Control method for realizing dynamic voltage support by coordinating multiple inverters in photovoltaic cluster

Abstract

The invention relates to the field of voltage regulation, in particular to a control method for realizing dynamic voltage support by coordinating multiple inverters in a photovoltaic cluster. Therefore, the system voltage stability and the control precision are improved. The method comprises the following steps: firstly, establishing a reactive power coordinated regulation and control mechanism of each photovoltaic unit, and then constructing a photovoltaic / cluster system dynamic characteristic mathematical characterization model containing an auxiliary variable z; it is ensured that the common connection point voltage converges to a reference value, optimal tracking control over the reference value by the PCC voltage is achieved, and multi-access photovoltaic local voltage collaborative dynamic adjustment is achieved; and finally, the PCC voltage is maintained at the optimal set value by regulating and controlling multiple inverters in a reactive power coordination cluster, the inverter control method is constructed based on a system stable track, and a technical path is provided for efficient voltage control of the photovoltaic cluster by designing a specific performance index tracking controller, guaranteeing system stability and control response and obtaining expected closed loop characteristics.

Description

Technical Field

[0001] This invention relates to the field of voltage regulation, and more particularly to a control method for coordinating multiple inverters within a photovoltaic cluster to achieve dynamic voltage support. Background Technology

[0002] The future smart distribution network is developing towards a high proportion of photovoltaic (PV) penetration. While large-scale PV integration provides important support for clean energy, it also significantly increases the complexity of system voltage management. Conventional voltage regulation methods (such as traditional on-load tap-changing transformers and fixed-capacity capacitor banks) can cope with normal load fluctuations, but their response speed is mostly on the order of minutes, making it difficult to match the second- or even millisecond-level voltage fluctuations in PV output caused by changes in sunlight and temperature. This leads to frequent voltage deviations at PV grid connection points. To overcome this dilemma, building a PV cluster control system through a point of common coupling (PCC) has become a key research direction. This strategy aims to maintain a stable voltage at the PCC node, which can improve the system management efficiency after a high proportion of distributed energy is connected to the grid. However, to achieve accurate tracking of the PCC reference voltage and ensure that the response speed matches the characteristics of PV power fluctuations, it is necessary to overcome the limitations of independent control of individual units and establish an efficient reactive power collaborative control mechanism among PV units.

[0003] Although various algorithms exist to achieve basic regulation of the PCC point voltage, their designs often focus on global voltage stability while neglecting the voltage state of the local nodes within the photovoltaic unit. This "global-centric, local-neglecting" control logic can easily lead to some photovoltaic units being forced to shut down due to their local voltage exceeding the allowable limit, triggering protection mechanisms. This not only causes power generation losses in individual photovoltaic systems but also disrupts the overall power balance of the cluster, further affecting the stability of the PCC voltage. Summary of the Invention

[0004] To completely solve the above problems, this invention proposes a control method for coordinating multiple inverters within a photovoltaic cluster to achieve dynamic voltage support. This method coordinates multiple photovoltaic inverters to form a dynamic voltage support cluster. It utilizes real-time data to construct a complete control framework, eliminating the need for prior acquisition of grid system parameters. This technology integrates a controller and a power allocation mechanism, collaboratively scheduling the operation of multiple photovoltaic inverters within the cluster, and prioritizing units with larger local voltage deviations for voltage support. The specific technical solution is as follows:

[0005] A control method for coordinating multiple inverters within a photovoltaic cluster to achieve dynamic voltage support includes:

[0006] The target is to reduce the voltage at the point of common link (PCC) of the photovoltaic cluster system. Adjust and maintain at the reference value;

[0007] A mathematical characterization model of the dynamic characteristics of a photovoltaic cluster system is constructed, where Z is an auxiliary variable in the model that forms the trajectory image space of the system.

[0008] Design a photovoltaic inverter controller to construct a stable trajectory Z in the image representation;

[0009] Constructing a reactive power collaborative allocation mechanism for photovoltaic clusters:

[0010] Conduct voltage and reactive power sensitivity analysis of photovoltaic clusters, and calculate the sensitivity of each photovoltaic inverter to PCC voltage and its local node voltage.

[0011] Based on the above sensitivity analysis results, a generalized expression for the droop gain of each photovoltaic inverter controller is proposed; a weighted bias mechanism is constructed by integrating the sensitivity of the PCC voltage of each photovoltaic inverter with the local node voltage sensitivity.

[0012] By monitoring the deviation between the PCC voltage and the voltages of each local node in real time, the droop gain parameters of each photovoltaic inverter controller are dynamically adjusted. Based on the adjusted droop gain parameters, the corresponding reactive power redistribution mechanism is triggered to rationally allocate reactive power regulation tasks to each photovoltaic inverter, achieving coordinated dynamic regulation of the local voltages of multiple connected photovoltaic systems. This enables optimal tracking control of the PCC voltage to the reference value.

[0013] (3)

[0014] Ensure that the PCC voltage converges to the reference value.

[0015] Furthermore, in the design of the photovoltaic inverter controller, each photovoltaic inverter adjusts the q-axis current in the dq rotating coordinate system through a built-in proportional-integral controller, thereby injecting or absorbing reactive power as needed. This provides a foundation for the subsequent construction of a photovoltaic cluster reactive power collaborative allocation mechanism and helps to achieve collaborative dynamic adjustment of the local voltage of multiple connected photovoltaic systems.

[0016] The formula for the coordinated dynamic adjustment of the local voltage of the multi-access photovoltaic system is as follows:

[0017] (4)

[0018] In the formula, j represents the bus number connected to the photovoltaic unit; For the corresponding The local voltage of bus j at time j; The optimal reference voltage for bus j; It represents the allowable deviation range of the voltage; it achieves optimal tracking control of the PCC voltage to the reference value, and ultimately achieves the control objective.

[0019] Furthermore, the specific steps for constructing the mathematical representation model of the dynamic characteristics of the photovoltaic cluster system are as follows:

[0020] For the port variable w, select a duration of The input and output measurement data trajectories are used to construct a Hankel matrix from the measured values, as defined below:

[0021] (8)

[0022] The order of the Hankel matrix is ​​set to . N represents the maximum lag order of the system;

[0023] To ensure that the system behavior can be accurately represented by the measurement data, the input signal must meet the condition of continuous excitation;

[0024] Let the system state space have dimension n. If the input sequence is given by... The constructed Hankel matrix If the rows are full rank, then the input sequence is said to have rank. Continuous excitation characteristics;

[0025] When the PE condition is satisfied, the Hankel matrix It can completely represent all legal trajectories of the system at lag order N; this conclusion holds true only if the trajectory length is measured. Must meet This ensures that the data margin is sufficient to accurately describe the dynamic characteristics of the system, laying the foundation for the subsequent design of the photovoltaic inverter controller and the construction of the stable trajectory Z;

[0026] When the continuous excitation condition is satisfied and the trajectory length is When the above requirements are met, the dynamic behavior characteristics of the system can be fully characterized based on the available measurement data;

[0027] Hankel matrix The Singular Value Decomposition (SVD) expression is: ,in and It is an orthogonal matrix. It includes A diagonal matrix of non-negative singular values;

[0028] Hankel matrix The rank of the singular value matrix is ​​denoted by r, and its value is equal to that of the singular value matrix. The number of non-zero singular values; due to exist There is There are zero rows, therefore U contains vectors with minified w; partition U into ,in Composed of the first r columns, the Hankel matrix can be derived. The explicit expression for the left kernel space is: ;for ,parameter and It can be determined;

[0029] These parameters can be expressed in the image representation as:

[0030] ;

[0031] In the formula, z is an auxiliary variable in the image space of the system trajectory, which is used in the design of photovoltaic inverter controllers to achieve the goal of constructing a stable trajectory Z.

[0032] Furthermore, when designing a photovoltaic inverter controller, the trajectory Z can be constructed as a solution to a difference equation of the following form:

[0033] (12);

[0034] The stability criterion for this controller requires that the roots of its characteristic equation lie inside the unit circle, satisfying the following condition: Where Re and Im represent the real and imaginary parts of the root, respectively;

[0035] Closed-loop control behavior can be specified by specifying the desired poles. and These poles are defined to determine the dynamic characteristics of the system, and are realized through the following characteristic equation:

[0036] ;

[0037] To achieve the desired time response characteristics (such as settling time) ), need to select and The parameter values; the relationship between the settling time and the location of the poles is determined by... Establish (among others) (where the sampling period is the reference period), this relationship provides a clear basis for parameter selection, ensuring optimal tracking control of the PCC voltage to the reference value in the subsequent implementation;

[0038] Controller gain parameters and The design must satisfy the relationship in equation (13); the expression can be obtained by direct derivation: , ;

[0039] The photovoltaic inverter controller defines and locks the closed-loop trajectory, and the system dynamic equation is represented as follows:

[0040] ;

[0041] Input sequence The generation strategy aims to ensure controlled quantities Precisely track the preset trajectory The typical implementation is as follows:

[0042] ;

[0043] Convergence rate parameter Directly determines the dynamic performance of the system; corresponding control input quantities Obtained through inversion calculation:

[0044] ;

[0045] When function When reversible, this control law ensures that the system response has a predictable and deterministic structure, completely avoiding the gain adjustment process, and providing a reliable guarantee for the stable operation of the photovoltaic cluster system and the achievement of control objectives.

[0046] Furthermore, when conducting voltage and reactive power sensitivity analysis of photovoltaic clusters, a sensitivity coefficient is introduced considering the characteristics of clustered distribution network systems. This is used to quantify the impact of execution unit j on the voltage of node i, providing key parameter support for the subsequent construction of a photovoltaic cluster reactive power collaborative allocation mechanism and the proposal of a droop gain generalization expression.

[0047] Sensitivity is defined as:

[0048] ;

[0049] In the formula, It is caused by changes in reactive power. The resulting change in node i voltage and These are the nominal values ​​of the voltage at node i and the reactive power of execution unit j, respectively; when node i is a PCC, This refers to the voltage change at the PCC, which can be used to analyze the sensitivity of each photovoltaic inverter to the PCC voltage.

[0050] This sensitivity can also be determined by the fluctuation of the following measurement data:

[0051] ;

[0052] ;

[0053] Using the least squares method, the sensitivity can be approximately calculated as follows:

[0054] ;

[0055] The sensitivity calculated in the above way can accurately reflect the correlation between the execution unit and the node voltage, laying a data foundation for subsequent dynamic adjustment of the droop gain parameters of each photovoltaic inverter and achieving reasonable allocation of reactive power.

[0056] Furthermore, the coordinated dynamic adjustment of the local voltage of the multi-access photovoltaic system is implemented as follows:

[0057] To achieve coordinated power distribution, a droop control equation is used to dynamically adjust the control reference values ​​of each photovoltaic inverter, thereby improving voltage performance indicators. Transition to To construct an effective negative feedback mechanism and ensure the stability and accuracy of voltage regulation;

[0058] For photovoltaic node j, the control reference value update strategy is as follows:

[0059] ;

[0060] ;

[0061] In the formula, The target voltage performance index for closed-loop control is consistent with the objective of ensuring that the PCC voltage converges to the reference value. This represents the actual reactive power output of the photovoltaic inverter at node j. This is the rated reference reactive power of the photovoltaic inverter;

[0062] Integral gain A smaller positive value is chosen to maintain system stability and suppress overshoot, thus avoiding the impact of improper parameters on the dynamic adjustment effect of the photovoltaic cluster system.

[0063] Given sensitivity The droop control gain directly characterizes the photovoltaic unit's ability to compensate for voltage fluctuations. To ensure that the compensation response strength matches the photovoltaic inverter's regulation capability, the droop control gain is defined as: ;

[0064] To address the need for a weighted bias mechanism, a generalized expression for the droop gain of the photovoltaic inverter controller is proposed based on the weighted values.

[0065] ;

[0066] in, It is an averaging factor; the voltage deviation at PCC and the voltage deviation at the local node j of the photovoltaic inverter are respectively defined as... and :

[0067] ;

[0068] ;

[0069] In the formula, and These represent the average voltage of PCC and local node j relative to the reference value within a time window of length T. and The deviation is an important quantitative result for real-time monitoring of the deviation between the PCC voltage and the voltage of each local node;

[0070] By integrating the sensitivity of the photovoltaic inverter to the PCC ( ) and its sensitivity to local node j ( We construct a weighted bias mechanism to further improve the droop gain generalization expression and ensure the rationality of parameter adjustment.

[0071] To implement the reactive power allocation strategy, each photovoltaic inverter controller is reconfigured in the following manner:

[0072] (32);

[0073] By adjusting node parameters in real time And its corresponding reactive power redistribution mechanism, to achieve distributed coordination and efficient voltage regulation among actuators, ultimately helping to achieve optimal tracking control of PCC voltage to the reference value, and achieving the control objective of the photovoltaic cluster system.

[0074] The intended effects of this invention are as follows:

[0075] 1. The primary advantage of this invention lies in completely eliminating the reliance of traditional control on prior knowledge of system parameters and structure, thereby improving control robustness and universality. Traditional methods require prior acquisition of information such as system topology and line impedance. Once the operating conditions change or parameters are incorrect, it can easily lead to control inaccuracy or even instability.

[0076] 2. The method of this invention relies solely on measured operating data such as voltage and reactive power. The model parameters are directly calculated from the data by singular value decomposition of the Hankel matrix. No pre-set model is required. It can be adapted to photovoltaic clusters of different sizes and topologies, greatly reducing engineering commissioning and adaptation costs.

[0077] 3. Regarding voltage regulation, this invention achieves a balance between the PCC voltage control target and local voltage safety, resolving the issue of ambiguous priorities in traditional solutions. Through a two-stage voltage regulation target design, it ensures that the PCC voltage accurately follows the ideal 1p.u. reference value in steady state using an optimized tracking algorithm.

[0078] 4. During the dynamic process of this invention, the method prioritizes the regulation of photovoltaic units with local voltage exceeding the limit. It first corrects the local voltage through reactive power injection / absorption, and then synchronously maintains the PCC voltage stability to avoid cluster-level failures caused by local instability, thus achieving the unity of "global stability" and "local safety".

[0079] 5. The fully distributed architecture design of this invention is the core breakthrough of this method in terms of system efficiency and scalability, overcoming the drawbacks of traditional centralized control. Traditional centralized solutions rely on a central controller to collect data and calculate instructions, which suffers from problems such as large data transmission volume, concentrated computing load, high risk of single point of failure, and a sharp drop in efficiency as the scale increases.

[0080] 6. In terms of reactive power distribution and voltage response, this invention achieves precise coordination and efficiency optimization, avoiding unit overload and capacity idleness caused by traditional average or fixed-ratio distribution. It quantifies the adjustment capability of each unit through voltage-reactive power sensitivity and accurately calculates sensitivity using the least squares method, ensuring that the compensation intensity matches the capability.

[0081] 7. This invention introduces a local voltage deviation weighting factor to prioritize the scheduling of low-voltage units, and dynamically updates the reference value using the droop control equation. Combined with a stable trajectory-based controller construction, the stabilization time is shortened by configuring poles, and manual adjustment of PI parameters is unnecessary, reducing maintenance costs and improving voltage regulation response speed and economy.

[0082] 8. This invention's method is highly adaptable to high-proportion photovoltaic grid-connected scenarios, effectively enhancing the grid's voltage support capability. With the increasing penetration rate of photovoltaics, the grid faces challenges such as frequent voltage fluctuations and insufficient inertia. Traditional inverters focus on active power output, resulting in weak voltage support.

[0083] 9. This invention enables photovoltaic inverters to dynamically adjust reactive power output in real time through PI control of the q-axis current in the dq coordinate system: rapidly injecting reactive power when voltage drops and absorbing reactive power when voltage rises. Furthermore, the distributed architecture avoids communication delays, ensuring real-time adjustment and effectively suppressing voltage spikes and drops caused by cloud cover, thereby improving grid stability under high-proportion grid connection conditions.

[0084] 10. This invention not only achieves precise control of the PCC voltage of photovoltaic clusters, but also ensures that the voltage of all nodes remains within the allowable range. This demonstrates its strong adaptability to system fluctuations and highlights the application potential of this technology in enhancing grid stability in renewable energy grid-connected scenarios. It is hoped that this invention will provide reference and guidance for photovoltaic voltage regulation. Attached Figure Description

[0085] Figure 1 This is a schematic diagram of voltage regulation for a photovoltaic inverter at the point of common coupling.

[0086] Figure 2This is a topology diagram of a large-scale photovoltaic system in cluster mode;

[0087] Figure 3 This is a diagram of the internal circuit topology of a single photovoltaic system;

[0088] Figure 4 This is the transient response curve of PCC voltage startup;

[0089] Figure 5 This is a comparison chart of local voltage control strategies for photovoltaic systems;

[0090] Figure 6 This is a diagram showing the steady-state voltage distribution characteristics of a dual-feeder system.

[0091] Figure 7 This is a graph showing the time-varying disturbance input characteristics of irradiance;

[0092] Figure 8 This is a timing response curve of PCC voltage immunity;

[0093] Figure 9 This is a comparison chart of local voltage control strategies for photovoltaic systems under irradiation disturbances;

[0094] Figure 10 This is a diagram showing the steady-state voltage distribution characteristics of the dual-feeder system after the disturbance.

[0095] Figure 11 This is a table showing the irradiance levels of various photovoltaic systems and their corresponding operating conditions. Detailed Implementation

[0096] Specific embodiments of the present invention are as follows:

[0097] Example 1

[0098] Referring to the attached diagram, a control method for coordinating multiple inverters within a photovoltaic cluster to achieve dynamic voltage support includes:

[0099] The target is to reduce the voltage at the point of common link (PCC) of the photovoltaic cluster system. Adjust and maintain at the reference value;

[0100] A mathematical characterization model of the dynamic characteristics of a photovoltaic cluster system is constructed, where Z is an auxiliary variable in the model that forms the trajectory image space of the system.

[0101] Design a photovoltaic inverter controller to construct a stable trajectory Z in the image representation;

[0102] Constructing a reactive power collaborative allocation mechanism for photovoltaic clusters:

[0103] Conduct voltage and reactive power sensitivity analysis of photovoltaic clusters, and calculate the sensitivity of each photovoltaic inverter to PCC voltage and its local node voltage.

[0104] Based on the above sensitivity analysis results, a generalized expression for the droop gain of each photovoltaic inverter controller is proposed; a weighted bias mechanism is constructed by integrating the sensitivity of the PCC voltage of each photovoltaic inverter with the local node voltage sensitivity.

[0105] By monitoring the deviation between the PCC voltage and the voltages of each local node in real time, the droop gain parameters of each photovoltaic inverter controller are dynamically adjusted. Based on the adjusted droop gain parameters, the corresponding reactive power redistribution mechanism is triggered to rationally allocate reactive power regulation tasks to each photovoltaic inverter, achieving coordinated dynamic regulation of the local voltages of multiple connected photovoltaic systems. This enables optimal tracking control of the PCC voltage to the reference value.

[0106] (3)

[0107] Ensure that the PCC voltage converges to the reference value.

[0108] Furthermore, in the design of the photovoltaic inverter controller, each photovoltaic inverter adjusts the q-axis current in the dq rotating coordinate system through a built-in proportional-integral controller, thereby injecting or absorbing reactive power as needed. This provides a foundation for the subsequent construction of a photovoltaic cluster reactive power collaborative allocation mechanism and helps to achieve collaborative dynamic adjustment of the local voltage of multiple connected photovoltaic systems.

[0109] The formula for the coordinated dynamic adjustment of the local voltage of the multi-access photovoltaic system is as follows:

[0110] (4)

[0111] In the formula, j represents the bus number connected to the photovoltaic unit; For the corresponding The local voltage of bus j at time j; The optimal reference voltage for bus j; It represents the allowable deviation range of the voltage; it achieves optimal tracking control of the PCC voltage to the reference value, and ultimately achieves the control objective.

[0112] Furthermore, the specific steps for constructing the mathematical representation model of the dynamic characteristics of the photovoltaic cluster system are as follows:

[0113] For the port variable w, select a duration of The input and output measurement data trajectories are used to construct a Hankel matrix from the measured values, as defined below:

[0114] (8)

[0115] The order of the Hankel matrix is ​​set to . N represents the maximum lag order of the system;

[0116] To ensure that the system behavior can be accurately represented by the measurement data, the input signal must meet the condition of continuous excitation;

[0117] Let the system state space have dimension n. If the input sequence is given by... The constructed Hankel matrix If the rows are full rank, then the input sequence is said to have rank. Continuous excitation characteristics;

[0118] When the PE condition is satisfied, the Hankel matrix It can completely represent all legal trajectories of the system at lag order N; this conclusion holds true only if the trajectory length is measured. Must meet This ensures that the data margin is sufficient to accurately describe the dynamic characteristics of the system, laying the foundation for the subsequent design of the photovoltaic inverter controller and the construction of the stable trajectory Z;

[0119] When the continuous excitation condition is satisfied and the trajectory length is When the above requirements are met, the dynamic behavior characteristics of the system can be fully characterized based on the available measurement data;

[0120] Hankel matrix The Singular Value Decomposition (SVD) expression is: ,in and It is an orthogonal matrix. It includes A diagonal matrix of non-negative singular values;

[0121] Hankel matrix The rank of the singular value matrix is ​​denoted by r, and its value is equal to that of the singular value matrix. The number of non-zero singular values; due to exist There is There are zero rows, therefore U contains vectors with minified w; partition U into ,in Composed of the first r columns, the Hankel matrix can be derived. The explicit expression for the left kernel space is: ;for ,parameter and It can be determined;

[0122] These parameters can be expressed in the image representation as:

[0123] ;

[0124] In the formula, z is an auxiliary variable in the image space of the system trajectory, which is used in the design of photovoltaic inverter controllers to achieve the goal of constructing a stable trajectory Z.

[0125] Furthermore, when designing a photovoltaic inverter controller, the trajectory Z can be constructed as a solution to a difference equation of the following form:

[0126] (12);

[0127] The stability criterion for this controller requires that the roots of its characteristic equation lie inside the unit circle, satisfying the following condition: Where Re and Im represent the real and imaginary parts of the root, respectively;

[0128] Closed-loop control behavior can be specified by specifying the desired poles. and These poles are defined to determine the dynamic characteristics of the system, and are realized through the following characteristic equation:

[0129] ;

[0130] To achieve the desired time response characteristics (such as settling time) ), need to select and The parameter values; the relationship between the settling time and the location of the poles is determined by... Establish (among others) (where the sampling period is the reference period), this relationship provides a clear basis for parameter selection, ensuring optimal tracking control of the PCC voltage to the reference value in the subsequent implementation;

[0131] Controller gain parameters and The design must satisfy the relationship in equation (13); the expression can be obtained by direct derivation: , ;

[0132] The photovoltaic inverter controller defines and locks the closed-loop trajectory, and the system dynamic equation is represented as follows:

[0133] ;

[0134] Input sequence The generation strategy aims to ensure controlled quantities Precisely track the preset trajectory The typical implementation is as follows:

[0135] ;

[0136] Convergence rate parameter Directly determines the dynamic performance of the system; corresponding control input quantities Obtained through inversion calculation:

[0137] ;

[0138] When function When reversible, this control law ensures that the system response has a predictable and deterministic structure, completely avoiding the gain adjustment process, and providing a reliable guarantee for the stable operation of the photovoltaic cluster system and the achievement of control objectives.

[0139] Furthermore, when conducting voltage and reactive power sensitivity analysis of photovoltaic clusters, a sensitivity coefficient is introduced considering the characteristics of clustered distribution network systems. This is used to quantify the impact of execution unit j on the voltage of node i, providing key parameter support for the subsequent construction of a photovoltaic cluster reactive power collaborative allocation mechanism and the proposal of a droop gain generalization expression.

[0140] Sensitivity is defined as:

[0141] ;

[0142] In the formula, It is caused by changes in reactive power. The resulting change in node i voltage and These are the nominal values ​​of the voltage at node i and the reactive power of execution unit j, respectively; when node i is a PCC, This refers to the voltage change at the PCC, which can be used to analyze the sensitivity of each photovoltaic inverter to the PCC voltage.

[0143] This sensitivity can also be determined by the fluctuation of the following measurement data:

[0144] ;

[0145] ;

[0146] Using the least squares method, the sensitivity can be approximately calculated as follows:

[0147] ;

[0148] The sensitivity calculated in the above way can accurately reflect the correlation between the execution unit and the node voltage, laying a data foundation for subsequent dynamic adjustment of the droop gain parameters of each photovoltaic inverter and achieving reasonable allocation of reactive power.

[0149] Furthermore, the coordinated dynamic adjustment of the local voltage of the multi-access photovoltaic system is implemented as follows:

[0150] To achieve coordinated power distribution, a droop control equation is used to dynamically adjust the control reference values ​​of each photovoltaic inverter, thereby improving voltage performance indicators. Transition to To construct an effective negative feedback mechanism and ensure the stability and accuracy of voltage regulation;

[0151] For photovoltaic node j, the control reference value update strategy is as follows:

[0152] ;

[0153] ;

[0154] In the formula, The target voltage performance index for closed-loop control is consistent with the objective of ensuring that the PCC voltage converges to the reference value. This represents the actual reactive power output of the photovoltaic inverter at node j. This is the rated reference reactive power of the photovoltaic inverter;

[0155] Integral gain A smaller positive value is chosen to maintain system stability and suppress overshoot, thus avoiding the impact of improper parameters on the dynamic adjustment effect of the photovoltaic cluster system.

[0156] Given sensitivity The droop control gain directly characterizes the photovoltaic unit's ability to compensate for voltage fluctuations. To ensure that the compensation response strength matches the photovoltaic inverter's regulation capability, the droop control gain is defined as: ;

[0157] To address the need for a weighted bias mechanism, a generalized expression for the droop gain of the photovoltaic inverter controller is proposed based on the weighted values.

[0158] ;

[0159] in, It is an averaging factor; the voltage deviation at PCC and the voltage deviation at the local node j of the photovoltaic inverter are respectively defined as... and :

[0160] ;

[0161] ;

[0162] In the formula, and These represent the average voltage of PCC and local node j relative to the reference value within a time window of length T. and The deviation is an important quantitative result for real-time monitoring of the deviation between the PCC voltage and the voltage of each local node;

[0163] By integrating the sensitivity of the photovoltaic inverter to the PCC ( ) and its sensitivity to local node j ( We construct a weighted bias mechanism to further improve the droop gain generalization expression and ensure the rationality of parameter adjustment.

[0164] To implement the reactive power allocation strategy, each photovoltaic inverter controller is reconfigured in the following manner:

[0165] (32);

[0166] By adjusting node parameters in real time And its corresponding reactive power redistribution mechanism, to achieve distributed coordination and efficient voltage regulation among actuators, ultimately helping to achieve optimal tracking control of PCC voltage to the reference value, and achieving the control objective of the photovoltaic cluster system.

[0167] A simulation analysis was conducted on a low-voltage distribution network cluster consisting of two 400V feeders and 10 photovoltaic systems, with the topology as follows: Figure 2 As shown. The upstream distribution network is represented by the Thevenin equivalent model, with parameters as follows: , For sampling measurements, a sampling frequency of 3.2kHz was used. ), window length The control frequency is 200Hz. The reference voltage of the PCC is set to 1p.u. (400V).

[0168] In the following simulation scenarios, the proposed control method is compared with two other methods. Method 1 employs a proportional power allocation strategy, which distributes reactive power equally among all photovoltaic units. This method focuses only on maintaining the voltage at the PCC and does not prioritize photovoltaic units with large local voltage deviations. Method 2 is a local voltage independent regulation strategy, in which each photovoltaic unit independently regulates its own local voltage without considering the voltage at the PCC.

[0169] Example 2

[0170] See the attached figures. The purpose of this invention is to provide a control method for coordinating multiple inverters within a photovoltaic (PV) cluster to achieve dynamic voltage support. Based on a distributed control strategy, this method coordinates multiple PV inverters to form a dynamic voltage support cluster. This method utilizes real-time data to construct a complete control framework, eliminating the need for prior acquisition of grid system parameters. This technology integrates a controller and a power distribution mechanism, collaboratively scheduling the operation of multiple PV inverters within the cluster, and prioritizing units with larger local voltage deviations for voltage support.

[0171] A robust control system that does not require prior knowledge of system parameters or structure is proposed, which achieves dynamic voltage support at the PCC point by coordinating multiple photovoltaic inverters within the cluster.

[0172] Prioritize the regulation of photovoltaic units whose local voltage exceeds the limit, and maintain the main control target of PCC voltage while ensuring that the voltage of all photovoltaic units is stable within the safe range;

[0173] By operating in a fully distributed manner without a central controller, the solution improves efficiency and scalability by reducing computing costs.

[0174] The voltage regulation objective proposed in this invention under the photovoltaic cluster mode is:

[0175] The core control objective of the system under photovoltaic cluster control mode is described as: to control the PCC voltage. Adjust and maintain at the optimal setting value The objective is formalized using the L2 norm, and its mathematical expression is as follows:

[0176]

[0177] In the formula This represents the vector of T consecutive effective voltage values ​​measured by the PCC, i.e.: ;in The index entries constructed for the time window satisfy the timestamp recursion relationship. Here This indicates that the sampling period is controlled, while data points within the window use a finer sampling period. (satisfy Reference voltage vector Defined as: The typical value of the reference voltage is To quantify the degree of achievement of this goal, an error evaluation index based on the L2 norm is introduced:

[0178]

[0179] The index defined by formula (2) The degree of voltage deviation at the PCC relative to the reference value was quantified and used as a basis for judging the performance of the photovoltaic control system. Based on this, the first control objective was established: to achieve optimized tracking of the voltage to the ideal reference value under steady-state conditions.

[0180] Objective 1: Achieve optimal tracking control of PCC voltage to the reference value.

[0181]

[0182] This objective ensures that the PCC voltage converges to the reference value, achieving the predetermined voltage regulation operation target. In addition to maintaining the PCC voltage at the reference value, the parallel operation coordination of the photovoltaic system must synchronously regulate the local voltage at each connected bus to dynamically bring it close to its optimal operating condition value.

[0183] Objective 2: To achieve coordinated dynamic regulation of local voltage across multiple connected photovoltaic systems.

[0184]

[0185] In the formula, j represents the node number; For the corresponding Local voltage of the bus connected at time node j; The optimal reference voltage for the busbar connected at node j; This represents the allowable voltage deviation range (5%).

[0186] Objective 2 is subordinate to Objective 1 (maintaining PCC voltage), which is the primary objective, but it is also crucial for ensuring the overall stability of the power grid voltage. Figure 1 As shown.

[0187] To achieve objectives 1 and 2, the photovoltaic inverter controller needs to maintain voltage stability by regulating reactive power. This process is achieved by dynamically adjusting the reactive power injection based on the system status. The photovoltaic inverter controller plays a crucial role in voltage regulation through dynamic control of reactive power. Each inverter regulates the q-axis current in the dq rotating coordinate system through a built-in proportional-integral (PI) controller, thereby injecting or absorbing reactive power as needed. This mechanism provides rapid adaptive voltage support for the PCC and local access point, ensuring that the voltage remains within permissible operating limits. This capability enables the inverter to serve as a grid auxiliary service unit, effectively enhancing system stability and ensuring voltage compliance.

[0188] The mathematical characterization method for the dynamic characteristics of photovoltaic / cluster systems proposed in this invention is as follows:

[0189] (a) Image representation

[0190] Consider establishing an Nth-order differential system model to describe the data relationship between photovoltaic units and the cluster. This system is either in an open-loop or closed-loop structure, and its mathematical expression is:

[0191]

[0192] In the formula, the external variable vector Defined as: Input quantity This indicates the reactive power output of the photovoltaic unit. The cluster node voltage deviation index defined by equation (2); each constant matrix , The dimension is 1×2, and its parameters are obtained through actual measurement data.

[0193] Equation (5) can be simplified as follows:

[0194]

[0195] Equation (6) is called kernel representation, where As a constant matrix, this representation fully describes the measurement dataset in discrete time. The dynamic relationships between them.

[0196] For control systems, an image representation method based on the left kernel space is adopted. The null space (left kernel space) and its column space (image space) of the matrix form an orthogonal complement relationship in the corresponding vector space. Therefore, the image representation of the system is established by constructing a matrix M that satisfies PM, where matrix P characterizes the dynamic characteristics of the system.

[0197] For a single-input single-output system, its image representation is determined by a matrix. ( It is constructed from this, and its specific form of expression is as follows:

[0198]

[0199] In the formula, auxiliary variables The column space (image space) of the system represents the degrees of freedom that characterize the system's behavior. The core advantage of this representation lies in... The unconstrained property, that is, arbitrary All trajectories are generated to satisfy the system dynamics and are valid input-output trajectories. This flexibility is used in design. The stable trajectory, thus To achieve stability.

[0200] (ii) Image representation based on measurement data

[0201] The coefficient matrices Pi and Mi in equations (6) and (7) are directly calculated from the data analysis. Based on this, let the duration of the input and output measurement data trajectory of the port variable W be Th, and define the Hankel matrix of the measurement values ​​as:

[0202]

[0203] The order of the Hankel matrix is ​​taken as . N represents the maximum lag order of the system.

[0204] To ensure that the system behavior can be represented by measurement data, the input signal must satisfy the persistency of excitation (PE) condition. Let the system state space have dimension n. If the input sequence u(1), The constructed Hankel matrix If a row is full rank, then the input sequence is said to have L+n order continuous excitation characteristics.

[0205] When the PE condition is satisfied, the Hankel matrix The complete representation of all legal trajectories of the system under lag order N; this conclusion holds only if the measured trajectory length Th satisfies... This ensures that the data margin is sufficient to describe the dynamic characteristics of the system.

[0206] When the continuous excitation condition is met and the trajectory length Th meets the above requirements, the dynamic behavior characteristics of the system can be fully characterized by measurement data.

[0207] Hankel matrix The singular value decomposition (z) is an auxiliary variable in the trajectory image space of the stretched system and is applied to control design.

[0208] IV. Photovoltaic Inverter Control Method Based on Stable Trajectory

[0209] Equation (3) requires This objective is achieved by designing a tracking controller that meets specific performance indicators. The desired closed-loop characteristics are obtained by configuring the system polynomial roots to ensure stability and achieve control response. To this end, a stable trajectory Z needs to be constructed in the image representation (9), the design of which will ensure that the input-output combination vector... The trajectory stability is achieved, ultimately realizing the control objective.

[0210] Specifically, the trajectory of z is constructed as a solution to the following form of difference equation:

[0211]

[0212] In the formula For reference values, k0 and k1 are the controller gain coefficients that determine the roots of the difference equation.

[0213] To simplify the analysis, let's assume... .

[0214] To achieve negative feedback compensation, the output quantity y in equation (9) is substituted into equation (10), that is:

[0215]

[0216] To ensure stability, it is crucial to avoid pole-zero cancellation. This requires that there are no common factors among the terms on both sides of equation (11). By grouping the terms of equation (11), we obtain:

[0217]

[0218] The stability criterion requires that the roots of the characteristic equation lie inside the unit circle, and it satisfies the following condition: Where Re and Im represent the real and imaginary parts of the root, respectively.

[0219] Closed-loop control behavior is achieved by specifying the desired poles. and These poles, as defined, determine the dynamic characteristics of the system and are realized through the following characteristic equation:

[0220]

[0221] To achieve the desired time response characteristics (such as settling time) ), need to select and The parameter values. The relationship between the settling time and the location of the poles is determined by... Establish (where Ts is the sampling period).

[0222] The controller gain parameters k0 and k1 must satisfy the relationship in equation (13). The expression is derived directly: , .

[0223] This control scheme explicitly defines and locks the closed-loop trajectory during the design phase, eliminating the need for repeated adjustments to gain parameters. This trajectory pre-setting characteristic has advantages in control, ensuring the controlled variable within the discrete system is properly positioned. Strictly track the preset trajectory. The system dynamic equations are represented as:

[0224]

[0225] Input sequence The generation strategy aims to ensure controlled quantities Tracking the preset trajectory Typical implementation methods include:

[0226]

[0227] Convergence rate parameter It directly determines the system's dynamic performance. The corresponding control input quantity... Obtained through inversion calculation:

[0228]

[0229] When function In reverse time, this control law ensures that the system response has a predictive and deterministic structure, completely avoiding the gain adjustment process. In contrast, discrete-time PI controllers rely on error... The control command is generated, and its control law expression is:

[0230]

[0231] The closed-loop system characteristics shown in Equation (17) depend entirely on the tuning of the parameters Kp and Ki, a process that presents significant challenges. Improper tuning can lead to system oscillations, overshoot, or slow convergence, making stability and dynamic performance highly sensitive to parameter selection. To visually compare the differences, a first-order discrete system is used as an example:

[0232]

[0233] The open-loop stability requirement in the formula is satisfied. Under a preset trajectory control strategy, the system is guided by a given target trajectory, for example:

[0234]

[0235] Substitute the target trajectory into the system equations:

[0236]

[0237] Direct solution requires the following control inputs:

[0238]

[0239] This method enables the system to strictly track the target trajectory, not only avoiding the need for gain adjustment but also ensuring a deterministic structure of the closed-loop response. When a PI controller is used, the system control input is:

[0240]

[0241] The error signal Substituting this into the system equations yields the closed-loop dynamic characteristics:

[0242]

[0243] The stability of this system depends on the characteristic equation:

[0244]

[0245] Ensuring stability requires careful tuning of the Kp and Ki parameters, making system performance highly dependent on parameter selection. Improper tuning can lead to oscillations, slow convergence, or even instability. In stark contrast, the proposed control strategy fundamentally avoids the iterative process of gain tuning by directly enforcing a preset trajectory, ensuring the predictability of the closed-loop response.

[0246] V. Reactive Power Coordination Allocation Mechanism for Photovoltaic Clusters

[0247] The core advantage of this control scheme lies in its modularity, with each photovoltaic inverter controller independently performing computational tasks. Compared to centralized control methods, this design improves practicality and greatly simplifies the integration process for new power generation units. For photovoltaic cluster systems containing multiple sets of closed-loop execution units, controller coordination is achieved through a power allocation mechanism.

[0248] (I) Voltage-Reactive Power Sensitivity Analysis

[0249] Based on actual operational data, this study quantifies the impact mechanism of reactive power variations in photovoltaic (PV) systems on PCC voltage. This characterization method relies on a sensitivity function, which provides a first-order approximation estimate of parameter and input variable fluctuation effects. In clustered distribution network systems, the sensitivity coefficient... Used to quantify the effect of execution unit j on the voltage of node i.

[0250] Sensitivity is defined as:

[0251]

[0252] In the formula It is caused by changes in reactive power. The resulting voltage change at PCC, relative to the nominal value and .

[0253] This sensitivity was also determined by the fluctuation of the following measurement data:

[0254]

[0255]

[0256] Using the least squares method, the sensitivity approximation is:

[0257]

[0258] A node sensitivity analysis method based on sampled signals is adopted, which belongs to the standard voltage / reactive power approximation technique for power grid analysis. New measurement data is integrated through a sliding data window, and the sensitivity analysis results are dynamically updated to ensure that the sensitivity parameters reflect the latest power grid operating conditions in real time.

[0259] (ii) Distributed Coordination Control of Photovoltaic Systems

[0260] To achieve coordinated power distribution, a droop control equation is used to dynamically adjust the control reference values ​​of each photovoltaic inverter, thereby improving voltage performance indicators. Transition to An effective negative feedback mechanism is constructed. For photovoltaic node j, the control reference value update strategy is as follows:

[0261]

[0262]

[0263] In the formula The target voltage performance index for closed-loop control. This represents the actual reactive power output of the photovoltaic inverter at node j. This is the rated reference reactive power of the photovoltaic inverter.

[0264] To achieve the control objective described in equation (1), an integral term is introduced (reflected in...). (Update process). Integral gain A relatively small positive value should be selected to maintain system stability and suppress overshoot. This mechanism enables distributed coordinated control of the photovoltaic system, effectively balancing local voltage regulation and global voltage control objectives.

[0265] Given sensitivity To directly characterize the photovoltaic unit's ability to compensate for voltage fluctuations, the compensation response strength must match the photovoltaic inverter's regulation capability. By combining equations (28) and (29), this proportional negative feedback is defined by the droop control gain in equation (29) as: .

[0266] Traditional droop control typically operates on the PCC (Power Control Center). When using clustered reactive power injection to regulate the PCC voltage, the control system will adjust the local voltage... Deviation from target value The largest photovoltaic unit is prioritized for regulation.

[0267] When setting the droop gain for each photovoltaic unit, dual sensitivities must be considered: one is its sensitivity to the PCC voltage, and the other is its sensitivity to its own local voltage. Priority should be given to photovoltaic units with lower local voltages, achieving a dual benefit: ensuring the effectiveness of coordinated control while improving the overall voltage stability of the power grid.

[0268] To achieve this goal, a generalized expression for the droop gain of the photovoltaic inverter controller at node j is proposed based on weighted values:

[0269]

[0270] in It is an averaging factor. The voltage deviation at PCC and the voltage deviation at the local node j of the photovoltaic inverter are respectively defined as... and : , .

[0271] In the formula and These represent the average voltage of PCC and local node j relative to the reference value within a time window of length T. and The deviation.

[0272] This generalized model simultaneously considers both global and local voltage fluctuations, by incorporating the sensitivity of the photovoltaic inverter at node j to PCC ( ) and its sensitivity to local node j ( ), and construct a weighted bias mechanism.

[0273] To implement this power allocation strategy, each photovoltaic inverter controller is reconfigured as follows:

[0274]

[0275] By adjusting node parameters in real time Its corresponding reactive power redistribution mechanism enables distributed coordination and efficient voltage regulation among actuators.

[0276] A simulation analysis was conducted on a low-voltage distribution network cluster consisting of two 400V feeders and 10 photovoltaic systems, with the topology as follows: Figure 2 As shown. The upstream distribution network is represented by the Thevenin equivalent model, with parameters as follows: , For sampling measurements, a sampling frequency of 3.2kHz was used. ), window length The control frequency is 200Hz. The reference voltage of the PCC is set to 1p.u. (400V).

[0277] In the following simulation scenarios, the proposed control method is compared with two other methods. Method 1 employs a proportional power allocation strategy, which distributes reactive power equally among all photovoltaic units. This method focuses only on maintaining the voltage at the PCC and does not prioritize photovoltaic units with large local voltage deviations. Method 2 is a local voltage independent regulation strategy, in which each photovoltaic unit independently regulates its own local voltage without considering the voltage at the PCC.

Claims

1. A control method for coordinating multiple inverters within a photovoltaic cluster to achieve dynamic voltage support, characterized in that, include: The target is to reduce the voltage at the point of common link (PCC) of the photovoltaic cluster system. Adjust and maintain at the reference value; A mathematical characterization model of the dynamic characteristics of a photovoltaic cluster system is constructed, where Z is an auxiliary variable in the model that forms the trajectory image space of the system. Design a photovoltaic inverter controller to construct a stable trajectory Z in the image representation; Constructing a reactive power collaborative allocation mechanism for photovoltaic clusters: Conduct voltage and reactive power sensitivity analysis of photovoltaic clusters, and calculate the sensitivity of each photovoltaic inverter to PCC voltage and its local node voltage. Based on the above sensitivity analysis results, a generalized expression for the droop gain of each photovoltaic inverter controller is proposed; a weighted bias mechanism is constructed by integrating the sensitivity of the PCC voltage of each photovoltaic inverter with the local node voltage sensitivity. By monitoring the deviation between the PCC voltage and the voltages of each local node in real time, the droop gain parameters of each photovoltaic inverter controller are dynamically adjusted. Based on the adjusted droop gain parameters, the corresponding reactive power redistribution mechanism is triggered to rationally allocate reactive power regulation tasks to each photovoltaic inverter, achieving coordinated dynamic regulation of the local voltages of multiple connected photovoltaic systems. This enables optimal tracking control of the PCC voltage to the reference value. (3)， Ensure that the PCC voltage converges to the reference value.

2. The control method for coordinating multiple inverters within a photovoltaic cluster to achieve dynamic voltage support according to claim 1, characterized in that, In the design of photovoltaic inverter controllers, each photovoltaic inverter adjusts the q-axis current in the dq rotating coordinate system through a built-in proportional-integral controller, thereby injecting or absorbing reactive power as needed. This provides a foundation for the subsequent construction of a reactive power collaborative allocation mechanism for photovoltaic clusters and helps to achieve coordinated dynamic adjustment of local voltages of multiple connected photovoltaic systems.

3. The control method for coordinating multiple inverters within a photovoltaic cluster to achieve dynamic voltage support according to claim 1, characterized in that, The formula for the coordinated dynamic adjustment of the local voltage of the multi-access photovoltaic system is as follows: (4)， In the formula, j represents the bus number connected to the photovoltaic unit; For the corresponding The local voltage of bus j at time j; The optimal reference voltage for bus j; It represents the allowable deviation range of the voltage; it achieves optimal tracking control of the PCC voltage to the reference value, and ultimately achieves the control objective.

4. The control method for coordinating multiple inverters within a photovoltaic cluster to achieve dynamic voltage support according to claim 1, characterized in that, The specific steps for constructing the mathematical representation model of the dynamic characteristics of the photovoltaic cluster system are as follows: For the port variable w, select a duration of The input and output measurement data trajectories are used to construct a Hankel matrix from the measured values, as defined below: （8）， The order of the Hankel matrix is ​​set to . N represents the maximum lag order of the system; To ensure that the system behavior can be accurately represented by the measurement data, the input signal must meet the condition of continuous excitation; Let the system state space have dimension n. If the input sequence is given by... The constructed Hankel matrix If the rows are full rank, then the input sequence is said to have rank. Continuous excitation characteristics; When the PE condition is satisfied, the Hankel matrix It can completely represent all legal trajectories of the system at lag order N; this conclusion holds true only if the trajectory length is measured. Must meet This ensures that the data margin is sufficient to accurately describe the dynamic characteristics of the system, laying the foundation for the subsequent design of the photovoltaic inverter controller and the construction of the stable trajectory Z; When the continuous excitation condition is satisfied and the trajectory length is When the above requirements are met, the dynamic behavior characteristics of the system can be fully characterized based on the available measurement data; Hankel matrix The Singular Value Decomposition (SVD) expression is: ,in and It is an orthogonal matrix. It includes A diagonal matrix of non-negative singular values; Hankel matrix The rank of the singular value matrix is ​​denoted by r, and its value is equal to that of the singular value matrix. The number of non-zero singular values; due to exist There is There are zero rows, therefore U contains vectors with minified w; partition U into ,in Composed of the first r columns, the Hankel matrix can be derived. The explicit expression for the left kernel space is: ;for ,parameter and It can be determined; These parameters can be expressed in the image representation as: ； In the formula, z is an auxiliary variable in the image space of the system trajectory, which is used in the design of photovoltaic inverter controllers to achieve the goal of constructing a stable trajectory Z.

5. The control method for coordinating multiple inverters within a photovoltaic cluster to achieve dynamic voltage support according to claim 1, characterized in that, When designing a photovoltaic inverter controller, the trajectory Z can be constructed as a solution to the following difference equation: （12）； The stability criterion for this controller requires that the roots of its characteristic equation lie inside the unit circle, satisfying the following condition: Where Re and Im represent the real and imaginary parts of the root, respectively; Closed-loop control behavior can be specified by specifying the desired poles. and These poles are defined to determine the dynamic characteristics of the system, and are realized through the following characteristic equation: ； To achieve the desired time response characteristics (such as settling time) ), need to select and The parameter values; the relationship between the settling time and the location of the poles is determined by... Establish (among others) (where the sampling period is the reference period), this relationship provides a clear basis for parameter selection, ensuring optimal tracking control of the PCC voltage to the reference value in the subsequent implementation; Controller gain parameters and The design must satisfy the relationship in equation (13); the expression can be obtained by direct derivation: , ; The photovoltaic inverter controller defines and locks the closed-loop trajectory, and the system dynamic equation is represented as follows: ； Input sequence The generation strategy aims to ensure controlled quantities Precisely track the preset trajectory The typical implementation is as follows: ； Convergence rate parameter Directly determines the dynamic performance of the system; corresponding control input quantities Obtained through inversion calculation: ； When function When reversible, this control law ensures that the system response has a predictable and deterministic structure, completely avoiding the gain adjustment process, and providing a reliable guarantee for the stable operation of the photovoltaic cluster system and the achievement of control objectives.

6. The control method for coordinating multiple inverters within a photovoltaic cluster to achieve dynamic voltage support according to claim 1, characterized in that, When conducting voltage and reactive power sensitivity analysis of photovoltaic clusters, a sensitivity coefficient is introduced considering the characteristics of clustered distribution network systems. This is used to quantify the impact of execution unit j on the voltage of node i, providing key parameter support for the subsequent construction of a photovoltaic cluster reactive power collaborative allocation mechanism and the proposal of a droop gain generalization expression. Sensitivity is defined as: ； In the formula, It is caused by changes in reactive power. The resulting change in node i voltage and These are the nominal values ​​of the voltage at node i and the reactive power of execution unit j, respectively; when node i is a PCC, This refers to the voltage change at the PCC, which can be used to analyze the sensitivity of each photovoltaic inverter to the PCC voltage. This sensitivity can also be determined by the fluctuation of the following measurement data: ； ； Using the least squares method, the sensitivity can be approximately calculated as follows: ； The sensitivity calculated in the above way can accurately reflect the correlation between the execution unit and the node voltage, laying a data foundation for subsequent dynamic adjustment of the droop gain parameters of each photovoltaic inverter and achieving reasonable allocation of reactive power.

7. The control method for coordinating multiple inverters within a photovoltaic cluster to achieve dynamic voltage support according to claim 1, characterized in that, The coordinated dynamic adjustment of the local voltage of the multi-access photovoltaic system is implemented as follows: To achieve coordinated power distribution, a droop control equation is used to dynamically adjust the control reference values ​​of each photovoltaic inverter, thereby improving voltage performance indicators. Transition to To construct an effective negative feedback mechanism and ensure the stability and accuracy of voltage regulation; For photovoltaic node j, the control reference value update strategy is as follows: ； ； In the formula, The target voltage performance index for closed-loop control is consistent with the objective of ensuring that the PCC voltage converges to the reference value. This represents the actual reactive power output of the photovoltaic inverter at node j. This is the rated reference reactive power of the photovoltaic inverter; Integral gain A smaller positive value is chosen to maintain system stability and suppress overshoot, thus avoiding the impact of improper parameters on the dynamic adjustment effect of the photovoltaic cluster system. Given sensitivity The droop control gain directly characterizes the photovoltaic unit's ability to compensate for voltage fluctuations. To ensure that the compensation response strength matches the photovoltaic inverter's regulation capability, the droop control gain is defined as: ; To address the need for a weighted bias mechanism, a generalized expression for the droop gain of the photovoltaic inverter controller is proposed based on the weighted values. ； in, It is an averaging factor; the voltage deviation at PCC and the voltage deviation at the local node j of the photovoltaic inverter are respectively defined as... and : ； ； In the formula, and These represent the average voltage of PCC and local node j relative to the reference value within a time window of length T. and The deviation is an important quantitative result for real-time monitoring of the deviation between the PCC voltage and the voltage of each local node; By integrating the sensitivity of the photovoltaic inverter to the PCC ( ) and its sensitivity to local node j ( We construct a weighted bias mechanism to further improve the droop gain generalization expression and ensure the rationality of parameter adjustment. To implement the reactive power allocation strategy, each photovoltaic inverter controller is reconfigured in the following manner: (32)； By adjusting node parameters in real time And its corresponding reactive power redistribution mechanism, to achieve distributed coordination and efficient voltage regulation among actuators, ultimately helping to achieve optimal tracking control of PCC voltage to the reference value, and achieving the control objective of the photovoltaic cluster system.

Images