] Figure 4 Here is a block diagram of the PI control for the sum current loop and the differential current loop;

[0083] Figure 5 This is a low-voltage ride-through curve for a photovoltaic power generation system.

[0084] Figure 6 This is a flowchart of the low-voltage ride-through strategy. Detailed Implementation

[0085] To make the technical objectives, technical solutions and beneficial effects of this application clearer, the following detailed explanation of this application is provided in conjunction with the accompanying drawings and embodiments.

[0086] The low-voltage ride-through control method for distributed photovoltaic grid-connected converters based on sum-difference coordinate transformation disclosed in this embodiment has a core innovation in achieving coordinated control of multiple inverter clusters through sum-difference coordinate transformation technology, significantly improving the transient stability and grid connection success rate of distributed photovoltaic systems during low-voltage ride-through. The photovoltaic low-voltage ride-through system topology and control architecture involved in this method are as follows: Figure 1 As shown, the distributed photovoltaic array, after being converted by a DC / DC converter and a DC / AC inverter, is connected in parallel to the microgrid through an inductor filter. By decoupling multiple inverters through sum and difference coordinate transformation, global control of the inverter cluster can be achieved during low voltage ride-through, accurately allocating active and reactive power, fundamentally overcoming the coupling interference problem in traditional multi-inverter collaborative control.

[0087] At the control process level, the sampled three-phase currents are decoupled into sum currents and differential currents through sum-difference coordinate transformation. After independent regulation, SVPWM modulation signals are generated through the sum-difference coordinate inverse transformation matrix. This ultimately enables the inverter cluster to distribute the load according to its rated power under rated operating conditions and to provide reactive power support output during low voltage ride-through, achieving efficient collaborative control of multiple inverters. The method specifically includes the following steps:

[0088] Step S1: Construct an equivalent symmetric model of the asymmetric inverter cluster.

[0089] The topology diagram of n asymmetric inverters is as follows: Figure 2 As shown, where u DC For DC side voltage, i g Let L be the grid current. Considering the heterogeneity of inverter parameters in practical applications, all inverters are assumed to use inductor filtering and are connected to the high-voltage grid through a step-up transformer. The filter inductance of the i-th inverter is defined as L. i =L / n i (i = 1, 2, ..., N), where L is the reference inductance value, and n i Let be the scaling factor of the i-th inverter. According to the superposition theorem, the voltage u at the point of common coupling (PCC) is... pcc It can be represented as:

[0090]

[0091] Among them, u i Z is the output voltage of the i-th inverter. L Let u be the filter impedance of the first filter. g Z is the grid voltage. g Let be the grid impedance, and s be the Laplace operator.

[0092] To construct a unified decoupling model, this step innovatively proposes an equivalent symmetry method for heterogeneous parameter systems: the i-th asymmetric converter is equivalent to n...i A parallel structure of a virtual converter with symmetrical parameters (filter inductor L), corresponding to... Figure 3 The diagram shows the equivalent topology of n asymmetric inverters, while reducing their rated current to 1 / n of their original value. i To satisfy the law of conservation of power:

[0093]

[0094] Among them, i i Let be the output current of the i-th inverter. This process transforms the complex asymmetric system into a symmetric system, which simplifies the modeling complexity and provides a consistent model basis for subsequent unified decoupling control.

[0095] Let the total number of inverters after equivalent transformation be:

[0096]

[0097] After the equivalent transformation, the voltage expression at the common coupling point is:

[0098]

[0099] After obtaining the above expression in the frequency domain, it can be transformed into a time-domain equation using the inverse Laplace transform. A p-dimensional coordinate system is chosen for modeling. Since the output current is controlled by the converter output voltage, the control variable and output variable in the p-dimensional coordinate system are defined as u0 and u1, respectively. j with i j：

[0100] U j =[u1,u2,…,u p ] T

[0101] I j =[i1,i2,…,i p ] T

[0102] Where j = 1, 2, ..., p.

[0103] Modeling the system in a p-dimensional coordinate system, the admittance model of multiple parallel inverters is as follows:

[0104] I(s)=Y(s)(U(s)-u g (s)E)

[0105]

[0106] Here, E is a unit vector, and Y(s) is the transfer function matrix of p equivalent inverters, representing the dynamic coupling effect between virtual converters. This model lays the theoretical framework for subsequent coordinate transformation and decoupling control.

[0107] Step S2: Design the extended dimension and difference coordinate transformation matrix T SD

[0108] Based on the equivalent symmetric system obtained in step S1, and addressing its decoupling requirements, this step designs an extended dimension and difference coordinate transformation matrix T. SD :

[0109]

[0110] The matrix construction strictly satisfies three core criteria: orthogonality and completeness, mode separation and parameter adaptability. Orthogonality and completeness ensure the mathematical rigor of the transformation process, mode separation achieves complete decoupling of modes and difference modes in the frequency domain, and parameter adaptability ensures that the transformation matrix is ​​compatible with the topology reconstruction rules of the virtual converter, ultimately achieving energy conservation in the transformed coordinate system.

[0111] The transformed output object is:

[0112]

[0113] By transforming the sum and difference coordinates, the control coordinate model is converted into a diagonal matrix. In the equivalent topology, the sum current is controlled only by the sum voltage, and the difference current is controlled only by the difference voltage. This completely eliminates the cross-coupling effects in traditional control, significantly improving control accuracy and response speed. SD The diagonalized impedance matrix obtained by the transformation:

[0114]

[0115] in:

[0116]

[0117] This matrix creates ideal conditions for subsequent independent channel control.

[0118] Step S3: Implement physical space mapping of the decoupled model

[0119] This transformation decouples the original coupled system into independent current channels and p-1 differential current channels. Considering that there are only N physical converters in the actual system, the p-dimensional decoupled model needs to be remapped to the N-dimensional physical space through inverse mapping, that is:

[0120] T′ SD I(s)=T′ SD Y(s)(U(s)-u g (s)E)

[0121]

[0122] Obtain the improved sum-difference coordinate transformation matrix T' applicable to asymmetric parametric systems SD With the inverse matrix T' ISD :

[0123]

[0124] This process solves the problem of mapping virtual models to physical devices, ultimately obtaining an improved sum-difference coordinate transformation matrix T' suitable for asymmetric parametric systems. SD With the inverse matrix T' ISD This ensures that theoretical control strategies can be directly applied to actual inverter systems.

[0125] Step S4: Design and implement a PI controller for the differential current loop.

[0126] Assume that each converter adopts a unified delay model G d (s) and current loop controller G c (s), then the open-loop transfer functions of the sum and difference channels after decoupling can be characterized as follows:

[0127]

[0128] The closed-loop transfer functions of the sum current loop and the differential current loop are:

[0129]

[0130] Based on this, a PI controller is designed with a current loop. Figure 4 (a) The differential current loop design PI controller is as follows: Figure 4 (b), k of the sum and differential current loop p The parameters are:

[0131]

[0132] Among them, T s To control the cycle, its reasonable PI controller parameters ensure that the current loop quickly tracks the reference value, enabling rapid output adjustment during low voltage surges, significantly improving the system's dynamic response performance, and providing stable control support for the precise allocation of active and reactive power.

[0133] Step S5: Low voltage ride-through control strategy in the 0.2 pu to 0.85 pu voltage range

[0134] According to the national standard GB / T 29319-2024, when the voltage is at... Figure 5 When a photovoltaic power generation system experiences a low-voltage crossing below curve 1, shutdown and grid disconnection can be considered. During the voltage dip, priority should be given to delivering the minimum reactive current specified by national standards, while simultaneously maximizing active power output within the inverter's rated current limits. Specific strategies are as follows:Figure 6 The low-voltage ride-through strategy flowchart is shown.

[0135] Under normal operating conditions, the inverters operate at their rated state. The rated current of each inverter is calculated based on the system's required power, and this current is input into the sum and difference coordinate transformation matrix to obtain the sum current and the difference currents, thus providing reference inputs to each PI controller; the sum current of the active reference current... and differential current The allocation is performed after matrix transformation, while the reactive power reference and current are... and differential current Set to zero.

[0136] When the grid voltage is between 0.2 pu and 0.85 pu, the reactive power reference and current... Given by the following formula:

[0137]

[0138] Among them, I G The dynamic reactive current U injected into the grid for photovoltaic power generation systems G I represents the per-unit voltage at the grid connection point of the photovoltaic power generation system. N The rated current of the photovoltaic power generation system is expressed in amperes (A).

[0139] The differential current of the reactive power reference is distributed according to its rated power, that is:

[0140]

[0141] in, Let i be the reactive reference current value that should be allocated to the i-th inverter. This is the reactive reference current value that should be allocated to the i-th control current loop, where the value of i is not 1.

[0142] At this point, the system switches to low voltage ride-through mode. If the calculated active power and current reference values ​​are less than the maximum active current under rated current, then... If the formula is correct, the photovoltaic array will continue to supply power to the grid at maximum active power.

[0143]

[0144] Among them, U gd This represents the per-unit value of the d-axis component of the grid voltage.

[0145] Conversely, the active power and current reference values ​​will switch to the rated current limit. Right now:

[0146]

[0147] Similarly, the reference value for the active power differential current is calculated as follows:

[0148]

[0149] in, Let i be the active reference current value that should be allocated to the i-th inverter. This is the active reference current value that should be allocated to the i-th control current loop, where the value of i is not 1.

[0150] Step 6: Low-voltage ride-through control strategy below 0.2 pu voltage range

[0151] When the mains voltage is less than 0.2 pu, Given 0.975×I N , The differential current distribution calculation for reactive power reference is the same as in step 5. and When the value is zero, the system still operates in low voltage ride-through mode. It stabilizes the grid voltage through reactive power compensation, while avoiding equipment overload caused by excessive active power output, effectively balancing system power loss and further improving the success rate of low voltage ride-through.

[0152] The present invention and its embodiments have been described above. This description is not restrictive, and the accompanying drawings are only one embodiment of the present invention; the actual structure is not limited thereto. In conclusion, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the invention, such designs should fall within the protection scope of the present invention.

Claims

1. A low-voltage ride-through control method for a distributed photovoltaic grid-connected converter based on sum and difference coordinate transformation, characterized in that, Includes the following steps: Step S1: Perform impedance modeling and equivalent transformation on the distributed photovoltaic inverter, and construct mutually coupled multidimensional matrix equations to lay the foundation for subsequent decoupling processing; Step S2: For a multi-inverter system with heterogeneous parameters, construct a sum-difference coordinate transformation matrix. Use this matrix to simplify the multidimensional matrix obtained in step S1 into a diagonal matrix, thereby achieving system decoupling. Step S3: Remap the p-dimensional model decoupled in step S2 to the N-dimensional physical space to obtain the equations corresponding to the actual inverter and the sum and difference coordinate transformation matrix adapted to the actual inverter. Step S4: Combining the control loop, derive the accurate expressions for the open-loop and closed-loop output impedances based on the results of step S3; Step S5: According to the low voltage ride-through technology specifications, calculate the active and reactive current setpoints of the system under normal operating conditions and when the grid voltage drops between 0.2 pu and 0.85 pu; Step S6: Calculate the active and reactive current setpoints of the system when the grid voltage drops below 0.2 pu.

2. The low-voltage ride-through control method for a distributed photovoltaic grid-connected converter based on sum and difference coordinate transformation according to claim 1, characterized in that, The implementation process of step S1 includes: First, all n distributed photovoltaic inverters use inductive filtering and are connected to the high-voltage grid through step-up transformers; Furthermore, the filter inductance of the i-th converter is defined as L. i =L / n i (i = 1, 2, ..., N), where L is the reference inductance value, n i The scaling factor for the i-th inverter; Based on this, the voltage u at the common coupling point is established using the root superposition theorem. pcc The expression: Among them, u i Z is the output voltage of the i-th inverter. L Let u be the filter impedance of the first filter. g Z is the grid voltage. g Let be the grid impedance, and s be the Laplace operator; To construct a unified decoupling model, an equivalent symmetry transformation method for parameter heterogeneous systems is proposed: the i-th asymmetric converter is equivalent to n i A parallel structure of virtual converters with symmetrical parameters, where the rated current is reduced to 1 / n of its original value. i To satisfy the power conservation, that is: Among them, i i Let i be the output current of the i-th inverter; Let the total number of inverters after equivalent transformation be: After obtaining the above expression in the frequency domain, it is transformed into a time-domain equation using the inverse Laplace transform, and then a p-dimensional coordinate system is chosen for modeling. Since the output current is controlled by the converter output voltage, the control variable and output variable in the p-dimensional coordinate system are defined as u, respectively. j with i j : U j =[u1,u2,…,u p ] T I j =[i1,i2,…,i p ] T Where j = 1, 2, ..., p; Finally, by modeling the system in a p-dimensional coordinate system, the admittance model of multiple parallel inverters is as follows: I(s)=Y(s)(U(s)-u g (s)E) Where E is a unit vector and Y(s) is the transfer function matrix of p equivalent inverters, representing the dynamic coupling effect between virtual converters.

3. The low-voltage ride-through control method for a distributed photovoltaic grid-connected converter based on sum and difference coordinate transformation according to claim 2, characterized in that, The implementation process of step S2 specifically includes: First, to address the decoupling requirements of the equivalent symmetric system, an extended dimension sum-difference coordinate transformation matrix T is designed. SD : Its construction must satisfy orthogonality, mode separation and parameter adaptability to ensure energy conservation in the transformed coordinate system, realize frequency domain decoupling of the main mode and the differential mode and be compatible with the topology reconstruction rules of the virtual converter; Therefore, after this matrix transformation, the output object is: By transforming the sum and difference coordinates, the control coordinate model is converted into a diagonal matrix, which means that in the equivalent topology, the sum current is controlled only by the sum voltage, and the difference current is controlled only by the difference voltage; finally, through T SD By performing a coordinate transformation, we can obtain the diagonalized impedance matrix: in:

4. The low-voltage ride-through control method for a distributed photovoltaic grid-connected converter based on sum and difference coordinate transformation according to claim 3, characterized in that, The specific implementation process of step S3 is as follows: After decoupling in step S2, the original coupled system is decoupled into independent sum current channels and p-1 differential current channels. Since there are only N physical converters in the actual system, it is necessary to remap the p-dimensional decoupling model to the N-dimensional physical space through inverse mapping, that is: T′ SD I(s)=T′ SD Y(s)(U(s)-u g (s)E) Through this inverse mapping, the improved sum-difference coordinate transformation matrix T' suitable for asymmetric parametric systems is finally obtained. SD With the inverse matrix T' ISD :

5. The low-voltage ride-through control method for a distributed photovoltaic grid-connected converter based on sum and difference coordinate transformation according to claim 4, characterized in that, The specific implementation process of step S4 is as follows: In conjunction with the control loop, it is assumed that each converter adopts a unified delay model G. d (s) and current loop controller G c Based on the sum and difference coordinates obtained in step S3, the open-loop transfer functions of the decoupled sum and difference channels can be derived respectively: Based on this, the closed-loop transfer functions of the sum current loop and the differential current loop are:

6. The low-voltage ride-through control method for a distributed photovoltaic grid-connected converter based on sum and difference coordinate transformation according to claim 5, characterized in that, The implementation process of step S5 includes: During voltage dips, priority is given to delivering the minimum reactive current specified by national standards, and the maximum active power is output as much as possible within the inverter's rated current limit. Under normal operating conditions, the inverters operate at their rated state. At this time, the rated current of each inverter needs to be calculated based on the system's required power. This current is then input into the sum-difference coordinate transformation matrix to obtain the sum current and the differential currents, providing reference inputs for each PI controller. Among these, the sum current of the active reference current... and differential current The allocation is performed after matrix transformation, while the reactive power reference and current are... and differential current Set to zero; When the grid voltage is between 0.2 pu and 0.85 pu, the reactive power reference and current... Given by the following formula: Among them, I G The dynamic reactive current U injected into the grid for photovoltaic power generation systems G I represents the per-unit voltage at the grid connection point of the photovoltaic power generation system. N The rated current of the photovoltaic power generation system is expressed in amperes (A). Furthermore, the differential current for reactive power reference is distributed according to the different rated power of each inverter, that is: in, Let i be the reactive reference current value that should be allocated to the i-th inverter. The reactive reference current value to be allocated to the i-th control current loop, where the value of i is not 1; At this point, the system switches to low voltage ride-through mode. If the calculated active power and current reference values ​​are less than the maximum active current under rated current, then... If so, the photovoltaic array will continue to supply power to the grid at maximum active power: Among them, U gd This represents the per-unit value of the d-axis component of the grid voltage; Conversely, the active power and current reference values ​​will switch to the rated current limit. Right now: Similarly, the formula for calculating the active power differential current reference value is as follows: in, Let i be the active reference current value that should be allocated to the i-th inverter. This is the active reference current value that should be allocated to the i-th control current loop, where the value of i is not 1.

7. The low-voltage ride-through control method for a distributed photovoltaic grid-connected converter based on sum and difference coordinate transformation according to claim 6, characterized in that, The specific implementation process of step S6 is as follows: When the mains voltage is less than 0.2 pu, Given 0.975×I N , The differential current distribution method for reactive power reference is the same as the calculation method in step S5. and The given value is zero; at this time, the system operates in low voltage ride-through mode to balance system power loss.

Images