A frequency-voltage coordinated support control method for distributed photovoltaic layered clusters

Abstract

The present invention discloses a distributed photovoltaic layered cluster frequency-voltage collaborative support control method. This method considers the impact of distributed photovoltaic participation in system frequency regulation on the distribution network voltage, and realizes distributed photovoltaic cluster division based on non-negative matrix decomposition technology; constructs a large-scale distributed photovoltaic layered cluster coordinated control scheme: in the global optimization layer, a distributed photovoltaic active / frequency and reactive / voltage local droop control parameter optimization model is established, which is decomposed into multiple sub-models and coordinated optimization is performed based on the alternating direction multiplier method; in the local coordination layer, the voltage self-correction control within the cluster is used to alleviate the local voltage limit of the distribution network; in the local control layer, the distributed photovoltaic droop control parameters are configured based on the calculation results of the global optimization layer, and then the distributed photovoltaic combines the droop control strategy to provide fast frequency-voltage collaborative support. The method of the present invention can efficiently coordinate large-scale distributed photovoltaics, provide frequency support for the transmission network, and alleviate the problem of voltage limit exceeding the distribution network.

Description

Technical Field

[0001] The present invention relates to a distributed photovoltaic layered cluster frequency-voltage collaborative support control method, belonging to the field of power system operation control. Background Art

[0002] The massive integration of distributed photovoltaic systems poses a serious threat to system operational safety. Distributed photovoltaic systems offer flexible power regulation, enabling them to simultaneously participate in system frequency regulation and local voltage control. Efficient coordinated control of large-scale distributed photovoltaic systems to fully tap and utilize their grid support potential is crucial for improving global frequency stability and ensuring local voltage safety within the distribution network.

[0003] In existing technical solutions, distributed photovoltaic active and reactive output is generally changed to participate in system frequency control and voltage regulation. However, existing control technologies or solutions have the following problems: Distributed photovoltaic systems have small single-unit capacity, large number, and wide distribution. Distributed photovoltaic coordinated control problems are characterized by high dimensionality and strong nonlinearity, making it difficult to achieve efficient coordinated control with traditional centralized control schemes. Existing zoning / hierarchical control schemes do not consider the coupling problem caused by distributed photovoltaics participating in both frequency and voltage control, and cannot guarantee the distributed photovoltaic power grid support effect; Random fluctuations in distributed photovoltaic output power and load demand lead to rapid changes in the system operating state. Traditional optimization control schemes have good effects but slow response speeds, while pure local control has fast response speeds but poor adaptability, making it impossible to achieve optimal coordinated control of distributed photovoltaics on multiple time scales. Summary of the Invention

[0004] In view of the limitations of the relevant background technology, the present invention provides a distributed photovoltaic layered cluster frequency-voltage collaborative support control method. The method proposes a distributed photovoltaic layered cluster coordinated control scheme of "global optimization-local coordination-local control", which collaboratively supports system frequency stability and distribution network voltage security through coordinated control of large-scale distributed photovoltaics. The method of the present invention takes into account the impact of distributed photovoltaic participation in system frequency regulation on distribution network voltage, improves the electrical distance index, and realizes cluster division based on non-negative matrix decomposition, thereby improving the rationality of distributed photovoltaic cluster division. In order to improve the adaptability of droop control, a distributed photovoltaic active / frequency and reactive / voltage droop control parameter optimization model is constructed in the global optimization layer, and inter-cluster coordinated optimization is performed based on the alternating direction multiplier method, and the droop control parameters are optimized and adjusted on an hourly time scale. Furthermore, in response to the problem of local voltage exceeding the limit in the distribution network caused by rapid changes in distributed photovoltaic power generation power or load demand, a cluster voltage self-correction control strategy is applied in the local coordination layer to achieve rapid adjustment of local voltage through a small amount of calculation. Finally, at the local control layer, combined with the calculation results from the global optimization layer, distributed photovoltaic active power / frequency and reactive power / voltage droop control parameters are configured, providing real-time active frequency-voltage support based on droop control. Through hierarchical cluster coordinated control, efficient multi-timescale coordinated control of large-scale distributed photovoltaics is achieved, providing aggregated frequency support for the transmission network (i.e., by adjusting the output power of multiple distributed photovoltaics, power at the distribution network substation outlet is regulated, thereby providing frequency modulation power support for the transmission network), while also alleviating the problem of voltage over-limit in the distribution network under complex operating conditions.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A distributed photovoltaic layered cluster frequency-voltage coordinated support control method is provided. The method is a large-scale distributed photovoltaic voltage-frequency layered coordinated support control method comprising a "global optimization layer-local coordination layer-local control layer", specifically comprising the following steps:

[0007] 1) Considering the impact of distributed photovoltaic participation in system frequency regulation on distribution network voltage, distributed photovoltaic clustering is performed based on non-negative matrix factorization technology to simultaneously integrate network topology and node attribute information, thereby improving the rationality of distributed photovoltaic clustering results;

[0008] 2) Build a hierarchical cluster coordinated control scheme consisting of a global optimization layer, a local coordination layer, and a local control layer to achieve hourly, minute-by-minute, and real-time multi-timescale coordinated control of distributed photovoltaics, thereby providing aggregated frequency support and alleviating voltage over-limit issues in the distribution network;

[0009] 3) At the global optimization layer, with the goal of minimizing the loss of distributed photovoltaic power generation benefits, reactive power usage, and network losses, and with the distribution network power flow equation, distributed photovoltaic operating characteristics, and local droop control function as constraints, an optimization model for the active power / frequency and reactive power / voltage local droop control parameters is constructed;

[0010] 4) Based on the cluster division results, frequency modulation-related constraints in the active / frequency and reactive / voltage local droop control parameter optimization model are reconstructed, and consistency constraints are added, thereby decomposing the active / frequency and reactive / voltage local droop control parameter optimization model into multiple sub-models. All sub-models are then coordinated and optimized based on the alternating direction multiplier method to reduce the computational burden, thereby ultimately achieving fair distribution of frequency modulation power and optimal design of local reactive / voltage droop control parameters, and obtaining optimized active / frequency and reactive / voltage droop control parameters;

[0011] 5) At the local coordination layer, for distributed photovoltaic clusters with voltage exceeding the limit, voltage self-correction control is achieved within the distributed photovoltaic cluster with voltage exceeding the limit by modifying the reactive power reference value of the distributed photovoltaic reactive power / voltage droop control within the cluster;

[0012] 6) At the local control layer, based on the parameters calculated at the global optimization layer (active power / frequency droop control gain, local reactive power / voltage control gain, voltage dead zone, and saturation parameters), the active power / frequency and reactive power / voltage droop control parameters of each distributed photovoltaic system are configured. Then, each distributed photovoltaic system combines local frequency and voltage measurements to quickly respond to frequency and voltage deviations, achieving immediate frequency and voltage coordinated support.

[0013] In the above technical solution, further, in step 1), the impact of distributed photovoltaic participation in system frequency regulation on the distribution network voltage is considered, the electrical distance index is improved, and the distributed photovoltaic is clustered based on the non-negative matrix decomposition technology. The specific implementation process is as follows:

[0014] Assume that the number of network nodes is , the electrical distance index between nodes can be defined based on the reactive-voltage sensitivity of the distribution network:

[0015]

[0016] Where: Representation node and Electrical distance indicators between; and Represent the reactive power-voltage sensitivity matrix of the distribution network No. and diagonal elements; express No. Rank Since distributed photovoltaic active power adjustment has a significant impact on node voltage, this impact must be considered when dividing clusters. Assume that the distribution network aggregate support power (i.e., the power adjustment at the distribution network substation outlet) is , and each distributed photovoltaic power plant bears the frequency regulation task fairly according to its capacity, then the node Distributed photovoltaic participation in frequency regulation of distribution network nodes The effect of voltage can be expressed as:

[0017]

[0018] Where: Representation node Upper voltage changes; Is a constant used to quantize nodes The impact of injection power adjustment on substation outlet power; Indicates that it is located at the node The distributed photovoltaic injection unit power to the node The magnitude of the voltage effect; represents the active-voltage sensitivity matrix, For its Rank Column elements; Representation node Based on this, the electrical distance index can be modified as follows:

[0019]

[0020] Where: Indicates the improved electrical distance indicator; Indicates that it is located at the node The distributed photovoltaic injection unit power to the node The impact of voltage, and Respectively represent nodes The distributed photovoltaic injection unit power to the node and nodes The magnitude of the voltage effect.

[0021] The node-edge adjacency matrix of the network Elements The corresponding expression is:

[0022]

[0023] Where: It represents the set of nodes in the distribution network except the reference node (i.e. the node at the substation exit).

[0024] The diagonal elements of the voltage sensitivity matrix reflect the impact of each distributed photovoltaic power adjustment on the local voltage, and the linearization coefficient It approximately represents the contribution of distributed photovoltaic active power regulation to the distribution network aggregate support power. Without loss of generality, the distribution network node attribute matrix is Defined as:

[0025]

[0026] Where: is a constant coefficient vector; It means taking the diagonal elements of the matrix and forming a column vector.

[0027] Node similarity matrix It can be constructed in combination with cosine similarity, and its elements are , representing a node and nodes The similarity between them can be expressed as:

[0028]

[0029] Where: and They are node attribute matrices No. and column vectors.

[0030] To find The cluster partitioning task is constructed as a non-negative matrix factorization problem by considering the node-edge adjacency matrix of the network. and node similarity matrix Decompose to obtain the node-cluster membership matrix Node-cluster affiliation matrix No. Rank Columns represent distribution network nodes For clusters Node-cluster membership matrix The row number corresponding to the maximum element in each column is the number of the cluster to which the corresponding node belongs. The non-negative matrix factorization problem for distributed photovoltaic cluster partitioning is expressed as:

[0031]

[0032] Where: and are two non-negative fundamental matrices; represents the Frobenius norm; is a positive real parameter used to balance the contribution of topology and attribute information; the third term of the objective function is a regularization term used to control the number of nodes in each cluster. is a regularization constant; Representation matrix No. elements, represents the total membership of each cluster; is a constant vector whose elements are all 1.

[0033] Furthermore, in step 2), based on the hierarchical cluster coordinated control scheme, the specific method for achieving hour-minute-real-time multi-time scale coordinated control of distributed photovoltaics is as follows:

[0034] The hierarchical cluster coordinated control scheme includes three control levels: global optimization layer, local coordination layer and local control layer; the global optimization layer performs distributed optimization once an hour to adjust the active power / frequency and reactive power / voltage droop control parameters of distributed photovoltaics; the local coordination layer performs cluster voltage self-correction control on a minute-level time scale, and controls the reactive power output of distributed photovoltaics in the cluster by modifying the reactive power reference value of the reactive power / voltage droop control of distributed photovoltaics in the cluster to alleviate the distribution network voltage over-limit; in the local control layer, the active power / frequency and reactive power / voltage droop control parameters of each distributed photovoltaic are configured in combination with the optimization design results of the global optimization layer. Distributed photovoltaics respond to system frequency and voltage deviations in real time based on local frequency and voltage measurements; through the mutual cooperation of the three control levels, coordinated control of distributed photovoltaics at multiple time scales is achieved, thereby providing aggregated frequency support for the system and effectively alleviating the distribution network voltage over-limit problem caused by fluctuations in distributed photovoltaic power generation and load demand.

[0035] Furthermore, in step 3), the active power / frequency and reactive power / voltage local droop control parameter optimization model can be specifically expressed as:

[0036]

[0037]

[0038]

[0039]

[0040]

[0041]

[0042]

[0043]

[0044]

[0045]

[0046]

[0047]

[0048]

[0049]

[0050]

[0051]

[0052]

[0053]

[0054]

[0055]

[0056]

[0057]

[0058]

[0059] Where: The objective function of the active power / frequency and reactive power / voltage local droop control parameter optimization model; , and are non-negative weight factors for distributed photovoltaic active reserve, reactive power and network loss, and satisfy ; Subscript Indicates the moment; and Respectively Time Node Upward and downward reserve power; for The reactive power adjustment of the inverter at the moment; Indicates line The square of the upper current amplitude; Indicates line resistance; Represents a collection of distribution network lines; Represents the set of nodes where distributed photovoltaics are installed; To optimize the time domain; and Represents each line separately There are active and reactive currents on the upper side; Representation and Node The set of downstream nodes connected by lines; and Represents nodes respectively Active and reactive power injected on Representation node The square of the voltage amplitude, For nodes Voltage amplitude; Indicates line reactance; and Respectively represent the upper and lower limits of voltage; and They are Time Node The maximum available and output active power of distributed photovoltaics; and are the maximum allowable upward and downward active reserve rates respectively; and Respectively Time Node The inverter output reactive power and reactive reference value; Indicates the distributed photovoltaic reactive power adjustment amount; express Time Node The maximum available reactive power of the inverter; and Respectively The participation of distributed photovoltaic upward and downward frequency regulation at all times; and Respectively Time Node Up and down FM gain; and They are The column vector composed of the up / down frequency modulation gains of each distributed photovoltaic at the moment; and They are Aggregate active power / frequency upward and downward frequency regulation gains expected by the transmission network operator at any given moment; and They are Distributed photovoltaics at all times The corresponding upper and lower limits of the frequency dead zone; and They are Distributed photovoltaics at all times The corresponding upper and lower limits of the frequency saturation threshold; and Respectively Time Node Upward and downward voltage droop control gains; and They are Time Node Upper and lower saturation thresholds for reactive / voltage droop control; and They are Time Node The upper and lower limits of the dead zone of reactive power / voltage droop control; is the frequency measurement value; They are Time Node 0-1 indicator variables related to the five linear segments of the reactive power / voltage droop control function; is a large constant; express Time Node The reactive power reference command that may be output by the distributed photovoltaic reactive power / voltage droop control; express Time Node The voltage droop controls the reactive power reference command of the actual output; represents the Hadamard product; for No. elements; and They are the upper and lower limits of voltage dead zone and saturation parameter respectively; is the scaling factor.

[0060] Furthermore, in step 4), based on the cluster division result, the constraints related to frequency regulation in the active power / frequency and reactive power / voltage local droop control parameter optimization model are reconstructed, and consistency constraints are added. The specific process is as follows:

[0061] Assume that the cluster division method described in step 1) divides the distributed photovoltaic into clusters, the distribution network can also be divided into Here, the tie line segmentation method is used to divide the distribution network. If there is a common line between two clusters m and n , then in A virtual node is introduced in the middle and the line is disconnected at the virtual node. The entire network is divided into subnets m and n, each of which has a contact line and a boundary node. In distributed optimization, subnets and Each has an optimization model, called a sub-model. The impedance value of each interconnection line is the common line half of , then the consistency constraint between subnets m and n is:

[0062]

[0063] Where: and Respectively Time by subnet Calculated flow through subnet and Active power and reactive power on the interconnection line; and Respectively Time by subnet Calculated flow through subnet and Active power and reactive power on the interconnection line; and They are Time by subnet and subnet Calculated voltages at boundary nodes.

[0064] Thus, the constraints related to frequency regulation in the active / frequency and reactive / voltage local droop control parameter optimization model are reconstructed (i.e., the equation constraints corresponding to the active / frequency droop control gains: ), specifically expressed as:

[0065]

[0066]

[0067] Where: and Respectively The downward and upward frequency regulation reference instructions issued by the transmission network operator at all times; and Respectively Time from subnet Send to Subnet Downward and upward frequency modulation reference instructions; and They respectively represent Time from subnet Send to Subnet Downward and upward frequency modulation reference instructions; represents the linearization constant coefficient corresponding to subnet m, which can be calculated separately for each subnet; and Represents subnets respectively The set of upstream and downstream subnets; It is a 0-1 indicator variable, 1 means that subnet m is connected to the substation, and the subnet receives the frequency regulation command issued by the transmission network operator, otherwise the subnet Only receive FM commands sent by adjacent subnets; Indicates a subnet A collection of nodes where distributed photovoltaics are installed.

[0068] At the same time, additional consistency constraints are added. The purpose of adding consistency constraints is to make the tie line power, boundary node voltage, frequency regulation reference instruction and distributed photovoltaic frequency regulation participation of each subnet the same to ensure that the calculation results are physically feasible. The specific expression is:

[0069]

[0070]

[0071] Where: and Respectively Time by subnet Calculate and send to subnet Downward and upward frequency modulation reference instructions; and Respectively Time by subnet Calculate and send to subnet Downward and upward frequency modulation reference instructions; and They represent the participation of distributed photovoltaic in subgrid m in upward and downward frequency regulation respectively; and Subnet The calculated participation of distributed photovoltaics in upward and downward frequency regulation.

[0072] The vector consisting of tie-line power, boundary node voltage, frequency regulation reference command, and distributed photovoltaic frequency regulation participation is defined as a coupling variable. The above consistency constraint can be rewritten into a vector form, specifically expressed as:

[0073]

[0074] Where: Representation and subnet The vector composed of the relevant tie-line power, boundary node voltage, frequency regulation reference command and distributed photovoltaic frequency regulation participation is called coupling variable; similarly, Representation and subnet A vector consisting of the relevant variables: tie line power, boundary node voltage, frequency regulation reference command, and distributed photovoltaic frequency regulation participation; and Subnet and subnet Calculated coupling variables and value.

[0075] Furthermore, in step 4), the objective function of each sub-model is:

[0076]

[0077] Where: Represents the objective function of the sub-model corresponding to sub-network m; , and Subnet The non-negative weight factors of distributed photovoltaic active reserve, reactive power and network loss, and satisfy ; Represents the set of lines in subnet m; Specifically, by solving the subnet The corresponding sub-model is obtained; is the Lagrange multiplier vector; represents the penalty coefficient. In addition to the equality constraints corresponding to the active power / frequency droop control gains, the constraints in the active power / frequency and reactive power / voltage local droop control parameter optimization models are all included in the constraints of each submodel. Furthermore, each submodel constraint also includes the aforementioned consistency constraints.

[0078] Furthermore, in step 4), all sub-models are coordinated and optimized based on the alternating direction multiplier method. The specific iterative process is:

[0079]

[0080] Where: and Represent sub-model m and sub-model respectively The feasible decision space is composed of the constraints of each sub-model; and Representation submodel and submodels Optimization variables of Indicates the current moment; and Respectively represent Sub-iteration model and submodels The calculated coupling variable values, the superscript indicates the number of iterations; and Represents sub-models and In the The optimization result obtained by the iteration; and Respectively expressed in The sub-model in the iterative calculation process and submodels The Lagrange multiplier vector of ; Indicates in The sub-model in the iterative calculation process The penalty coefficient.

[0081] Definition After the iteration calculation, the sub-model The original residual and the dual residual They are:

[0082]

[0083] Where: Is with subnet Related coupled variables; Indicates that the subnet Calculate and send to subnet The value of the coupling variable; and are the Lagrange multipliers calculated for the current iteration step and the previous iteration step respectively.

[0084] Rule No. The judgment condition of whether the algorithm converges after iterations is expressed as:

[0085] ,

[0086] ,

[0087] ,

[0088] Where: and Respectively represent The residual vector composed of the original residuals and dual residuals of all sub-models after iterations; is the maximum allowable residual of the alternating direction multiplier method, which can be set to 1×10 -3 ; represents the infinite norm; Represents the number of sub-models, that is, the number of distributed photovoltaic clusters.

[0089] In order to accelerate iterative convergence and reduce the number of iterations, a penalty coefficient dynamic update strategy is introduced. As an example, the corresponding penalty coefficient dynamic update strategy is specifically expressed as follows:

[0090]

[0091] Where: is the adjustment factor; is the balance factor; represents the two-norm.

[0092] Furthermore, in step 5), the reactive reference value of the internal distributed photovoltaic reactive power / voltage droop control is modified to adjust the distributed photovoltaic reactive output, thereby achieving voltage self-correction control within the distributed photovoltaic cluster with voltage exceeding the limit. The correction amount of the reactive reference value of the distributed photovoltaic reactive power / voltage droop control is derived as follows:

[0093] Assuming that the output power of distributed photovoltaic increases, the cluster Internal Node Voltage by Rise to , and each reactive power / voltage droop controller operates in the linear region, the distributed photovoltaic will change the reactive power output, and the steady-state voltage is expressed as:

[0094]

[0095] Where: Represents distributed photovoltaic Reactive power / voltage droop control gain; and Represents nodes respectively Initial voltage and voltage after disturbance when no control is applied; It means that after reactive power / voltage droop control, the node The steady-state voltage on .

[0096] If the steady-state voltage Exceeding the safe upper limit of distribution network voltage or lower limit , such as 1.05 pu or 0.95 pu, the distributed photovoltaic reactive power / voltage droop control must be adjusted to compensate for the reactive power and adjust the voltage to a safe range. When the voltage exceeds the limit, the reactive reference value of the reactive power / voltage droop control can be modified. , the droop control curve is shifted downward along the Q axis, so that the distributed photovoltaic inverter absorbs more reactive power. The actual reactive power compensation required to maintain voltage safety is:

[0097]

[0098]

[0099] Where: Indicates the voltage correction amount; Indicates the actual reactive power compensation amount. , then the reduction in distributed photovoltaic reactive output is It can be calculated according to the following formula:

[0100]

[0101] Therefore, the correction amount of the reactive reference value of distributed photovoltaic reactive / voltage droop control is Should meet the following requirements:

[0102]

[0103] Where: Indicates the correction amount of the reactive power reference value for distributed photovoltaic reactive power / voltage droop control.

[0104] The above analysis is only for the case where a single distributed photovoltaic participates in distribution network voltage regulation. When multiple distributed photovoltaics participate in voltage regulation at the same time, considering the interaction between multiple distributed photovoltaics, the correction amount of the reactive power reference value of each distributed photovoltaic reactive power / voltage droop control can be calculated by the following formula:

[0105]

[0106] Where: Indicates the droop control gain of each distributed photovoltaic in the cluster The vector of components; The diagonal elements are The diagonal matrix of ; ; Represents a distributed photovoltaic cluster The number of internal nodes.

[0107] By calculating the correction amount of the reactive reference value of distributed photovoltaic reactive power / voltage droop control It is then sent to each distributed photovoltaic in the cluster to modify the reactive reference value of each distributed photovoltaic reactive / voltage droop control, thereby adjusting the distributed photovoltaic reactive output to solve the voltage limit problem in the cluster.

[0108] Finally, in step 5), the local control layer configures the active / frequency and reactive / voltage droop control parameters of each distributed photovoltaic system based on the parameters calculated by the global optimization layer. The local frequency and voltage values are measured in real time. Once the frequency or voltage values exceed the frequency or voltage dead zone, the distributed photovoltaic system will quickly adjust its active or reactive output according to the active / frequency or reactive / voltage droop control, thereby participating in system frequency regulation and distribution network voltage control in real time. The active / frequency and reactive / voltage droop control of the local control layer is specifically expressed as follows:

[0109]

[0110]

[0111] Where: and Represents distributed photovoltaic exist Real-time active and reactive adjustments corresponding to frequency and voltage deviations; express The measured value of system frequency at the moment; express Time Node Voltage measurement value.

[0112] The beneficial effects of the present invention are:

[0113] To meet the needs of large-scale distributed photovoltaic (PV) frequency-voltage coordinated control, this paper proposes a distributed PV cluster partitioning and hierarchical coordinated control method. Considering the impact of distributed PV participation in system frequency regulation on distribution network voltage, and integrating network topology and node attribute information, a non-negative matrix factorization technique is used to achieve a reasonable partitioning of distributed PV clusters. Furthermore, a hierarchical cluster coordinated control approach based on "global optimization, local coordination, and local control" is employed to achieve efficient multi-timescale coordinated control and frequency-voltage coordinated support for large-scale distributed PV. Specifically, at the global optimization layer, parameter optimization models for reactive power / voltage and active power / frequency droop control are constructed, and multiple clusters are coordinated using the alternating direction multiplier method to balance control speed and optimality. At the local coordination layer, an intra-cluster voltage self-correction control strategy is proposed. This strategy modifies the reactive power reference values of distributed PV reactive power / voltage droop control to change the distributed PV active and reactive power output, thereby addressing local voltage over-limit issues caused by real-time source-load fluctuations. At the local control layer, distributed PV rapidly responds to system frequency and voltage deviations based on optimized and adjusted active power / frequency and reactive power / voltage droop control. BRIEF DESCRIPTION OF THE DRAWINGS

[0114] Figure 1 This is an implementation process of a distributed photovoltaic layered cluster frequency-voltage coordinated support control method according to an embodiment of the present invention;

[0115] Figure 2 The control effects of the method of the present invention and the existing control scheme on the improved IEEE 33-node test system are compared, where (a) is the comparison of frequency control effects and (b) is the comparison of voltage control effects. DETAILED DESCRIPTION

[0116] The present invention is further described below with reference to the accompanying drawings and embodiments. The embodiment of the present invention provides a distributed photovoltaic layered cluster frequency-voltage collaborative support control method process as follows: Figure 1 shown.

[0117] A distributed photovoltaic layered cluster frequency-voltage coordinated support control method specifically includes the following steps:

[0118] (1) Considering the impact of distributed photovoltaics participating in system frequency regulation on distribution network voltage, the electrical distance index is improved. The specific method is to initialize basic data such as distribution network topology, line parameters, and source-load data, obtain distribution network voltage sensitivity information, and calculate the improved electrical distance index to form an adjacency matrix and a node attribute matrix; and use non-negative matrix decomposition technology to cluster distributed photovoltaics, as follows:

[0119] Assume that the number of distribution network nodes is , the non-negative matrix factorization problem of distributed photovoltaic cluster partitioning is expressed as:

[0120] ,

[0121] ,

[0122] ,

[0123] ,

[0124] ,

[0125] ,

[0126] ,

[0127] Where: Indicates the number of distributed photovoltaic clusters; Indicates the number of network nodes; Represents the set of nodes in the distribution network except the reference node (i.e. the node at the substation exit); and are two non-negative fundamental matrices; represents the Frobenius norm; is a positive real parameter used to balance the contribution of topology and attribute information; the third term of the objective function is a regularization term used to control the number of nodes in each cluster. is a regularization constant; Representation matrix No. elements, represents the total membership of each cluster; is a constant vector whose elements are all 1; represents the node-cluster membership matrix; represents the node similarity matrix; Represents the distribution network node attribute matrix; and They are the distribution network node attribute matrices No. and column vectors; is a constant coefficient vector whose elements For quantization nodes The impact of injection power adjustment on substation outlet power; It means taking the diagonal elements of the matrix and forming a column vector; Represents the node-edge adjacency matrix of the network, whose elements are , representing a node and nodes Electrical connections between Indicates the improved electrical distance indicator; Is a constant used to quantize nodes The impact of injection power adjustment on substation outlet power; and Respectively represent nodes The distributed photovoltaic injection unit power to the node and nodes The impact of voltage, and Respectively represent nodes The distributed photovoltaic injection unit power to the node and nodes The magnitude of the voltage effect; represents the active-voltage sensitivity matrix, For its Rank Column elements; Representation node The installed capacity of distributed photovoltaics; Indicates reactive-voltage sensitivity; Representation node and Electrical distance indicators between; and Respectively No. and diagonal elements; express No. Rank Column element.

[0128] (2) A hierarchical cluster coordinated control scheme consisting of a global optimization layer, a local coordination layer, and a local control layer is constructed to achieve multi-time-scale coordinated control of distributed photovoltaics to provide fast frequency support and alleviate the problem of voltage exceeding the limit in the distribution network.

[0129] The hierarchical cluster coordinated control scheme includes three control levels: global optimization layer, local coordination layer and local control layer. The global optimization layer performs distributed optimization once an hour to optimize the active power / frequency and reactive power / voltage droop control parameters of distributed photovoltaics. The local coordination layer performs voltage self-correction control within the cluster on a minute-level time scale. By modifying the reactive reference value of the reactive power / voltage droop control of distributed photovoltaics within the cluster, the reactive power output of distributed photovoltaics within the cluster is adjusted, thereby alleviating the voltage over-limit of the distribution network. In the local control layer, the active power / frequency and reactive power / voltage droop control parameters of each distributed photovoltaic are configured in combination with the optimization design results of the global optimization layer. Distributed photovoltaics respond to system frequency and voltage deviations in real time based on local frequency and voltage measurements. Through the mutual cooperation of the three control levels, distributed photovoltaic coordinated control at multiple time scales is achieved.

[0130] (3) In the global optimization layer, with the goal of minimizing the loss of distributed photovoltaic power generation benefits, the use of reactive power, and network losses, and with the distribution network flow equation, distributed photovoltaic operation characteristics, and local droop control function as constraints, an optimization model for the active / frequency and reactive / voltage local droop control parameters is constructed.

[0131] The optimization model of active power / frequency and reactive power / voltage local droop control parameters is as follows:

[0132]

[0133]

[0134]

[0135]

[0136]

[0137]

[0138]

[0139]

[0140]

[0141]

[0142]

[0143]

[0144]

[0145]

[0146]

[0147]

[0148]

[0149]

[0150]

[0151]

[0152]

[0153]

[0154]

[0155] Where: The objective function of the active power / frequency and reactive power / voltage local droop control parameter optimization model; , and are all non-negative weight factors for distributed photovoltaic active backup, reactive power and network loss, and satisfy ; Subscript Indicates the moment; and Respectively represent upward and downward reserve power; is the reactive power adjustment value of the inverter; Indicates line The square of the upper current amplitude; Indicates line resistance; Represents a collection of distribution network lines; Represents the set of nodes where distributed photovoltaics are installed; To optimize the time domain; and Represents each line separately There are active and reactive currents on the upper side; Representation and Node The set of downstream nodes connected by lines; and Represents nodes respectively Active and reactive power injected on Representation node The square of the voltage amplitude, For nodes Voltage amplitude; Indicates line reactance; and Respectively represent the upper and lower limits of voltage; and are the maximum available and output active power of distributed photovoltaics, respectively; and are the maximum allowable upward and downward active reserve rates respectively; and Respectively represent the inverter output reactive power and reactive reference value; Indicates the distributed photovoltaic reactive power adjustment amount; Indicates the maximum available reactive power of the inverter; and They represent the participation of distributed PV in upward and downward frequency regulation respectively; and Respectively represent the upward and downward FM gains; and are column vectors composed of up / down frequency modulation gains of each distributed photovoltaic; and are the aggregated active power / frequency upward and downward frequency regulation gains expected by the transmission grid operator, respectively; and They are Distributed photovoltaics at all times The corresponding upper and lower limits of the frequency dead zone; and They are Distributed photovoltaics at all times The corresponding upper and lower limits of the frequency saturation threshold; and Respectively represent the upward and downward voltage droop control gains; and are the upper and lower saturation thresholds for reactive / voltage droop control respectively; and They are the upper and lower limits of the dead zone for reactive power / voltage droop control respectively; is the frequency measurement value; are 0-1 indicator variables related to the five linear segments of the reactive / voltage droop control function; is a large constant; Indicates the reactive power reference command that may be output by distributed photovoltaic reactive power / voltage droop control; Indicates the reactive power reference command actually output by voltage droop control; represents the Hadamard product; for No. elements; and They are the upper and lower limits of voltage dead zone and saturation parameter respectively; is the scaling factor.

[0156] (4) Based on the cluster division results, the constraints related to frequency regulation in the active / frequency and reactive / voltage local droop control parameter optimization model are reconstructed, and consistency constraints are added, so that the active / frequency and reactive / voltage local droop control parameter optimization model is decomposed into multiple sub-models, and the coordinated optimization of multiple sub-models is performed based on the alternating direction multiplier method, so as to achieve the optimal design of the active / frequency and reactive / voltage droop control parameters of each distributed photovoltaic.

[0157] Reconstruct the frequency regulation-related constraints in the active / frequency and reactive / voltage local droop control parameter optimization model (i.e., the equality constraints corresponding to the active / frequency droop control gains: ), specifically expressed as:

[0158] ,

[0159] ,

[0160] Where: subscript Indicates the moment, and Respectively represent the subnet Send to Subnet Downward and upward frequency modulation reference instructions; and They represent the subnet Send to Subnet Downward and upward frequency modulation reference instructions; represents the linearization constant coefficient corresponding to subnet m, which can be calculated separately for each subnet; and Represents distributed photovoltaic in the cluster Downward and upward active power / frequency droop control gains; and Represents subnets respectively The set of upstream and downstream subnets; It is a 0-1 indicator variable, 1 means that subnet m is connected to the substation, and the subnet receives the frequency regulation command issued by the transmission network operator, otherwise the subnet Only receive FM commands sent by adjacent subnets; Indicates a subnet A collection of nodes where distributed photovoltaics are installed.

[0161] The consistency constraint is expressed as:

[0162]

[0163] ,

[0164] ,

[0165] Where: and Subnet Calculated flow through subnet and Active power and reactive power on the interconnection line; and Subnet Calculated flow through subnet and Active power and reactive power on the interconnection line; and Subnet and subnet The calculated voltages on the boundary nodes; and Subnet Calculate and send to subnet Downward and upward frequency modulation reference instructions; and Subnet Calculate and send to subnet Downward and upward frequency modulation reference instructions; and They represent the participation of distributed PV in the upward and downward frequency regulation in subgrid m respectively; and Represents subnets respectively The calculated participation of distributed photovoltaics in upward and downward frequency regulation.

[0166] The vector consisting of tie-line power, boundary node voltage, frequency regulation reference command, and distributed photovoltaic frequency regulation participation is defined as a coupling variable. The above consistency constraint can be rewritten into a vector form, specifically:

[0167]

[0168] Where: Representation and subnet The vector composed of the relevant tie-line power, boundary node voltage, frequency regulation reference command and distributed photovoltaic frequency regulation participation is called coupling variable; similarly, Representation and subnet A vector consisting of the relevant variables: tie line power, boundary node voltage, frequency regulation reference command, and distributed photovoltaic frequency regulation participation; and Subnet and subnet Calculated coupling variables and value.

[0169] The objective function of each sub-model is:

[0170] ,

[0171] Where: Represents the objective function of the sub-model corresponding to sub-network m; , and Subnet The non-negative weight factors of distributed photovoltaic active reserve, reactive power and network loss, and satisfy ; Represents the set of lines in subnet m; Specifically, by solving the subnet The corresponding sub-model is obtained; is the Lagrange multiplier vector; represents the penalty coefficient. In addition to the equality constraints corresponding to the active power / frequency droop control gains, the constraints in the active power / frequency and reactive power / voltage local droop control parameter optimization models are all included in the constraints of each submodel. Furthermore, each submodel constraint also includes the aforementioned consistency constraints.

[0172] All sub-models are coordinated and optimized based on the alternating direction multiplier method. The specific iterative process is as follows:

[0173] ,

[0174] Where: and Represent sub-model m and sub-model respectively The feasible decision space is composed of the constraints of each sub-model; and Representation submodel and submodels Optimization variables of Indicates the current moment; and Respectively represent Sub-iteration model and submodels The calculated coupling variable values, the superscript indicates the number of iterations; and Represents sub-models and In the The optimization result obtained by the iteration; and Respectively expressed in The sub-model in the iterative calculation process and submodels The Lagrange multiplier vector of ; Indicates in The sub-model in the iterative calculation process The penalty coefficient.

[0175] Definition After the iteration calculation, the sub-model The original residual and the dual residual They are:

[0176] ,

[0177] Where: Is with subnet Related coupled variables; Indicates that the subnet Calculate and send to subnet The value of the coupling variable; and are the Lagrange multipliers calculated for the current iteration step and the previous iteration step respectively.

[0178] Rule No. The judgment condition of whether the algorithm converges after iterations is expressed as:

[0179] ,

[0180] ,

[0181] ,

[0182] Where: and Respectively represent The residual vector composed of the original residuals and dual residuals of all sub-models after iterations; is the maximum allowable residual of the alternating direction multiplier method, which can be set to 1×10 -3 ; represents the infinite norm; Represents the number of sub-models, that is, the number of distributed photovoltaic clusters.

[0183] Introduce a penalty coefficient dynamic update strategy to reduce the number of iterations and cluster As an example, the corresponding penalty coefficient dynamic update strategy is specifically expressed as follows:

[0184] ,

[0185] Where: is the adjustment factor; is the balance factor; represents the two-norm.

[0186] At the global optimization layer, by combining the alternating direction multiplier method to coordinate the optimization of multiple sub-models, multiple distributed photovoltaic active / frequency droop control-related parameters (upward reserve, downward reserve, droop gain) and reactive / voltage droop control-related parameters (droop gain, voltage dead zone, voltage saturation threshold, reactive power reference value, etc.) can be obtained, and these parameters can be sent to each distributed photovoltaic.

[0187] (5) The voltage self-correction control strategy within the cluster is implemented at the local coordination layer to quickly coordinate and control the distributed photovoltaics within the cluster where the voltage exceeds the limit. By modifying the reactive reference value of the distributed photovoltaic reactive / voltage droop control within the cluster, the distributed photovoltaic output reactive power is changed, thereby alleviating the distribution network voltage exceeding the limit problem caused by rapid source-load fluctuations.

[0188] The correction amount of the reactive reference value for distributed photovoltaic reactive power / voltage droop control in a cluster with voltage exceeding the limit is calculated by the following formula:

[0189] ,

[0190] ,

[0191] ,

[0192] Where: Represents a vector composed of reactive power correction quantities of each distributed photovoltaic in the cluster; Indicates the node after reactive power / voltage droop control Steady-state voltage on represents the vector consisting of the voltage correction amount of each node, Represents a distributed photovoltaic cluster Number of internal nodes; Representation node Voltage correction amount; and Respectively represent the upper and lower safety limits of the distribution network voltage; Indicates the droop control gain of each distributed photovoltaic in the cluster The vector of components; The diagonal elements are The diagonal matrix of .

[0193] By sending the correction amount of the reactive reference value of reactive / voltage droop control to each distributed photovoltaic in the cluster, the distributed photovoltaic will change its reactive power output accordingly, thereby adjusting the voltage of each node in the distributed photovoltaic cluster to a safe range.

[0194] (6) At the local control layer, based on the calculation results of the global optimization layer, the active / frequency and reactive / voltage droop control parameters of each distributed photovoltaic are configured. When a frequency or voltage deviation is detected, the distributed photovoltaic will adjust its active and reactive output in real time according to the droop control, thereby providing immediate frequency and voltage support. Among them, the active / frequency and reactive / voltage droop control is specifically expressed as:

[0195]

[0196]

[0197] Where: and Represents distributed photovoltaic exist Real-time active and reactive adjustments corresponding to frequency and voltage deviations; express The measured value of system frequency at the moment; express Time Node Voltage measurement value.

[0198] On an IEEE 33-node distribution network, the control scheme of the present invention was compared with three other control schemes to illustrate the effectiveness of the proposed control scheme in frequency-voltage coordinated support. Schemes 1-3 are:

[0199] Option 1: No control;

[0200] Solution 2: Update the distributed photovoltaic active and reactive power set points every 5 seconds;

[0201] Scheme 3: adopts the global optimization layer and local control layer in the proposed control scheme but does not include the local coordination layer.

[0202] Table 1 compares the control performance of the method of the present invention and the other three schemes.

[0203] Table 1 Performance comparison of different control schemes (24 hours)

[0204]

[0205] In order to intuitively compare the frequency support effects of different methods, the aggregate support power deficit is defined here, which characterizes the difference between the actual power regulation at the outlet of the distribution network substation and the expected power regulation of the transmission network. In terms of frequency support effect, the smaller the aggregate support power deficit, the better. As can be seen from the table, Schemes 1, 2 and 3 all have voltage limits that last for a long time, and the method of the present invention can fully guarantee the voltage safety of the distribution network. Scheme 3 and the method of the present invention are optimized with a cycle of 1h, while Scheme 2 is optimized every 5s. Due to the large changes in the system operating state, the reactive compensation amount and network loss of Scheme 3 and the method of the present invention are higher than those of Scheme 2. Scheme 3 and the method of the present invention have similar effects in frequency support, and the aggregate support power deficit is much smaller than that of Scheme 2. In summary, in terms of distribution network voltage regulation and system frequency support effects, the method of the present invention performs relatively well under various operating conditions, and the control effect is better than that of traditional methods.

[0206] Figure 2 The dynamic performance of different control methods is demonstrated. The frequency ( Figure 2 (a)) and voltage control results ( Figure 2 Compared with (b) in [1], the proposed method exhibits superior dynamic response characteristics. Building on Scheme 3, the proposed method further applies intra-cluster voltage self-correction control, significantly improving voltage over-limit performance. The frequency support is superior to Schemes 1 and 2, with a maximum frequency deviation of 0.1860 Hz. In summary, the proposed method offers a more balanced performance in terms of frequency and voltage support, and outperforms other schemes overall.

[0207] The above description of the specific implementation methods of the present invention in conjunction with the accompanying drawings is not intended to limit the scope of protection of the present invention. All equivalent models or equivalent algorithm processes made using the contents of the present invention specification and accompanying drawings, which are directly or indirectly applied to other related technical fields, are within the scope of patent protection of the present invention.

Claims

1. A distributed photovoltaic layered cluster frequency-voltage coordinated support control method, characterized in that: The following steps are involved: 1) Considering the impact of distributed photovoltaic participation in system frequency regulation on distribution network voltage, the electrical distance index is improved, and distributed photovoltaic clusters are divided based on non-negative matrix factorization technology; 2) Construct a hierarchical cluster coordinated control scheme consisting of a global optimization layer, a local coordination layer, and a local control layer to achieve multi-timescale coordinated control of distributed photovoltaics to provide rapid frequency support and alleviate voltage over-limit issues in the distribution network; 3) At the global optimization layer, with the goal of minimizing the loss of distributed photovoltaic power generation benefits, reactive power usage, and network losses, and using the distribution network power flow equation, distributed photovoltaic operating characteristics, and local droop control function as constraints, an optimization model for the active power / frequency and reactive power / voltage local droop control parameters is constructed; 4) Based on the clustering results, frequency regulation-related constraints in the active / frequency and reactive / voltage local droop control parameter optimization model are reconstructed and consistency constraints are added, thereby decomposing the active / frequency and reactive / voltage local droop control parameter optimization model into multiple sub-models. All sub-models are then coordinated and optimized based on the alternating direction multiplier method to achieve optimal design of the active / frequency and reactive / voltage droop control parameters of each distributed photovoltaic system. 5) At the local coordination layer, for distributed photovoltaic clusters with voltage exceeding the limit, the reactive power reference value for distributed photovoltaic reactive power / voltage droop control within the cluster is modified to achieve voltage self-correction control within the distributed photovoltaic cluster with voltage exceeding the limit; 6) At the local control layer, based on the parameter optimization design results obtained at the global optimization layer, the active power / frequency and reactive power / voltage droop control parameters of each distributed photovoltaic are configured. Each distributed photovoltaic system combines local real-time frequency and voltage measurement values to quickly respond to frequency and voltage deviations, achieving immediate frequency and voltage coordinated support.

2. A distributed photovoltaic layered cluster frequency-voltage coordinated support control method according to claim 1, characterized in that: In step 1), the influence of distributed photovoltaic participation in system frequency regulation on the distribution network voltage is considered, the electrical distance index is improved, and the distributed photovoltaic is clustered based on the non-negative matrix decomposition technology. The specific method is: ， ， ， ， ， ， ， ， Where: Represents the node-edge adjacency matrix of the network, whose elements are , representing a node and nodes Electrical connections between represents the number of distributed photovoltaic clusters, Indicates the number of nodes in the power distribution network; Represents the set of nodes in the distribution network except the reference node; and are two non-negative fundamental matrices; Node-cluster membership matrix, matrix The row number corresponding to the maximum value of each column element in is the number of the cluster to which each node belongs; Represents the node similarity matrix, whose elements are , representing a node and nodes similarities between represents the Frobenius norm; is a positive real parameter used to balance the contribution of topological and attribute information; The third term of the objective function is a regularization term, which is used to control the number of nodes in each cluster. is the regularization constant; Representation matrix No. elements, represents the total membership of each cluster, is a constant vector whose elements are all 1; Represents the distribution network node attribute matrix; and They are the distribution network node attribute matrices No. and column vectors; is a constant coefficient vector whose elements For quantization nodes The impact of injection power adjustment on substation outlet power; represents the active-voltage sensitivity matrix, For its Rank Column elements; Indicates reactive-voltage sensitivity; It means taking the diagonal elements of the matrix and forming a column vector; Representation node and The electrical distance indicator between It indicates the improved electrical distance index; and Respectively represent nodes The distributed photovoltaic injection unit power to the node and nodes The impact of voltage, and Respectively represent nodes The distributed photovoltaic injection unit power to the node and nodes The magnitude of the voltage's influence; Representation node The installed capacity of distributed photovoltaics; and Respectively No. and diagonal elements; express No. Rank Column element.

3. A distributed photovoltaic layered cluster frequency-voltage coordinated support control method according to claim 1, characterized in that: In step 2), the specific method for achieving multi-time-scale coordinated control of distributed photovoltaics based on the hierarchical cluster coordinated control scheme is as follows: The hierarchical cluster coordinated control scheme includes three control levels: global optimization layer, local coordination layer and local control layer. The global optimization layer performs distributed optimization once an hour to optimize the active power / frequency and reactive power / voltage droop control parameters of distributed photovoltaics. The local coordination layer performs voltage self-correction control within the cluster on a minute-level time scale. By modifying the reactive reference value of the reactive power / voltage droop control of distributed photovoltaics within the cluster, the reactive power output of distributed photovoltaics within the cluster is adjusted, thereby alleviating the voltage over-limit of the distribution network. In the local control layer, the active power / frequency and reactive power / voltage droop control parameters of each distributed photovoltaic are configured in combination with the optimization design results of the global optimization layer. Distributed photovoltaics respond to system frequency and voltage deviations in real time based on local frequency and voltage measurements. Through the mutual cooperation of the three control levels, distributed photovoltaic coordinated control at multiple time scales is achieved.

4. A distributed photovoltaic layered cluster frequency-voltage coordinated support control method according to claim 1, characterized in that: In step 3), in the active / frequency and reactive / voltage local droop control parameter optimization model, the objective function is specifically expressed as: ， Where: The objective function of the active power / frequency and reactive power / voltage local droop control parameter optimization model; , and are all non-negative weight factors for distributed photovoltaic active backup, reactive power and network loss, and satisfy ; Subscript Indicates the moment; and Respectively Time Node Upward and downward reserve power; for Time Node The reactive power adjustment of the inverter; Indicates line The square of the upper current amplitude; Indicates line resistance; Represents a collection of distribution network lines; Represents the set of nodes with distributed photovoltaic installed in the entire distribution network; To optimize the time domain; The distribution network power flow equation constraints are as follows: ， ， ， ， ， ， Where: and Represents each line separately There are active and reactive currents on the upper side; Representation and Node The set of downstream nodes connected by lines; and Represents nodes respectively Active and reactive power injected on the Representation node The square of the voltage amplitude, For nodes Voltage amplitude; Indicates line reactance; and Respectively represent the upper and lower safety limits of the distribution network voltage; Represents the set of nodes in the distribution network except the reference node.

5. A distributed photovoltaic layered cluster frequency-voltage coordinated support control method according to claim 4, characterized in that: In step 4), based on the cluster division results, the constraints related to frequency regulation in the active / frequency and reactive / voltage local droop control parameter optimization model are reconstructed, and consistency constraints are added. The specific method is: If two clusters and There is a public line between , then in A virtual node is introduced in the middle and the line is disconnected at the virtual node; the entire network is divided into subnets and , each subnet has a contact line and a boundary node; in distributed optimization, the subnet and Each has an optimization model, called a sub-model; The constraints related to frequency modulation obtained after reconstruction are specifically expressed as: ， ， Where: and Respectively The downward and upward frequency regulation reference instructions issued by the transmission network operator at all times; and Respectively Time from subnet Send to Subnet Downward and upward frequency modulation reference instructions; and They respectively represent Time from subnet Send to Subnet Downward and upward frequency modulation reference instructions; Indicates that it corresponds to a subnet The linearization constant coefficient is calculated separately for each subnet; and Represents distributed photovoltaic in the cluster Downward and upward active power / frequency droop control gains; and Represents subnets respectively The set of upstream and downstream subnets; It is a 0-1 indicator variable, 1 indicates the subnet Connected to the substation, receiving the frequency regulation instructions issued by the transmission network operator, otherwise the subnet Only receive FM commands sent by adjacent subnets; Indicates a subnet The collection of nodes with distributed photovoltaic installed; The added consistency constraints specifically include: ， ， ， Where: and Respectively Time by subnet Calculated flow through subnet and Active power and reactive power on the interconnection line; and Respectively Time by subnet Calculated flow through subnet and Active power and reactive power on the interconnection line; and They are Time by subnet and subnet The calculated voltages on the boundary nodes; and Respectively Time by subnet Calculate and send to subnet Downward and upward frequency modulation reference instructions; and Respectively Time by subnet Calculate and send to subnet Downward and upward frequency modulation reference instructions; and They represent the distributed photovoltaic upward and downward frequency regulation participation calculated by subgrid m respectively; and Subnet The calculated participation of distributed photovoltaic upward and downward frequency regulation; The above consistency constraint is further expressed as the following vector form: ， Where: , is related to the subnet Related coupling variables; similarly, , is related to the subnet Related coupled variables; and Subnet and subnet Calculated coupling variables and value.

6. A distributed photovoltaic layered cluster frequency-voltage coordinated support control method according to claim 5, characterized in that: In step 4), the objective function of each sub-model is: ， Where: Indicates a subnet The objective function of the corresponding sub-model; , and All subnets The non-negative weight factors of distributed photovoltaic active reserve, reactive power and network loss, and satisfy ; Represents the set of lines in subnet m; is the Lagrange multiplier vector; Represents the penalty coefficient.

7. A distributed photovoltaic layered cluster frequency-voltage coordinated support control method according to claim 6, characterized in that: The coordinated optimization of all sub-models based on the alternating direction multiplier method is carried out, and the specific iterative process is as follows: ， Where: and Represent sub-model m and sub-model respectively feasible decision space; and Representation submodel and submodels Optimization variables of Indicates the current moment; and Respectively represent Sub-iteration model and submodels The calculated coupling variable values; and Represents sub-models and In the The optimization result obtained by the iteration; and Respectively expressed in The sub-model in the iterative calculation process and submodels The Lagrange multiplier vector of ; Indicates in The sub-model in the iterative calculation process The penalty coefficient; Definition After the iteration calculation, the sub-model The original residual and the dual residual They are: ， Where: Is with subnet Related coupled variables; Indicates that the subnet Calculate and send to subnet The value of the coupling variable; and are the Lagrange multipliers calculated for the current iteration step and the previous iteration step respectively; Rule No. The judgment condition of whether the algorithm converges after iterations is expressed as: ， ， ， Where: and Respectively represent The residual vector composed of the original residuals and dual residuals of all sub-models after iterations; is the maximum allowable residual of the alternating direction multiplier method; represents the infinite norm; represents the number of sub-models, that is, the number of distributed photovoltaic clusters; The penalty coefficient dynamic update strategy is introduced, which is specifically expressed as follows: ， Where: is the adjustment factor; is the balance factor; represents the two-norm.

8. A distributed photovoltaic layered cluster frequency-voltage coordinated support control method according to claim 1, characterized in that: In step 5), in the local coordination layer, the method for modifying the reactive reference value of the distributed photovoltaic reactive power / voltage droop control within the cluster to realize voltage self-correction control is specifically as follows: If the distributed photovoltaic cluster If an internal node voltage exceeds the limit, the voltage correction amount is calculated using the following formula: ， ， Where: Indicates the voltage correction amount of each node The vector formed; and Respectively represent the upper and lower safety limits of the distribution network voltage; Indicates the node after reactive power / voltage droop control Steady-state voltage on Represents a distributed photovoltaic cluster Number of internal nodes; The correction amount of the reactive power reference value for each distributed photovoltaic reactive power / voltage droop control is calculated according to the following formula: ， Where: Represents a vector composed of reactive power correction quantities of each distributed photovoltaic in the cluster; Indicates reactive-voltage sensitivity; Indicates the droop control gain of each distributed photovoltaic in the cluster The vector of components; The diagonal elements are The diagonal matrix of ; By calculating the correction amount of the reactive reference value of distributed photovoltaic reactive power / voltage droop control It is then sent to each distributed photovoltaic unit in the cluster to modify the reactive power reference value of each distributed photovoltaic unit's reactive power / voltage droop control.

9. A computer-readable storage medium having computer instructions stored thereon, characterized in that: The computer instructions are used to enable a computer to execute the steps of the method according to any one of claims 1 to 8.

10. An electronic device, characterized in that: include: one or more processors; a memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 8.

Images