A distributed photovoltaic multi-cluster voltage control method and system and storage medium

Abstract

The application relates to a distributed photovoltaic multi-cluster voltage control method and system and a storage medium, which comprises the following steps: preliminarily determining safe clusters and dangerous clusters by using cluster division results and cluster voltage deviation degrees of each cluster, and locally regulating the voltage of the dangerous clusters; if the local voltage regulation fails, calculating reactive power-voltage sensitivity factors between each cluster, and re-determining safe clusters and dangerous clusters according to the reactive power-voltage sensitivity factors; selecting a safe cluster with the maximum sensitivity factor between a leading node in each cluster and a leading node in the dangerous cluster, making the total reactive power amount of inverters in the safe cluster increase by △Q, and distributing reactive power to each distributed photovoltaic power supply in the cluster based on the total reactive power amount; cyclically detecting the voltage deviation degrees of each cluster, re-determining safe clusters if the clusters are still in the dangerous clusters, and vice versa, so that the voltage of each cluster is within a safe margin, and the control of grid-level voltage stability is realized. The application can be widely applied in the field of voltage control of new energy power system distribution networks.

Description

Technical Field

[0001] This invention relates to the field of voltage control in distribution networks of new energy power systems, and in particular to a distributed photovoltaic multi-cluster voltage control method, system, and storage medium based on cluster voltage deviation. Background Technology

[0002] Clean energy, represented by photovoltaics (PV), has developed rapidly worldwide due to its economic, clean, and environmentally friendly advantages, significantly alleviating the pressure of fossil fuel depletion and ecological degradation. However, with the increasing penetration rate of distributed photovoltaic (PV) power in distribution networks, problems such as voltage and frequency fluctuations and exceeding limits caused by the intermittent, random, and large power fluctuations of PV power output have become prominent, posing significant challenges to distribution network operation. To achieve rapid and real-time voltage stability control of the distribution network under the new power system and lay a theoretical foundation for the safe and stable operation of the power system under the new energy structure, it is necessary to study the stability control after distributed PV integration. PV clusters have low inertia, low damping, and rapid response, making cluster voltage stability control a challenge. Inappropriate control modes and parameter settings of distributed PV inverters lead to low PV utilization, poor economic efficiency, and large grid voltage fluctuations. Therefore, it is urgent to conduct research on dynamic reactive power-voltage switching control technology for distributed PV clusters under the new circumstances, as well as frequency and voltage active support technology for coordination between distributed PV clusters, to provide decision-making basis for projects with high proportions of new energy integration.

[0003] In recent years, a number of publications have focused on stabilization methods after distributed photovoltaic (PV) grids are connected to distribution networks. Their main focus includes: inverter control modes, coordination with energy storage, changing transformer taps, and corresponding improvement measures.

[0004] Currently, voltage regulation methods for distributed photovoltaic (PV) systems mainly include reactive power compensation and active power reduction. Regarding local reactive power compensation, the German Association for Electrical, Electronic & Information Technologies (DAAD) has proposed four reactive power control strategies suitable for distributed PV: constant reactive power Q control, constant power factor cosφ control, cosφ(P) control based on PV active power output, and Q(U) control based on grid connection point voltage amplitude. However, as a fundamental method, each has its advantages and disadvantages. The constant reactive power method cannot participate in grid voltage regulation in a timely manner; the power factor method results in significant losses due to excess power transmission when PV power generation is high and local load is high; and the control mode based on grid connection point voltage has weak voltage regulation capability at the grid connection point. Active power reduction methods are used when the inverter output capacity reaches its rated capacity but the voltage still exceeds the limit. This study investigates the reactive power regulation capability of photovoltaic inverters and establishes a multi-mode voltage control model for low-voltage distribution networks based on the reactive power and voltage sensitivity matrix. When the network faces the risk of voltage exceeding limits, the reactive power of the inverter is adjusted with the goal of risk suppression. When the network is safe, the optimization of network loss and power factor is used as the basis for reactive power regulation of the inverter. However, the data volume of the entire distribution network is large, the efficiency is low, and the real-time performance is poor, resulting in poor voltage fluctuation control. Summary of the Invention

[0005] To address the shortcomings in voltage stability after distributed photovoltaic (PV) grid integration, the present invention aims to provide a distributed PV multi-cluster voltage control method, system, and storage medium that can effectively adapt to and regulate grid operating voltage.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a distributed photovoltaic multi-cluster voltage control method, comprising: initially determining safe clusters and dangerous clusters using cluster division results and cluster voltage deviation of each cluster, and performing local voltage regulation on the dangerous clusters; if the local voltage regulation fails, calculating the reactive power-voltage sensitivity factor between each cluster, and re-determining safe clusters and dangerous clusters based on the reactive power-voltage sensitivity factor; selecting the safe cluster with the largest sensitivity factor between the dominant node in each cluster and the dominant node of the dangerous cluster, increasing the total reactive power ΔQ generated by the inverter in the safe cluster, and allocating reactive power to each distributed photovoltaic power source in the cluster based on the total reactive power; cyclically detecting the voltage deviation of each cluster, and if it is still in a dangerous cluster, re-determining a safe cluster, otherwise ensuring that the voltage of each cluster is within the safety margin, thereby achieving grid-level voltage stability control.

[0007] Furthermore, the preliminary determination of safe and dangerous clusters using the cluster division results and the cluster voltage deviation of each cluster includes: using the cluster division results and the cluster voltage deviation of each cluster to determine whether the voltage in each cluster exceeds the limit; if it exceeds the limit, it is a dangerous cluster, otherwise it is a safe cluster.

[0008] Furthermore, the local voltage regulation of the hazardous cluster includes: scheduling inverters with adjustable reactive power capacity within the hazardous cluster and switching the inverters to an improved droop control mode for local voltage regulation.

[0009] Furthermore, the determination of the cluster voltage deviation includes: obtaining each photovoltaic cluster microgrid based on the cluster division results; calculating the reactive voltage sensitivity of each node to the dominant node for each node within the cluster; and calculating the voltage deviation of each cluster based on the number of nodes within the cluster and the reactive voltage sensitivity.

[0010] Furthermore, the voltage deviation is:

[0011]

[0012] In the formula, M θ U is the cluster voltage deviation; N is the number of nodes in the cluster; U i For the real-time operating voltage of the i-node, S iθ U is the reactive voltage sensitivity coefficient of node i with respect to the dominant node θ. min U represents the minimum terminal voltage of each node within the cluster. max This represents the maximum voltage at each node within the cluster.

[0013] Furthermore, the total reactive power is:

[0014]

[0015] In the formula, ΔQ j Increase the total reactive power generated by the photovoltaic inverters within the safety cluster; S ij U is the reactive power-voltage sensitivity factor between the dominant node j of the safe cluster and the dominant node i of the dangerous cluster. min U represents the minimum terminal voltage of each node within the cluster. max U represents the maximum voltage at each node within the cluster. i Provides the real-time operating voltage for the i-node.

[0016] Furthermore, the allocation of reactive power to each distributed photovoltaic power source within the cluster based on the total reactive power includes:

[0017] Establish active / reactive voltage sensitivity matrices between clusters and between nodes;

[0018] The relationship between the change in node voltage amplitude and the change in power is obtained from the active / reactive voltage sensitivity matrix of the node, and the difference between the real-time voltage and the rated voltage of the node is obtained.

[0019] The reactive power change of each node is calculated based on the difference between the real-time voltage and the rated value of the node. The reactive power generated by each distributed photovoltaic power source in the cluster is obtained from the reactive power change K of each node and the reactive power-voltage sensitivity factor, so as to distribute the total reactive power to each distributed photovoltaic power source in the cluster.

[0020] A distributed photovoltaic multi-cluster voltage control system includes: a primary partitioning module, which initially determines safe and dangerous clusters based on cluster partitioning results and cluster voltage deviation of each cluster, and performs local voltage regulation on the dangerous clusters; a cluster determination module, which calculates the reactive power-voltage sensitivity factor between each cluster if local voltage regulation fails, and redetermines safe and dangerous clusters based on the reactive power-voltage sensitivity factor; a power distribution module, which selects the safe cluster with the largest sensitivity factor between the dominant node in each cluster and the dominant node of the dangerous cluster, increases the total reactive power ΔQ generated by the inverters in the safe cluster, and distributes reactive power to each distributed photovoltaic power source in the cluster based on the total reactive power; and a detection module, which cyclically detects the voltage deviation of each cluster, and if it is still in a dangerous cluster, redetermines a safe cluster; otherwise, the voltage of each cluster is within the safety margin, achieving grid-level voltage stability control.

[0021] A computer-readable storage medium storing one or more programs, said one or more programs including instructions that, when executed by a computing device, cause the computing device to perform any of the methods described above.

[0022] A computing device includes: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include instructions for performing any of the methods described above.

[0023] The present invention has the following advantages due to the adoption of the above technical solutions:

[0024] This invention is based on dual closed-loop control of voltage and current, employing improved droop control and PQ control mode switching for grid-connected inverters to establish a distributed photovoltaic (PV) power grid connection model. Then, using the cluster partitioning results and voltage deviation level indicators of each cluster, the risk of voltage exceeding limits in each cluster is assessed, establishing reactive power and voltage sensitivity matrices between clusters and between nodes. Finally, a voltage adaptive control mode is adopted for the distributed PV inverters within each cluster to achieve grid voltage stability, effectively adapting to and adjusting grid operating voltage. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of a distributed photovoltaic multi-cluster voltage control method according to an embodiment of the present invention;

[0026] Figure 2 This is an equivalent circuit diagram of a photovoltaic power source according to an embodiment of the present invention;

[0027] Figure 3 This is a grid connection diagram of a photovoltaic power source according to an embodiment of the present invention;

[0028] Figure 4 This is a control block diagram of a droop control inverter according to an embodiment of the present invention;

[0029] Figure 5 This is a simulation system for distributed photovoltaic access to IEEE 33 nodes in one embodiment of the present invention;

[0030] Figure 6 This is a schematic diagram of the voltage level of the IEEE 33-node system in one embodiment of the present invention;

[0031] Figure 7a This is a schematic diagram of the active power output of distributed photovoltaic power generation under PQ mode in one embodiment of the present invention;

[0032] Figure 7b This is a schematic diagram of the reactive power output of distributed photovoltaic systems under PQ mode in one embodiment of the present invention;

[0033] Figure 7c This is a voltage variation diagram of distributed photovoltaic system under Q(U) mode in one embodiment of the present invention;

[0034] Figure 7d This is a diagram showing the reactive power variation of distributed photovoltaic systems under the Q(U) mode in one embodiment of the present invention;

[0035] Figure 8 This is a diagram showing the reactive voltage sensitivity relationship of an improved IEEE 33-node system in one embodiment of the present invention.

[0036] Figure 9 This is a diagram showing the voltage changes of IEEE 33 nodes before and after the control mode based on cluster voltage deviation in one embodiment of the present invention. Detailed Implementation

[0037] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention are within the scope of protection of the present invention.

[0038] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this application. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0039] This invention proposes a distributed photovoltaic (PV) multi-cluster voltage control method, system, and storage medium, comprising: dividing the actual distribution network containing distributed PV into clusters based on improved electrical distances, and setting the grid-connected inverter control mode to automatically switch according to cluster voltage deviation. The voltage exceedance level of each cluster is assessed based on its cluster voltage deviation, and a dominant node is selected based on a reactive power-voltage sensitivity matrix established between clusters and nodes, and the sensitivity relationship between the dominant nodes is calculated. Finally, a control mode based on cluster voltage deviation is used to coordinate the safe and dangerous clusters to achieve grid voltage stability. This method avoids the increased workload and reduced real-time effectiveness of voltage control caused by processing large amounts of data from various nodes in the distribution network, and has stronger targeting and adaptability, enabling the grid voltage to stabilize more quickly within a safe range.

[0040] This invention targets distribution networks with a high proportion of distributed photovoltaic (PV) power. It uses impedance-based electrical distance as the clustering criterion to group the PV-enabled distribution network, calculates the voltage deviation of each cluster, and employs improved droop control based on virtual impedance for clusters with severe over-limit conditions. Furthermore, it adjusts the reactive power within each safely operating cluster to regulate the voltage of each cluster within the grid. Based on this, it monitors the grid's operating status in real time, achieving dynamic cluster division and corresponding control mode switching. This method simplifies distribution network voltage control, avoiding the need to control each node, which would significantly increase the complexity of data processing and make it difficult to operate. Simultaneously, by grouping the distribution network with a high proportion of PV power, multi-level voltage control is achieved, enabling the control method adopted in this invention to track grid operating characteristics in real time and adopt corresponding control modes to maintain grid voltage stability.

[0041] In one embodiment of the present invention, a distributed photovoltaic multi-cluster voltage control method is provided. In this embodiment, as shown... Figure 1 As shown, the method includes the following steps:

[0042] 1) Based on the cluster division results and the cluster voltage deviation of each cluster, the safe clusters and dangerous clusters are initially determined, and the dangerous clusters are regulated locally.

[0043] 2) If local voltage regulation fails, calculate the reactive power-voltage sensitivity factor between each cluster, and redetermine the safe and dangerous clusters based on the reactive power-voltage sensitivity factor.

[0044] 3) Select the safe cluster with the largest sensitivity factor between the dominant node in each cluster and the dominant node in the dangerous cluster, so that the inverters in the safe cluster generate an additional total reactive power ΔQ, and distribute reactive power to each distributed photovoltaic power source in the cluster based on this total reactive power.

[0045] 4) Circularly detect the voltage deviation of each cluster. If it is still in a dangerous cluster, then redetermine the safe cluster. Otherwise, the voltage of each cluster is within the safety margin, thus achieving grid-level voltage stability control.

[0046] In this embodiment, a distributed photovoltaic (PV) power generation grid connection model needs to be established. First, a PV power generation side model is established, such as... Figure 2 As shown, the output current I of the photovoltaic power source is:

[0047]

[0048] Among them I SC I is the short-circuit current of the photovoltaic cell. d I is the diode saturation current. Sh I0 is the leakage current of the photovoltaic cell, I0 is the reverse saturation current, and R0 is the reverse saturation current. S For the series equivalent resistance, R sh U is the parallel equivalent resistance, and U is the output voltage of the photovoltaic power supply.

[0049] Considering engineering practicality, a practical mathematical model of the photovoltaic power source is established. The output current I of the photovoltaic power source is then:

[0050]

[0051]

[0052]

[0053] Where U is the output voltage of the photovoltaic power supply, C1 is the current proportional coefficient, and C2 is the current exponential coefficient; U oc U is the open-circuit voltage. m For the maximum power voltage, I m This represents the maximum power current. The active power P output (absorbed by the load) of the PV power supply is:

[0054]

[0055] Where R1 is the load resistance, X1 is the load impedance, R0 is the power supply internal resistance, and X0 is the power supply internal reactance.

[0056] When X0 + X1 = 0, the power can have a maximum value. Substituting this into equation (1-5) and taking the derivative, we get:

[0057]

[0058] Therefore, the condition for the load to obtain maximum power is R1 = R0 and X1 = -X0. At this time, the load Z... L The maximum active power consumed is It should be noted that the same amount of power is consumed at Z0, the maximum power transfer efficiency is 50%, and Z0 is the power supply impedance. Maximum power point tracking control can be achieved using methods such as interference observation.

[0059] Then, a photovoltaic grid-connected model is established, such as... Figure 3 As shown, the photovoltaic grid-connected inverter passes through a filter inductor and capacitor after being connected to the power grid. The filter inductor L m The voltage equation is:

[0060]

[0061]

[0062] Where m is a controllable sinusoidal modulation signal, I pv V is the inverter output current. L Let V be the load voltage vector, k be the voltage modulation coefficient, and V be the voltage vector. pv ω is the inverter output voltage, ω is the angular frequency of the three-phase electrical quantity, and t is the operating time. Let be the initial phase angle, and i be the phase coefficient. Ignore the filter resistor R connected to the inverter. m (Since its value is very small), the current equation for the filter capacitor can be written as:

[0063]

[0064] Among them, C m I represents the size of the filter capacitor. L For the magnitude of the current flowing to the load, I Pcc This refers to the magnitude of the current flowing to the grid connection point.

[0065] The outer loop control is a voltage loop, which, combined with PI control, stabilizes the load voltage to the given voltage. The inner loop control is a current loop, which improves the system's dynamic response capability.

[0066] When the photovoltaic power grid-connected model is three-phase symmetrical, the filter output load voltage vector (load-side voltage) V L With inverter output voltage V pv The transfer function is shown in equation (1-10).

[0067]

[0068] In the formula,

[0069] Considering distributed photovoltaic (PV) grid integration, when the PV system employs unity power factor control and its active power output is significantly greater than the local load, and reactive power losses are negligible, let the grid voltage be V. PCC Then we have:

[0070]

[0071] In the formula, P pv X represents the active power output of the inverter, and X represents the equivalent reactance of the system.

[0072] If we want to maintain a constant grid connection voltage before and after photovoltaic installation, ignoring the voltage difference increment on the vertical axis, the required reactive power Q is... PV Size:

[0073]

[0074] In the formula, Q L P represents the reactive power of the load. L This indicates the active power of the load.

[0075] In step 1) above, the safe and dangerous clusters are initially determined by using the cluster division results and the cluster voltage deviation of each cluster. Specifically, the voltage in each cluster is judged to be exceeded by using the cluster division results and the cluster voltage deviation of each cluster. If it exceeds the limit, it is a dangerous cluster, otherwise it is a safe cluster.

[0076] Among them, local voltage regulation is performed on dangerous clusters: the inverters with adjustable reactive power capacity in the dangerous clusters are dispatched and switched to the improved droop control mode for local voltage regulation.

[0077] In this embodiment, the basic control principle of Q(U) in inverter control is as follows: Figure 4 As shown, when a voltage source inverter topology is used for grid-connected photovoltaic systems, the active power P and reactive power Q of the inverter output are calculated based on the abc three-phase stationary coordinate system as follows:

[0078]

[0079]

[0080] Where, θ i Z is the impedance power factor angle between the inverter and the grid connection point, α is the phase angle difference between the inverter output and the grid connection point voltage, and Z is the voltage difference between the inverter output and the grid connection point. m U is the line impedance. pv U is the inverter terminal voltage. pcc This is the voltage at the grid connection point.

[0081] Taking two distributed photovoltaic grid-connected inverters connected in parallel as an example, when the line impedance Z m When primarily expressing emotions:

[0082]

[0083]

[0084] The phase angle is obtained only after integrating the frequency, that is:

[0085]

[0086] Among them, f σ For the rated frequency, α σ X is the rated power angle. m Let t0 be the load reactance, and t0 be the end time of the integral calculation.

[0087] The corresponding droop control expression is:

[0088]

[0089]

[0090] Among them, P m f is the measured active power output of the photovoltaic system, and f is the measured frequency of the photovoltaic system. σ K is the rated frequency for photovoltaics. P K is the active power droop coefficient. q U is the reactive power droop factor. σ This is the rated output voltage for photovoltaics.

[0091] By performing a Parker transformation on the real-time measured three-phase voltage and current at the distributed photovoltaic grid-connected inverter terminals, the dq-axis components of the voltage and current can be obtained, enabling decoupled power calculation. The Parker transformation matrix T abc / dq for:

[0092]

[0093] Where ω is the angular frequency of the three-phase electrical quantity. The dq-axis components of the three-phase voltage and current after Parker transformation are u... md u mq i md i mq Calculate the inverter output power at this time:

[0094] p = u md i md +u mq i mq (1-21)

[0095] q = u mq i md -u md i mq (1-22)

[0096] By adopting grid voltage-oriented vector control, the output current of the photovoltaic grid-connected inverter and the d-axis of the synchronous rotating coordinate system are rotated synchronously with the grid voltage vector, and the d-axis of the synchronous rotating coordinate system is in the same direction as the grid voltage vector, thus achieving power decoupling. Based on this, the inverter PQ control and droop control strategies are implemented.

[0097] After measuring the instantaneous power output of the inverter, the high-frequency components change rapidly and generally have a small amplitude, which can cause unnecessary frequent operation of the controller and reduce its lifespan. Therefore, a low-pass filter is needed to remove the high-frequency components to enhance system stability.

[0098]

[0099]

[0100] Where ω σ This is the cutoff frequency of the low-pass filter.

[0101] For distributed photovoltaic clusters using droop control, the internal power should be allocated according to capacity to prevent inverter overload damage that could cause further voltage fluctuations or even exceed limits. When the system is in a stable operating state, the operating frequencies of all units within the cluster are the same, i.e., ω1 = ω2. Therefore, according to equations (1-15) and (1-18), it can be seen that as long as all inverters in the cluster have the same reference frequency under rated active power and the droop coefficient (K) is the same, the system can achieve the desired voltage distribution. 1P K 2P constant power That is, satisfying equations (1-25) and (1-26):

[0102] ω1 σ =ω2 σ (1-25)

[0103]

[0104] At this point, the inverter's output active power can be evenly distributed within the cluster according to its rated power.

[0105] K 1P P 1m =K 2P P 2m (1-27)

[0106] According to equations (1-16) and (1-19), under the premise that (1-28) and (1-29) are true, the premise for achieving reactive power distribution according to capacity, i.e., the premise that (1-30) is true, is that capacity E1 = E2.

[0107] U1 σ =U2 σ (1-28)

[0108]

[0109] At this point, the reactive power output by the inverter can be evenly distributed within the cluster according to its rated power.

[0110] K 1q Q 1m =K 2q Q 2m (1-30)

[0111] The slope of the reactive voltage droop control curve is generally small. Small disturbances in voltage can lead to a large reactive voltage difference, resulting in inverter overcurrent. When equations (1-28) and (1-29) hold true, the voltage difference between distributed photovoltaic systems within the cluster is:

[0112] ΔU=U2-U1=K 2q Q 2m -K 1q Q 1m (1-31)

[0113] Substituting equation (1-19) into equation (1-16) yields

[0114]

[0115] Substituting equation (1-32) into equation (1-31) yields

[0116]

[0117] From the final derivation (1-33), it can be seen that, under the premise that (1-28) and (1-29) are true, U2 = U1 can only be guaranteed if the reactive voltage droop coefficient of the distributed photovoltaic inverter in the cluster is inversely proportional to the impedance, thereby realizing the equal distribution of reactive power within the cluster according to capacity.

[0118] The improved droop control coefficient range determination method used in this embodiment is as follows: Power is allocated according to capacity by considering the virtual impedance of distributed photovoltaic access. The droop coefficient ratio within the distributed photovoltaic cluster is determined, and to select an appropriate value, the reactive power droop control coefficient is derived:

[0119] Suppose the cluster has N nodes, b branches, and E vir E σ P, Q, P L Q L U and U represent the virtual grid-connected voltage, rated reference voltage, active power output of distributed photovoltaic / energy storage at each node, reactive power output of distributed photovoltaic at each node, active load, reactive load, and voltage matrix of each node, respectively; E0 is the rated voltage matrix; R m Xm K q These represent an n×n diagonal matrix of virtual resistance, virtual reactance, and reactive droop coefficient, respectively. b P b Q b These represent the line voltage drop, active power transmission, and reactive power transmission matrices, respectively. R b X b Let M be the line resistance and reactance matrix, and M be the branch correlation matrix.

[0120] Where Q' represents the reactive power output of each distributed PV unit when reactive power is allocated according to capacity within the cluster. The virtual output voltage of each distributed PV power source is:

[0121] E vir =E σ -K q Q (1-34)

[0122] The voltage at each node within the cluster is as follows:

[0123]

[0124] The voltage drop on the branch is:

[0125]

[0126] According to Kirchhoff's current law, we have:

[0127] U b =M T U (1-37)

[0128] The power flow equilibrium equations at each node are:

[0129] P b =M T (PP L (1-38)

[0130] Q b =M T (QQ L (1-39)

[0131] From equations (1-34) to (1-39), we can obtain:

[0132] [M(E0K q +X m )+X b M T Q = E0M T E σ -(M T R m +R b M T )P+Rb M T P L +X b M T Q L (1-40)

[0133] Based on the preceding analysis, active power can be allocated according to capacity, therefore:

[0134] P = KE 1×n P L (1-41)

[0135]

[0136] Since the active and reactive power loads of distributed generation are generally proportional, then:

[0137] Q'=KE 1×n Q L (1-43)

[0138] The reactive power matrix equation for actual power grid operation is as follows:

[0139] Q = Q' + ΔQ (1-44)

[0140] Q L Q = Q' + ΔQ L (1-45)

[0141] Furthermore, due to:

[0142]

[0143] Equation (1-40) can then be expressed as:

[0144] [M(E0K q +X m )+X b M T Q

[0145] =(M T X m +X b M T )Q'-[(M T R m +R b M T )KE 1×n -R b M T ]P L +X b M T Q L (1-47)

[0146] By substituting equations (1-43) to (1-47) into the system parameters of the microgrid cluster, the range of droop coefficient selection can be determined.

[0147] In step 1) above, determining the cluster voltage deviation includes the following steps:

[0148] 1.1) Based on the cluster partitioning results, the microgrids of each photovoltaic cluster are obtained;

[0149] 1.2) For each node in the cluster, calculate the reactive voltage sensitivity of each node to the dominant node;

[0150] 1.3) The voltage deviation of each cluster is calculated based on the number of nodes in the cluster and the reactive voltage sensitivity.

[0151] The voltage deviation is:

[0152]

[0153] In the formula, M θ U is the cluster voltage deviation; N is the number of nodes in the cluster; U i For the real-time operating voltage of the i-node, S iθ U is the reactive voltage sensitivity coefficient of node i with respect to the dominant node θ. min U represents the minimum terminal voltage of each node within the cluster. max This represents the maximum voltage at each node within the cluster.

[0154] In this embodiment, the reactive voltage sensitivity relationship between clusters and between nodes is calculated. The voltage of all nodes in the cluster is input into the cluster control system to obtain the voltage deviation of each cluster, thereby obtaining the operation control mode of each cluster and the system parameters under the corresponding mode.

[0155] In step 1.1) above, the cluster partitioning method includes the following steps:

[0156] Determining Modularity Metrics: The strength of a cluster structure is typically explained by its external characteristics, such as the degree of internal cohesion, the degree of cohesion between clusters, the number of clusters, the size of the clusters, and the rationality of the cluster logic. Modularity metrics can quantitatively describe the external characteristics of a community. Modularity metrics quantify the structural strength of a community and determine the optimal number of partitions between each partition, defined as follows:

[0157]

[0158] In the formula A ij The weight of the edge between node i and node j is represented by ∑ j A ijδ(i,j) is the sum of the weights of all edges connected to node i. If node i and node j are in the same cluster, then δ(i,j) is 1; otherwise, it is 0. A modularity close to 1 reflects the tightness of the connections between nodes in the cluster.

[0159] Determining Electrical Distance: Electrical relationships between nodes are more significant than spatial distances; therefore, electrical distance is chosen as the modularity index for calculating the weighted adjacency matrix, simultaneously measuring the structural performance of the cluster partitioning and the strength of electrical connections. In engineering practice, to simplify the method of obtaining electrical distance, the node impedance matrix is ​​often used to represent the electrical distance matrix. In the cluster partitioning method mentioned in this invention, the electrical distance between grounding points is still represented by an impedance matrix, but the two-port network input impedance Z is used. ij ′ indicates the electrical distance from the non-grounded point:

[0160] Z ij ′=Z ii +Z jj -2Z ij (2-2)

[0161] In step 1.2) above, the dominant node is selected primarily based on the monitoring and control of node voltage. That is, the selected dominant node must be both observable and controllable. Based on the characteristics of the dominant node, the comprehensive sensitivity S of all nodes in the distributed power generation cluster is calculated, and the node with the largest S value is the dominant node.

[0162] maxS=max(V+dC) (2-4)

[0163] Where V represents the observability of a node, C represents the controllability of a node, and d is the weight coefficient.

[0164]

[0165]

[0166] Where N is the set of all nodes in the cluster. Let be the node voltage sensitivity of node j to node i, and n be the set of controllable nodes in the cluster connected to distributed photovoltaic / energy storage devices. The reactive voltage sensitivity of the voltage amplitude at node i relative to the reactive power injected at node j.

[0167] In step 3) above, the total reactive power is:

[0168]

[0169] In the formula, ΔQ j Increase the total reactive power generated by the photovoltaic inverters within the safety cluster; S ijU is the reactive power-voltage sensitivity factor between the dominant node j of the safe cluster and the dominant node i of the dangerous cluster. min U represents the minimum terminal voltage of each node within the cluster. max U represents the maximum voltage at each node within the cluster. i Provides the real-time operating voltage for the i-node.

[0170] In step 3) above, the reactive power is allocated to each distributed photovoltaic power source within the cluster based on the total reactive power, including the following steps:

[0171] 3.1) Establish active / reactive voltage sensitivity matrices between clusters and between nodes;

[0172] In this embodiment, based on the power system load flow Jacobian matrix, the power flow calculation in the distribution network satisfies the following equation:

[0173]

[0174] The above equation, when transformed into a matrix, yields:

[0175]

[0176] The active / reactive voltage sensitivity matrix is ​​as follows:

[0177]

[0178]

[0179] 3.2) Based on the active / reactive voltage sensitivity matrix of the node, the relationship between the change in node voltage amplitude and the change in power is obtained, and the difference ΔU between the real-time voltage and the rated voltage of the node is obtained.

[0180] In this embodiment, based on the active / reactive voltage sensitivity matrix of the nodes, when n nodes in the cluster contain distributed photovoltaic / energy storage, the relationship matrix between the node voltage amplitude change and the power change is as follows:

[0181] ΔU=S UP ΔP+S UQ ΔQ (3-5)

[0182] The voltage Vi of each node in the cluster is affected not only by its own power changes, but also by the magnitude of active and reactive power injected by other nodes:

[0183]

[0184] 3.3) The reactive power change of each node is calculated based on the difference between the real-time voltage and the rated value of the node. The reactive power generated by each distributed photovoltaic power source in the cluster is obtained from the reactive power change K of each node and the reactive power-voltage sensitivity factor, so as to realize the distribution of reactive power from the total reactive power to each distributed photovoltaic power source in the cluster.

[0185] In this embodiment, as shown in equation (3-1), changing the reactive power output of distributed photovoltaic (PV) alters the voltage of this cluster or other clusters, and the impact of changes in the output power of distributed PV at different locations on the voltage amplitude change at the same point varies. To calculate the reactive power change K at each node, it is necessary to rationally allocate the power based on the reactive power-voltage sensitivity relationship of each node within the cluster. Given that the difference between the real-time voltage and the rated value of node a is ΔU, we have:

[0186]

[0187] in, This is the reactive power-voltage sensitivity coefficient.

[0188] The reactive power generated by each distributed photovoltaic power source within the cluster is then calculated as follows:

[0189]

[0190] In the formula ε i It is a Boolean value. When node i in the cluster is connected to a distributed photovoltaic power source and has adjustable power, its value is 1, otherwise it is 0.

[0191] Example:

[0192] A 33-node IEEE distribution network model with a high proportion of distributed photovoltaic power generation was built using MATLAB / Simulink, as shown in the figure. All generating units are of the same model. Figure 5 As shown, distributed photovoltaic power is connected to the distribution network via a transformer (311V / 12.66kV), where node 1 is the grid connection point for the distribution network to connect to the main power grid.

[0193] Clusters and nodes are defined as follows: In this embodiment, the nodes considered are the various busbars of the distribution network. In the example, the nodes are the 33 buses in the IEEE 33-node system. Furthermore, the clusters referred to in this embodiment specifically refer to the various cluster microgrids after clustering in a distribution network containing distributed photovoltaic access. In the example, each cluster is... Figure 5 The microgrid system consists of multiple nodes within the dashed box.

[0194] The clustering results of the IEEE 33-node distribution network system partitioning based on the electrical distance clustering algorithm are shown in Table 1.

[0195] Table 1 Clustering Results

[0196]

[0197] Without a photovoltaic system connected, if the grid connection point voltage is set to the rated voltage of 12.66 kV, the following results will be obtained: Figure 6 The voltage diagram of IEEE 33-node is shown.

[0198] A distributed photovoltaic (PV) power source grid-connected model via an inverter was constructed, including a PV power source model employing PQ power decoupling control and a PV power source model employing improved droop control suitable for the grid connection point voltage. The former assumes that the irradiance remains constant (P remains constant) for a short period, and changes in reactive power enable the distributed PV to participate in voltage regulation between / within clusters. The latter assumes that the grid connection point voltage rises at 0.02s.

[0199] Depend on Figures 7a to 7d As shown, in the power decoupling control mode, when the active output remains unchanged, changing the reactive output to participate in voltage regulation does not affect the active output, thus achieving decoupling control. In the Q(U) control mode, when the voltage at the monitoring point rises, the reactive output is automatically reduced to participate in voltage regulation, which can reduce voltage fluctuations and effectively regulate the local voltage to be close to the rated voltage.

[0200] Calculate the voltage change at other nodes caused by a unit change in reactive power at a certain node, i.e., reactive voltage sensitivity, and generate data such as... Figure 8 The sensitivity factor diagram shown has the x-axis representing nodes with unit reactive power changes, the y-axis representing nodes with voltage changes, and the z-axis representing the per-unit increase in voltage at the corresponding node after a unit reactive power change.

[0201] Calculations show that for clusters containing distributed photovoltaic (PV) systems, the control modes adopted based on the cluster voltage deviation are as follows: Cluster 5, containing PV nodes 20 and 22, has a voltage deviation within a stability margin of 0.005348. Therefore, PQ control can be used to participate in grid voltage regulation. Furthermore, the PV nodes connected to nodes 20 and 22 within cluster 5 still have reactive power margin. Based on the reactive power-voltage sensitivity relationship among the clusters, it is known that when nodes 20 and 22 increase reactive power generation, the voltage of the critical clusters 3, 4, and 7 increases, verifying the rationality of coordination between clusters. Other safe clusters can participate in grid-level voltage regulation using the same method.

[0202] Photovoltaic cluster 7, containing 31 nodes, exhibits a relatively large voltage deviation of 0.0533. An improved droop control based on virtual impedance is employed, with photovoltaic cells participating in voltage regulation. Photovoltaic cluster 3, containing 14 and 16 nodes, also shows a relatively large voltage deviation of 0.0606. An improved droop control based on virtual impedance is also employed, with photovoltaic cells participating in voltage regulation. Photovoltaic cluster 4, containing 18 nodes, also shows a relatively large voltage deviation of 0.0647. An improved droop control based on virtual impedance is also employed, with photovoltaic cells participating in voltage regulation. Comparing the cluster-based PQ control method, the system voltage magnitudes before and after optimization are as follows: Figure 9 As shown.

[0203] Depend on Figure 9 It can be seen that under the conventional PQ control method, cluster 3 still exceeds the limit, with a voltage deviation of 0.0506. However, the cluster voltage control method proposed in this paper significantly restores the distribution network voltage level. Calculations show that each cluster recovers to a safe voltage deviation level, all of which are reduced to below 0.05, indicating that it has good voltage regulation capability. This verifies the rationality of the voltage regulation control method based on cluster voltage deviation proposed in this paper.

[0204] In complex power distribution networks with numerous nodes, single grid-level control requires processing massive amounts of data, making it difficult to meet real-time control requirements and resulting in low data processing efficiency, thus failing to achieve the desired control effect. Single station-level control, relying solely on local information, struggles to maximize voltage coordination among nodes. Therefore, cluster-level control, as a comprehensive approach, significantly improves voltage control performance. First, the actual power distribution network containing distributed photovoltaic (PV) power is divided into clusters based on improved electrical distance. The control mode for grid-connected inverters is set to automatically switch based on voltage exceedance levels. Then, the voltage exceedance level of each cluster is evaluated based on the cluster voltage deviation index. The dominant node of each cluster is selected based on the reactive power-voltage sensitivity matrix between nodes, and the voltage sensitivity relationship between clusters is calculated. Finally, a voltage control mode based on cluster voltage deviation is applied to the distributed PV inverters within each cluster to achieve grid voltage stability. The voltage deviation control mode for power clusters is as follows: when a voltage deviation in a cluster exceeds a set threshold, it is considered a dangerous cluster, and the improved Q(U) mode is preferentially used for local voltage control in this cluster. For other safe clusters with voltage deviations within the threshold safety range, the safe cluster with the largest sensitivity factor between the dominant node in each cluster and the dominant node of the dangerous cluster is selected. The inverters in this safe cluster generate an additional ΔQ, and then the system returns to determine whether the dangerous cluster has entered the safe domain. If not, a safe cluster is selected for voltage regulation. The reactive power allocation criterion within the safe cluster is that inverters with adjustable reactive power capacity are allocated proportionally according to their sensitivity factors. This method simplifies grid voltage control, avoiding the increased workload and reduced real-time effectiveness of voltage control caused by processing large amounts of data from various nodes in the distribution network. Furthermore, by using different control modes for each cluster, it has stronger targeting and adaptability, and considering the coordinated control between clusters, it enables the grid voltage to stabilize within the safe range more quickly.

[0205] In summary, this invention, based on dual closed-loop voltage and current control, employs improved droop control and PQ control mode switching for grid-connected inverters to establish a distributed photovoltaic (PV) power grid connection model. Then, using the cluster partitioning results and voltage deviation level indicators of each cluster, the risk of voltage exceeding limits in each cluster is assessed, establishing reactive power and voltage sensitivity matrices between clusters and nodes. Finally, a voltage adaptive control mode is applied to the distributed PV inverters within each cluster to achieve grid voltage stability. Simulation results demonstrate that the proposed cluster voltage support method effectively adapts to and regulates grid operating voltage, thus verifying the effectiveness of the method.

[0206] In one embodiment of the present invention, a distributed photovoltaic multi-cluster voltage control system is provided, comprising:

[0207] The initial partitioning module uses the cluster partitioning results and the cluster voltage deviation of each cluster to initially determine safe and dangerous clusters, and performs local voltage regulation on dangerous clusters.

[0208] If the local voltage regulation fails in the cluster determination module, the reactive power-voltage sensitivity factor between each cluster is calculated, and the safe and dangerous clusters are re-determined based on the reactive power-voltage sensitivity factor.

[0209] The power distribution module selects the safe cluster with the largest sensitivity factor between the dominant node in each cluster and the dominant node in the dangerous cluster, increases the total reactive power ΔQ generated by the inverters in the safe cluster, and distributes reactive power to each distributed photovoltaic power source in the cluster based on the total reactive power.

[0210] The detection module continuously monitors the voltage deviation of each cluster. If a cluster is still in danger, a safe cluster is redefined. Otherwise, the voltage of each cluster is within the safety margin, thus achieving grid-level voltage stability control.

[0211] The system provided in this embodiment is used to execute the above-described method embodiments. For specific processes and details, please refer to the above embodiments, which will not be repeated here.

[0212] One embodiment of the present invention provides a computing device structure, which can be a terminal and may include: a processor, a communication interface, memory, a display screen, and an input device. The processor, communication interface, and memory communicate with each other via a communication bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and a computer program, which, when executed by the processor, implements a control method; the internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The communication interface is used for wired or wireless communication with external terminals. Wireless communication can be achieved through Wi-Fi, a management network, NFC (Near Field Communication), or other technologies. The display screen can be a liquid crystal display or an e-ink display. The input device can be a touch layer covering the display screen, or buttons, a trackball, or a touchpad mounted on the casing of the computing device, or an external keyboard, touchpad, or mouse, etc. The processor can call logic instructions in memory to execute the following methods: First, preliminarily determine safe and dangerous clusters based on the cluster partitioning results and the cluster voltage deviation of each cluster, and perform local voltage regulation on dangerous clusters; if local voltage regulation fails, calculate the reactive power-voltage sensitivity factor between each cluster, and re-determine safe and dangerous clusters based on the reactive power-voltage sensitivity factor; select the safe cluster with the largest sensitivity factor between the dominant node in each cluster and the dominant node of the dangerous cluster, increase the total reactive power ΔQ generated by the inverters in this safe cluster, and allocate reactive power to each distributed photovoltaic power source within the cluster based on this total reactive power; cyclically detect the voltage deviation of each cluster; if it is still in a dangerous cluster, re-determine a safe cluster; otherwise, the voltage of each cluster is within the safety margin, achieving grid-level voltage stability control.

[0213] Furthermore, the logical instructions in the aforementioned memory can be implemented as software functional units and sold or used as independent products, and can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0214] Those skilled in the art will understand that the structure of the computer device described above is only a part of the structure related to the solution of this application, and does not constitute a limitation on the computing device on which the solution of this application is applied. The specific computing device may include more or fewer components, or combine certain components, or have different component arrangements.

[0215] In one embodiment of the present invention, a computer program product is provided, the computer program product including a computer program stored on a non-transitory computer-readable storage medium, the computer program including program instructions, when the program instructions are executed by a computer, the computer can execute the methods provided in the above-described method embodiments, for example including: initially determining safe clusters and dangerous clusters using cluster partitioning results and cluster voltage deviation of each cluster, and performing local voltage regulation on dangerous clusters; if local voltage regulation fails, calculating the reactive power-voltage sensitivity factor between each cluster, and re-determining safe clusters and dangerous clusters based on the reactive power-voltage sensitivity factor; selecting the safe cluster with the largest sensitivity factor between the dominant node in each cluster and the dominant node of the dangerous cluster, increasing the total reactive power ΔQ of the inverter in the safe cluster, and distributing reactive power to each distributed photovoltaic power source in the cluster based on the total reactive power; cyclically detecting the voltage deviation of each cluster, if it is still in a dangerous cluster, re-determining a safe cluster, otherwise the voltage of each cluster is within the safety margin, achieving grid-level voltage stability control.

[0216] In one embodiment of the present invention, a non-transitory computer-readable storage medium is provided, which stores server instructions that cause a computer to execute the methods provided in the above embodiments, including, for example,: initially determining safe clusters and dangerous clusters using cluster partitioning results and cluster voltage deviation of each cluster, and performing local voltage regulation on dangerous clusters; if local voltage regulation fails, calculating the reactive power-voltage sensitivity factor between each cluster, and re-determining safe clusters and dangerous clusters based on the reactive power-voltage sensitivity factor; selecting the safe cluster with the largest sensitivity factor between the dominant node in each cluster and the dominant node of the dangerous cluster, increasing the total reactive power ΔQ of the inverters in the safe cluster, and distributing reactive power to each distributed photovoltaic power source in the cluster based on the total reactive power; cyclically detecting the voltage deviation of each cluster, and if it is still in a dangerous cluster, re-determining a safe cluster, otherwise ensuring that the voltage of each cluster is within the safety margin, thereby achieving grid-level voltage stability control.

[0217] The computer-readable storage medium provided in the above embodiments has a similar implementation principle and technical effect to the above method embodiments, and will not be described again here.

[0218] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0219] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0220] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0221] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A distributed photovoltaic multi-cluster voltage control method, characterized in that, include: Based on the cluster partitioning results and the cluster voltage deviation of each cluster, safe and dangerous clusters are initially identified, and the dangerous clusters are regulated locally. If the local voltage regulation fails, the reactive power-voltage sensitivity factor between each cluster is calculated, and the safe cluster and the dangerous cluster are re-determined based on the reactive power-voltage sensitivity factor. Select the safe cluster with the largest sensitivity factor between the dominant node in each cluster and the dominant node in the dangerous cluster, increase the total reactive power ΔQ of the inverters in the safe cluster, and allocate reactive power to each distributed photovoltaic power source in the cluster based on the total reactive power. The voltage deviation of each cluster is continuously monitored. If the cluster is still in danger, a safe cluster is redefined. Otherwise, the voltage of each cluster is within the safety margin, thus achieving grid-level voltage stability control. The determination of the cluster voltage deviation includes: The microgrids of each photovoltaic cluster are obtained based on the cluster partitioning results; For each node in the cluster, calculate the reactive voltage sensitivity of each node to the dominant node; The voltage deviation of each cluster is calculated based on the number of nodes in the cluster and the reactive voltage sensitivity. The voltage deviation is: ， In the formula, N represents the cluster voltage deviation; N is the number of nodes in the cluster. For the real-time operating voltage of the i-node, For node i, for the dominant node The reactive voltage sensitivity coefficient, This represents the minimum voltage at each node within the cluster. This represents the maximum voltage at each node within the cluster.

2. The distributed photovoltaic multi-cluster voltage control method as described in claim 1, characterized in that, The preliminary determination of safe and dangerous clusters using the cluster division results and the cluster voltage deviation of each cluster includes: using the cluster division results and the cluster voltage deviation of each cluster to determine whether the voltage in each cluster exceeds the limit; if it exceeds the limit, it is a dangerous cluster, otherwise it is a safe cluster.

3. The distributed photovoltaic multi-cluster voltage control method as described in claim 1, characterized in that, The on-site pressure regulation of the hazardous cluster includes: The inverters with adjustable reactive power capacity within the hazardous cluster are scheduled to switch to an improved droop control mode for local voltage regulation.

4. The distributed photovoltaic multi-cluster voltage control method as described in claim 1, characterized in that, The total reactive power is: In the formula, To increase the total reactive power generated by the photovoltaic inverters within the safety cluster; The reactive power-voltage sensitivity factor between the dominant node j of the safe cluster and the dominant node i of the dangerous cluster. This represents the minimum voltage at each node within the cluster. This represents the maximum voltage at each node within the cluster. Provides the real-time operating voltage for the i-node.

5. The distributed photovoltaic multi-cluster voltage control method as described in claim 1, characterized in that, The allocation of reactive power to each distributed photovoltaic power source within the cluster based on the total reactive power includes: Establish active / reactive voltage sensitivity matrices between clusters and between nodes; The relationship between the change in node voltage amplitude and the change in power is obtained from the active / reactive voltage sensitivity matrix of the node, and the difference between the real-time voltage and the rated voltage of the node is obtained. The photovoltaic reactive power variation at each node is calculated based on the difference between the real-time voltage and the rated value. The reactive power generated by each distributed photovoltaic power source within the cluster is obtained by combining the reactive power-voltage sensitivity factor, thereby realizing the distribution of reactive power from the total reactive power to each distributed photovoltaic power source within the cluster.

6. A distributed photovoltaic multi-cluster voltage control system, characterized in that, include: The initial partitioning module uses the cluster partitioning results and the cluster voltage deviation of each cluster to initially determine safe clusters and dangerous clusters, and performs local voltage regulation on the dangerous clusters. If the local voltage regulation fails, the cluster determination module calculates the reactive power-voltage sensitivity factor between each cluster and redetermines the safe and dangerous clusters based on the reactive power-voltage sensitivity factor. The power distribution module selects the safe cluster with the largest sensitivity factor between the dominant node in each cluster and the dominant node in the dangerous cluster, increases the total reactive power ΔQ generated by the inverters in the safe cluster, and distributes reactive power to each distributed photovoltaic power source in the cluster based on the total reactive power. The detection module cyclically detects the voltage deviation of each cluster. If it is still in a dangerous cluster, it redetermines a safe cluster. Otherwise, the voltage of each cluster is within the safety margin, thus achieving grid-level voltage stability control. The determination of the cluster voltage deviation includes: The microgrids of each photovoltaic cluster are obtained based on the cluster partitioning results; For each node in the cluster, calculate the reactive voltage sensitivity of each node to the dominant node; The voltage deviation of each cluster is calculated based on the number of nodes in the cluster and the reactive voltage sensitivity. The voltage deviation is: ， In the formula, N represents the cluster voltage deviation; N is the number of nodes in the cluster. For the real-time operating voltage of the i-node, For node i, for the dominant node The reactive voltage sensitivity coefficient, This represents the minimum voltage at each node within the cluster. This represents the maximum voltage at each node within the cluster.

7. A computer-readable storage medium for storing one or more programs, characterized in that, The one or more programs include instructions that, when executed by a computing device, cause the computing device to perform any of the methods described in claims 1 to 5.

8. A computing device, characterized in that, include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions for performing any of the methods described in claims 1 to 5.

Images