Distributed photovoltaic grid-connected flexible voltage regulation method and system

Abstract

The present application relates to the technical field of low-voltage distributed photovoltaic regulation, and particularly relates to a distributed photovoltaic grid-connected flexible voltage regulation method and system, each photovoltaic grid-connected unit monitors local voltage and output to obtain an initial virtual responsibility coefficient; exchanges state information with neighborhood units periodically to calculate neighborhood pressure factors; combines local regulation sensitivity to iteratively adjust the virtual responsibility coefficient until a collaborative consensus is reached; determines a power regulation amount according to the steady-state responsibility coefficient to control the inverter to perform voltage regulation; and gradually reduces the responsibility coefficient to exit the regulation when the voltage is continuously restored. The present application realizes a centerless voltage regulation decision through a completely distributed collaborative mechanism, and improves the expansibility and regulation accuracy of the system.

Description

Technical Field

[0001] This invention relates to the field of low-voltage distributed photovoltaic (PV) regulation technology, and in particular to a flexible voltage regulation method and system for grid-connected distributed PV. Background Technology

[0002] With the large-scale integration of low-voltage distributed photovoltaic (PV) power into the distribution network, the intermittency and randomness of its output pose significant challenges to the power quality of the distribution area, especially voltage stability. To address this challenge, the industry has developed a typical flexible control scheme based on the technical framework of "observable, measurable, controllable, and adjustable." This scheme mainly constructs a comprehensive operation monitoring system by deploying distributed power source access units, smart IoT energy meters, and PV-specific circuit breakers as underlying sensing and control devices. At the "observable" level, micro-applications are deployed to achieve centralized visualization of PV installation data, real-time operating status, execution of control commands, and equipment anomaly information. At the "measurable" level, relying on minute-level or higher data acquisition capabilities, real-time sensing and operational monitoring of massive distributed PV power generation output are achieved, providing a data foundation for analysis and decision-making. At the control level, a "rigid-flexible" system has been formed: on the one hand, "controllable" capabilities such as remote and rapid switching are provided through photovoltaic-specific circuit breakers, serving as a rigid backup means to ensure grid security; on the other hand, distributed power source access units or smart IoT energy meters with remote communication and adjustment functions are used to receive instructions from the upper-level master station or local controller to limit and adjust the output power of the photovoltaic inverter, thereby forming an "adjustable" flexible adjustment capability. Currently, this technical system has become the mainstream practical path for achieving effective management of low-voltage distributed photovoltaic power.

[0003] While existing solutions have established basic regulation capabilities, several specific technical challenges remain in practical operation, particularly in addressing voltage exceedance issues caused by photovoltaic backfeeding. Existing flexible regulation largely relies on centralized or semi-centralized control modes, where the master station system or local centralized controller calculates and issues unified or node-specific regulation commands based on global monitoring data. This mode places high demands on the reliability and real-time performance of the communication network, leading to problems such as high data aggregation pressure, control command delays, and heavy central computing burden in scenarios with massive node access, thus limiting system scalability. More importantly, centralized decision-making often fails to fully and in real-time consider the local characteristics and state differences of each node in the grid, potentially resulting in a "one-size-fits-all" approach to regulation commands, failing to achieve precise and optimized allocation of regulation resources in time and space. For example, some nodes may over-regulate, impacting power generation efficiency, while others may under-regulate, failing to effectively suppress voltage exceedances. Furthermore, existing methods lack effective autonomous coordination mechanisms among photovoltaic units during regulation, making it impossible to dynamically adjust their responsibilities based on local neighborhood information, hindering the formation of rapid, smooth, and adaptive global voltage coordination control effects. Furthermore, how to enable the system to automatically and smoothly exit the regulation state after voltage recovery, maximizing the self-consumption benefits of photovoltaic power, is also a problem that needs further optimization in existing solutions. Therefore, there is an urgent need for a more distributed, intelligent, and collaborative flexible voltage regulation method to improve control efficiency, economy, and adaptability.

[0004] Chinese Patent Publication No. CN119231553A discloses a distributed photovoltaic (PV) reactive power voltage regulation method and system, including fitting the voltage / reactive power sensitivity relationship matrix of each distributed PV node; calculating the voltage / reactive power droop curve of each distributed PV node and the reactive power that needs to support other nodes; calculating the expected reactive power output of the PV inverter based on the voltage / reactive power droop curve and the reactive power that needs to support other nodes; and when the expected reactive power output of each PV inverter is lower than or equal to the upper limit of reactive power, and the expected reactive power of each PV inverter is... If the output is higher than or equal to the lower limit of reactive power, the expected reactive power is used as the reactive power command value of the photovoltaic inverter. When the expected reactive power output of each photovoltaic inverter is lower than the lower limit of reactive power or higher than the upper limit of reactive power, the reactive power output of the photovoltaic inverter is adjusted to the lower limit of reactive power or the upper limit of reactive power. The reactive power deficit caused by capacity limitation is calculated and used as the input information for the next cycle. It can be seen that the distributed photovoltaic reactive power regulation method and system have the following problems: it relies on centralized control for overall regulation, the regulation accuracy is insufficient, and the exit process is prone to grid impact. Summary of the Invention

[0005] To address these issues, this invention provides a distributed photovoltaic grid-connected flexible voltage regulation method and system to overcome the problems of strong reliance on centralized control, insufficient regulation accuracy, and easy grid impact during the shutdown process in existing technologies.

[0006] To achieve the above objectives, the present invention provides a distributed photovoltaic grid-connected flexible voltage regulation method and system, comprising:

[0007] Step S1: Each photovoltaic grid-connected unit monitors the voltage deviation rate and its own output percentage at the local grid connection point in real time to obtain the initial virtual responsibility coefficient of each local grid connection point.

[0008] Step S2: Each photovoltaic grid-connected unit periodically broadcasts a coordination message, including its current voltage deviation rate and the current initial virtual responsibility coefficient, to other photovoltaic grid-connected units in the neighborhood, and receives coordination messages from other photovoltaic grid-connected units in the neighborhood.

[0009] Step S3: Each photovoltaic grid-connected unit calculates a neighborhood pressure factor based on all received neighborhood coordination messages. The neighborhood pressure factor is a weighted average of the voltage deviation rates of other photovoltaic grid-connected units in the neighborhood, with their respective initial virtual responsibility coefficients as weights.

[0010] Step S4: Each photovoltaic grid-connected unit obtains the pre-stored local regulation sensitivity coefficient, and adjusts the initial virtual responsibility coefficient based on the neighborhood pressure factor and the local regulation sensitivity coefficient to obtain the virtual responsibility coefficient;

[0011] Step S5: Repeat steps S2 to S4. Based on the comparison results of each photovoltaic grid-connected unit determining that its virtual responsibility coefficient is less than or equal to the preset responsibility coefficient convergence threshold, and the rate of change of voltage deviation rate of key photovoltaic grid-connected units in the neighborhood is less than or equal to the preset voltage deviation rate convergence threshold, a consensus is reached and each photovoltaic grid-connected unit enters parameter steady state.

[0012] Step S6: After the photovoltaic grid-connected unit enters the parameter steady state, the corresponding power regulation amount is determined according to the virtual responsibility coefficient at the parameter steady state through a preset mapping relationship, and the connected photovoltaic inverter is controlled to adjust based on the power regulation amount to complete the coordinated voltage regulation.

[0013] Step S7: When each photovoltaic grid-connected unit detects that the local voltage deviation rate is continuously lower than the recovery threshold, the virtual responsibility coefficient is adjusted in a negative iterative manner until the photovoltaic inverter recovers to normal output.

[0014] Further, step S1 includes:

[0015] Step S11: Measure the current voltage at the local grid connection point in real time, and calculate the voltage deviation rate based on the current voltage and the rated voltage;

[0016] Step S12: Obtain the current output power of the connected photovoltaic inverter in real time, and calculate the output percentage based on the current output power and the rated capacity of the photovoltaic inverter;

[0017] Step S13: Determine the initial virtual responsibility coefficient based on the voltage deviation rate and the output percentage.

[0018] Furthermore, in step S13, the first component of the initial virtual responsibility coefficient is positively correlated with the voltage deviation rate; the second component of the initial virtual responsibility coefficient is positively correlated with the output percentage; and the initial virtual responsibility coefficient is the sum of the first component and the second component.

[0019] Further, step S3 includes:

[0020] Step S31: Extract the voltage deviation rate and the initial virtual responsibility coefficient of each sending photovoltaic grid-connected unit from all received neighborhood cooperative messages;

[0021] Step S32: Determine the weight of the photovoltaic grid-connected unit in the weighted calculation based on the initial virtual responsibility coefficient of each transmitting photovoltaic grid-connected unit;

[0022] Step S33: Based on the voltage deviation rate of each transmitting photovoltaic grid-connected unit and its corresponding weight, calculate the weighted average value to obtain the neighborhood pressure factor.

[0023] Further, in step S32, the weight of the photovoltaic grid-connected unit in the weighted calculation is the ratio of the initial virtual responsibility coefficient of a single transmitting photovoltaic grid-connected unit to the sum of the initial virtual responsibility coefficients of all transmitting photovoltaic grid-connected units.

[0024] Further, step S5 includes:

[0025] Step S51: After each completion of step S4, calculate the instantaneous rate of change of the current virtual responsibility coefficient compared to the previous calculation result;

[0026] Step S52: Traverse all received neighborhood coordination messages, and by comparing the voltage deviation rate in each neighborhood coordination message, obtain at least one photovoltaic grid-connected unit with the largest voltage deviation rate, which is then used as the key photovoltaic grid-connected unit.

[0027] Step S53: Monitor the voltage deviation rate of the key photovoltaic grid-connected unit and calculate its instantaneous change rate;

[0028] Step S54: When the instantaneous rate of change of the current virtual responsibility coefficient is lower than the first convergence threshold and the instantaneous rate of change of the voltage deviation rate of the key photovoltaic grid-connected unit is lower than the second convergence threshold, it is determined that the system has reached a consensus and the current photovoltaic grid-connected unit enters the parameter steady state.

[0029] Step S55: If the determination condition of step S54 is not met, return to step S2 to perform message broadcasting, pressure factor calculation and virtual responsibility coefficient adjustment for the next cycle.

[0030] Further, S7 includes:

[0031] Step S71: When the photovoltaic grid-connected unit detects that the local voltage deviation rate is continuously less than the recovery threshold, the first adjustment amount is calculated based on the current virtual responsibility coefficient and the current local voltage deviation rate.

[0032] Step S72: Reduce the current virtual responsibility coefficient according to the first adjustment amount to obtain the updated virtual responsibility coefficient;

[0033] Step S73: Based on the updated virtual responsibility coefficient, determine the corresponding updated power regulation command based on the preset mapping relationship, and control the connected photovoltaic inverter to execute the updated power regulation command;

[0034] Step S74: After executing the updated power adjustment command, monitor the local voltage deviation rate again. If it is still lower than the recovery threshold, repeat steps S71 to S73 to perform the next round of adjustment.

[0035] Step S75: Repeat the above adjustment until the virtual responsibility coefficient is reduced to the virtual responsibility coefficient threshold, and the photovoltaic inverter returns to normal output state.

[0036] Furthermore, the feature is that the other photovoltaic grid-connected units within the neighborhood are all other photovoltaic grid-connected units that are connected to the same transformer area as the current photovoltaic grid-connected unit.

[0037] Furthermore, in step S4, the adjustment range of the virtual responsibility coefficient is positively correlated with the product of the neighborhood pressure factor and the local adjustment sensitivity coefficient, and negatively correlated with the virtual responsibility coefficient before adjustment.

[0038] On the other hand, the present invention also provides a processing system, comprising:

[0039] The status monitoring unit is deployed in each photovoltaic grid-connected unit to measure the current voltage of the local grid connection point in real time and calculate the voltage deviation rate, as well as to obtain the current output power of the connected photovoltaic inverter in real time and calculate the output percentage.

[0040] A responsibility coefficient determination unit, which is connected to the status monitoring unit, is used to determine the initial virtual responsibility coefficient based on the voltage deviation rate and the output percentage;

[0041] The communication unit, which is deployed in each photovoltaic grid-connected unit, is used to periodically broadcast a coordination message including its current voltage deviation rate and current virtual responsibility coefficient to other photovoltaic grid-connected units in the neighborhood, and to receive coordination messages from other photovoltaic grid-connected units in the neighborhood.

[0042] A pressure factor calculation unit, which is connected to the communication unit, is used to calculate the neighborhood pressure factor based on all received neighborhood cooperation messages.

[0043] The responsibility coefficient adjustment unit is connected to the pressure factor calculation unit, the responsibility coefficient determination unit, and the pre-stored local adjustment sensitivity coefficient, respectively, and is used to dynamically adjust the virtual responsibility coefficient based on the neighborhood pressure factor and the local adjustment sensitivity coefficient.

[0044] A consensus determination unit, which is connected to the responsibility coefficient adjustment unit and the communication unit, is used to determine whether a collaborative consensus has been reached and whether the parameters have entered a steady state based on the instantaneous change rate of the adjusted virtual responsibility coefficient and the instantaneous change rate of the voltage deviation rate of the key photovoltaic grid-connected unit determined from the received message.

[0045] A power regulation unit, which is connected to the consensus determination unit and the photovoltaic inverter respectively, is used to determine the corresponding power regulation amount according to the virtual responsibility coefficient of the steady state through a preset mapping relationship after entering the parameter steady state, and control the photovoltaic inverter to adjust based on the power regulation amount;

[0046] The recovery control unit is connected to the status monitoring unit, the responsibility coefficient adjustment unit, and the power regulation unit respectively. When the local voltage deviation rate is detected to be continuously lower than the recovery threshold, it initiates a negative iterative adjustment of the virtual responsibility coefficient and updates the power regulation amount synchronously until the photovoltaic inverter recovers to normal output state.

[0047] Compared with the prior art, the beneficial effects of the present invention are that by constructing a fully distributed autonomous and coordinated voltage regulation mechanism, each photovoltaic grid-connected unit can spontaneously form a globally consistent consensus on voltage regulation responsibility allocation through multiple rounds of iterations, relying only on local measurements and limited neighborhood information interaction, and execute precise power regulation accordingly. This fundamentally overcomes the dependence of the traditional centralized control mode on the central master station and high-reliability communication, and effectively improves the overall scalability and adaptability of the system when facing massive node access, communication anomalies and topology changes.

[0048] Furthermore, this invention introduces a virtual responsibility coefficient determined by fusing local voltage deviation with its own power output status, and uses it as the core medium for neighborhood information interaction and collaborative computing. This enables the allocation of regulation responsibility to simultaneously reflect the severity of the node's problems and its regulation potential, laying a quantitative foundation for the optimization and fair allocation of regulation resources in time and space, and avoiding the loss of power generation benefits or insufficient regulation caused by one-size-fits-all regulation.

[0049] Furthermore, by designing a pressure factor calculation rule with the responsibility coefficient of neighboring units as the weight, this invention enables each unit to perceive the local power grid situation more intelligently, and to focus on responding to the group pressure exerted by neighboring nodes that they consider to have a heavy responsibility and serious voltage overruns. This guides the collaborative process to quickly focus on key problem areas, accelerates consensus convergence, and improves the overall efficiency and targeting of collaborative voltage regulation.

[0050] Furthermore, by employing a dual convergence criterion that includes the stability of the local responsibility coefficient and the stability of the voltage state of critical neighbors, this invention ensures that the system only enters the steady-state execution stage when the internal responsibility allocation reaches a consensus and the most severe external voltage problem has been stably suppressed. This guarantees that the collaborative result is not only a mathematically balanced solution, but also an effective and reliable control scheme for actual voltage problems.

[0051] Furthermore, by establishing a power regulation mapping relationship based on the virtual responsibility coefficient under steady-state parameters, this invention enables each unit to independently and in parallel convert the negotiation results into specific inverter control commands after reaching a consensus. This achieves seamless connection from collaborative decision-making to execution control, and the control accuracy matches the responsibility allocation, ensuring the consistency between the voltage regulation effect and the theoretical design.

[0052] Furthermore, this invention achieves a gradual and smooth exit mechanism by performing negative iterative adjustments to the virtual responsibility coefficient based on the local voltage deviation and simultaneously relaxing power limits after the voltage problem is alleviated. This mechanism can intelligently control the exit pace according to the responsibility of each node and the degree of voltage recovery, effectively avoiding secondary voltage shocks that may be caused by sudden power recovery, and maximizing photovoltaic power generation revenue while ensuring voltage stability. Attached Figure Description

[0053] Figure 1 This is a flowchart of the distributed photovoltaic grid-connected flexible voltage regulation method of the present invention;

[0054] Figure 2 This is a flowchart of step S1 of the distributed photovoltaic grid-connected flexible voltage regulation method of the present invention;

[0055] Figure 3 This is a flowchart of step S3 of the distributed photovoltaic grid-connected flexible voltage regulation method of the present invention;

[0056] Figure 4 This is a flowchart of step S5 of the distributed photovoltaic grid-connected flexible voltage regulation method of the present invention. Detailed Implementation

[0057] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0058] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0059] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate the direction or positional relationship, are based on the direction or positional relationship shown in the drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0060] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0061] Please see Figure 1 As shown, it is a flowchart of the distributed photovoltaic grid-connected flexible voltage regulation method of the present invention;

[0062] This invention provides a distributed photovoltaic grid-connected flexible voltage regulation method, comprising:

[0063] Step S1: Each photovoltaic grid-connected unit monitors the voltage deviation rate and its own output percentage at the local grid connection point in real time to obtain the initial virtual responsibility coefficient of each local grid connection point.

[0064] Please continue reading. Figure 2 The diagram shows a flowchart of step S1 of the distributed photovoltaic grid-connected flexible voltage regulation method of the present invention. Specifically, step S1 includes:

[0065] Step S11: Measure the current voltage at the local grid connection point in real time, and calculate the voltage deviation rate based on the current voltage and the rated voltage;

[0066] In one specific embodiment, the photovoltaic grid-connected unit acquires the instantaneous voltage value U of the local grid connection point in real time through its built-in or external voltage measurement module. i Voltage deviation rate ΔU i The specific calculation formula is as follows:

[0067] ΔU i =(U i -U N ) / U N ×100%;

[0068] Wherein, ΔU i U represents the voltage deviation rate of the i-th photovoltaic grid-connected unit, expressed as a percentage (%). i The current voltage measured in real time by this unit, in volts (V); U N The rated voltage specified for the power distribution network, in volts (V).

[0069] Understandably, the above formula is a commonly used calculation method in power systems to characterize the degree of voltage deviation from the standard value. Voltage measurement modules belong to mature industrial electronic technology, and their selection must meet the measurement range, accuracy (usually requiring no less than 0.5 class), and electrical isolation safety standards of the voltage level of the power grid. Preferably, to filter out transient interference and obtain a stable operating condition characterization, the moving average of multiple continuously collected voltage cycle data can be performed before being used in the calculation.

[0070] Step S12: Obtain the current output power of the connected photovoltaic inverter in real time, and calculate the output percentage based on the current output power and the rated capacity of the photovoltaic inverter;

[0071] In one specific embodiment, the photovoltaic grid-connected unit obtains its real-time output active power value from the photovoltaic inverter it is connected to via a standard communication interface (such as RS485, Ethernet, or wireless module) and protocol (such as Modbus TCP / RTU, DL / T645, or SunSpec protocol). The formula for calculating the output percentage is as follows:

[0072] P peri =P curi / P ratedi ×100%;

[0073] Among them, P peri P represents the power output percentage of the i-th grid-connected photovoltaic unit, expressed as a percentage (%). curi P represents the current output active power of the photovoltaic inverter connected to this unit, in kilowatts (kW). ratedi This refers to the rated (nominal) capacity of the photovoltaic inverter, expressed in kilowatts (kW).

[0074] Understandably, obtaining real-time power from the inverter via communication protocols is a common practice in the photovoltaic monitoring field. The output percentage calculated by the formula represents the actual utilization rate of the inverter under current environmental conditions. ratedi These are fixed parameters on the inverter's nameplate. This step does not require complex calculations; the key lies in the reliability of the communication link and the setting of the data update cycle (usually on the order of seconds).

[0075] Step S13: Determine the initial virtual responsibility coefficient based on the voltage deviation rate and the output percentage.

[0076] Specifically, in step S13, the first component of the initial virtual responsibility coefficient is positively correlated with the voltage deviation rate; the second component of the initial virtual responsibility coefficient is positively correlated with the output percentage; and the initial virtual responsibility coefficient is the sum of the first component and the second component.

[0077] In a specific embodiment, the formula for calculating the initial virtual liability coefficient is as follows:

[0078] K i =α×(ΔU i / ΔU max )+β×(P peri / 100%);

[0079] Among them, K i Let ΔU be the initial virtual responsibility coefficient for the i-th photovoltaic grid-connected unit, a dimensionless value; i The voltage deviation rate calculated in step S11; ΔU max The maximum reference value for voltage deviation rate preset by the system, expressed as a percentage (%), is used to normalize the voltage deviation component; P peri The output percentage is calculated in step S12; α is the voltage deviation weighting coefficient, which is dimensionless; β is the output percentage weighting coefficient, which is dimensionless.

[0080] The voltage deviation weighting coefficient α and the output percentage weighting coefficient β are used to balance the relative importance of the two factors in the initial responsibility assessment. Their values ​​range from 0 to 1, and they satisfy α + β = 1 to ensure that the two components are comparable in magnitude. Preferably, α is 0.6 and β is 0.4 to emphasize the urgency of the current voltage deviation while also considering the unit's adjustment potential. The maximum reference value for the voltage deviation rate is ΔU. max According to the allowable deviation upper limit calibration in the distribution network voltage operation regulations, the preferred value is ΔU. max Take 10%.

[0081] Understandably, the initial virtual responsibility coefficient serves as the starting point for subsequent distributed collaborative iterations. Its design principle lies in comprehensively reflecting two key dimensions: one is the severity of the problem, namely the local voltage deviation rate (ΔU). i The more the voltage is limited, the more severe the problem becomes (ΔU). i The larger the node (in terms of its size), the greater its contribution to or the deeper its impact on the global voltage problem; therefore, it should be given a higher priority in the initial responsibility allocation. The second dimension is the potential dimension of resources, i.e., its own percentage output (P...). peri Units with higher output have a larger adjustable power margin and stronger potential regulation capabilities, which should be reflected in the initial responsibility. The two are weighted and summed using weighting coefficients and then normalized to generate a dimensionless initial responsibility value with a relatively controllable range. This value is not the final regulation command, but rather an "initial proposal" to initiate subsequent inter-neighbor information interaction and negotiation. It aims to allow the system to evolve from a reasonable state that simultaneously reflects local voltage problems and its own regulation capabilities, thereby accelerating the convergence of the subsequent consensus process and improving the technical and economic rationality of the final regulation scheme.

[0082] Step S2: Each photovoltaic grid-connected unit periodically broadcasts a coordination message, including its current voltage deviation rate and the current initial virtual responsibility coefficient, to other photovoltaic grid-connected units in the neighborhood, and receives coordination messages from other photovoltaic grid-connected units in the neighborhood.

[0083] Specifically, the other photovoltaic grid-connected units in the neighborhood are all other photovoltaic grid-connected units that are connected to the same transformer area as the current photovoltaic grid-connected unit.

[0084] In one specific embodiment, the photovoltaic grid-connected unit, through its integrated communication module, broadcasts and receives a coordination message once within each control cycle T. The coordination message is a structured data frame, including at least the following three data fields: the unique identifier (ID) of the photovoltaic grid-connected unit sending the message, and the voltage deviation rate ΔU currently calculated by the unit. iand the current virtual responsibility coefficient of the unit (in the first cycle, this value is the initial virtual responsibility coefficient K calculated in step S1). i The "neighborhood" of communication is physically defined as all other photovoltaic grid-connected units connected to the same transformer substation as the current photovoltaic grid-connected unit. To achieve broadcasting within this neighborhood, mature industrial communication technologies can be employed, such as power line communication (PLC) technology or a local wireless LAN (e.g., based on Wi-Fi or Zigbee protocols) built under the transformer substation. At the beginning of each cycle, each unit packages its aforementioned status data into a fixed-format message frame and broadcasts it to all other units in the neighborhood via the selected communication network. Simultaneously, the unit continuously listens to the communication channel during the cycle, receiving and parsing similar message frames broadcast from other units in the neighborhood.

[0085] Understandably, the communication period T is a crucial time parameter for the system, determining the frequency of state information synchronization and collaborative computation updates. A period that is too short can lead to excessive communication overhead and potentially unnecessary consumption of computing resources; a period that is too long will affect the system's response speed to voltage changes. Considering the relatively slow voltage changes in low-voltage distribution networks and communication reliability, the period T typically ranges from 1 to 30 seconds. Preferably, T is set to 5 seconds to achieve a balance between response timeliness and system overhead. The unique identifier is used to distinguish the message source at the receiving end and can typically be the device's factory MAC address or a logical ID assigned by the operation and maintenance system.

[0086] In one specific embodiment, after receiving a cooperative message from a neighboring unit, the photovoltaic grid-connected unit parses and stores it. Specifically, the unit extracts three key data points from each received message frame: the sender identifier, the sender voltage deviation rate, and the sender virtual responsibility coefficient, and caches them in a local neighborhood status table. This table is indexed by the sender ID and dynamically updated with its latest status data. During each control cycle, the table is refreshed, retaining only the neighbor information successfully received in the current cycle for use in subsequent step S3 calculations; where the identifier of each sending photovoltaic grid-connected unit is denoted as IDj, the voltage deviation rate of each sending photovoltaic grid-connected unit is ΔUj, and the virtual responsibility coefficient of each sending photovoltaic grid-connected unit is denoted as Kj.

[0087] Understandably, by having each grid-connected photovoltaic unit periodically broadcast its core local status (voltage deviation and responsibility awareness) to physically connected and electrically coupled neighboring units (like a transformer substation), each unit can obtain a "neighborhood situation map" of its local power grid environment without relying on a central master station to collect global information. This periodic, one-to-many broadcast communication mode is simpler and more reliable within the local network than continuous point-to-point communication or star communication relying on a central node, and naturally achieves information dissemination. Receiving and storing these messages allows each unit not only to know its own status but also to perceive the status and "attitude" (responsibility coefficient) of its neighbors, thus providing the necessary input data for the next step of collective decision-making based on local information. This mechanism replaces the cumbersome two-way communication between the master station and all terminals in traditional centralized control, distributing the task of perceiving and processing global information to each local unit in a distributed manner.

[0088] Step S3: Each photovoltaic grid-connected unit calculates a neighborhood pressure factor based on all received neighborhood coordination messages. The neighborhood pressure factor is a weighted average of the voltage deviation rates of other photovoltaic grid-connected units in the neighborhood, with their respective initial virtual responsibility coefficients as weights.

[0089] Please continue reading. Figure 3 The diagram shows a flowchart of step S3 in the distributed photovoltaic grid-connected flexible voltage regulation method of the present invention. Specifically, step S3 includes:

[0090] Step S31: Extract the voltage deviation rate and the initial virtual responsibility coefficient of each sending photovoltaic grid-connected unit from all received neighborhood cooperative messages;

[0091] In a specific embodiment, the photovoltaic grid-connected unit sequentially reads each record from its local data set, which consists of the data received and stored in step S2 during the current control cycle. Each record corresponds to another photovoltaic grid-connected unit that sent a coordination message to it (i.e., "other photovoltaic grid-connected units in the neighborhood," whose physical range has been defined as described in step S2). For each record, the photovoltaic grid-connected unit reads two key state quantities contained therein: one is the voltage deviation rate ΔU of its local grid connection point reported by the sending unit. j The parameter is a percentage value; the second is the virtual responsibility factor K currently used by the sending unit. j This parameter is a dimensionless value. During the first complete "monitoring-communication-computation" cycle after the distributed collaborative voltage regulation method begins execution, the virtual responsibility coefficient broadcast by each unit is the initial virtual responsibility coefficient K determined in step S1. jIn subsequent periodic executions, this coefficient is the virtual responsibility coefficient obtained after adjustment in step S4 and used for the current periodic broadcast.

[0092] Understandably, this step involves data reading and preparation, relying on the neighborhood state table constructed in step S2. The extraction operation is performed once per control cycle T, ensuring that the pressure factor used for calculation is based on the latest neighborhood state information. If a neighbor's message is missing in the current cycle due to communication issues, that neighbor's entry may not exist in the current cycle's state table, and its state will not participate in this calculation. This demonstrates the distributed algorithm's fault tolerance for momentary packet loss during communication.

[0093] Step S32: Determine the weight of each photovoltaic grid-connected unit in the weighted calculation based on the initial virtual responsibility coefficient of each transmitting unit;

[0094] Specifically, in step S32, the weight of the photovoltaic grid-connected unit in the weighted calculation is the ratio of the initial virtual responsibility coefficient of a single transmitting photovoltaic grid-connected unit to the sum of the initial virtual responsibility coefficients of all transmitting photovoltaic grid-connected units.

[0095] Step S33: Based on the voltage deviation rate of each transmitting photovoltaic grid-connected unit and its corresponding weight, calculate the weighted average value to obtain the neighborhood pressure factor.

[0096] In a specific embodiment, for the current computing unit i, its neighborhood pressure factor P locali Calculated using the following weighted average formula:

[0097] P locali =Σ{j∈N i}[w j ×ΔU j ];

[0098] The formula for calculating the weight wj is as follows:

[0099] w j =K j / (Σ{m∈N i}K m );

[0100] Therefore, the overall calculation formula is:

[0101] P locali =(Σ{j∈N i}[K j ×ΔU j ]) / (Σ{m∈N i}K m );

[0102] Among them, Plocali Represents the neighborhood pressure factor calculated for the i-th photovoltaic grid-connected unit, expressed as a percentage (%). Its physical meaning is the weighted average of the voltage deviation rates of all its neighboring units, evaluated from the perspective of unit i.

[0103] N i The "neighbor set" of the i-th grid-connected photovoltaic unit is the set of all grid-connected photovoltaic units that are in the same communication neighborhood (such as those connected to the same transformer in the same area) and have successfully exchanged messages with it, excluding itself.

[0104] j and m are both used to traverse the neighbor set N. i The index variables for different units are: j is used to identify a specific neighboring unit when calculating the weighted numerator, and m is used to identify a specific neighboring unit when calculating the sum of the denominators.

[0105] ΔU j The voltage deviation rate is extracted from the cooperative message of neighboring cell j, in (%).

[0106] K j Let be the virtual responsibility coefficient extracted from the collaborative message of neighboring unit j, which is dimensionless;

[0107] K m The virtual responsibility coefficient is extracted from the cooperative messages of neighboring unit m and is dimensionless.

[0108] Symbol Σ{j∈N i} represents the neighbor set N i Summing the quantities corresponding to all units in the equation.

[0109] It is understandable that in this formula, the weight w assigned to each neighboring unit j is... j It is equal to the neighbor's own virtual responsibility coefficient K. j The sum of virtual responsibility coefficients of all units in the entire neighbor set (i.e., the denominator Σ{m∈N) i}K m The proportion of a neighboring cell to the perceived neighborhood voltage situation. This means that the greater the self-declared responsibility of a neighboring cell, the greater its weight in the neighborhood voltage situation perceived by the current cell i. The denominator of the formula normalizes all weights to ensure that all weights w are normalized. j The sum of these values ​​is 1, which makes the calculated neighborhood pressure factor similar in both dimensions and numerical range to the voltage deviation rate ΔU. j They are directly comparable.

[0110] Understandably, the principle behind calculating the neighborhood pressure factor is to quantify the voltage anomalies in the local power grid. However, this quantification is not a simple arithmetic average; instead, it incorporates "perceived responsibility" as a weight. The core purpose of this is to make each unit's perception of neighborhood pressure more directional and rational. If a neighboring unit reports a high voltage deviation and simultaneously assesses and declares a high regulatory responsibility (a large virtual responsibility coefficient), its high voltage state will constitute a stronger "pressure" signal to the current unit. This design implies a collaborative logic: units that perceive serious problems or have significant regulatory potential (and therefore high initial responsibility coefficients) should receive more attention within their region and more significantly influence their neighbors' decisions. Conversely, if a unit has a high voltage but a low responsibility coefficient (possibly due to low output and limited regulatory margin), the "pressure" it exerts on its neighbors will be correspondingly weakened. Therefore, the neighborhood pressure factor not only reflects the average voltage level of neighbors but also a weighted combination of the "importance" or "urgency" of these voltage issues in the neighbors' self-perception. This guides the system to prioritize responding to regional pressures that are most needed and capable of being addressed by “high-responsibility, high-voltage” neighbors in subsequent adjustments, thereby enabling the entire collaborative process to evolve in a more efficient and resource-focused direction.

[0111] Step S4: Each photovoltaic grid-connected unit obtains the pre-stored local regulation sensitivity coefficient, and adjusts the initial virtual responsibility coefficient based on the neighborhood pressure factor and the local regulation sensitivity coefficient to obtain the virtual responsibility coefficient;

[0112] Specifically, in step S4, the adjustment range of the virtual responsibility coefficient is positively correlated with the product of the neighborhood pressure factor and the local adjustment sensitivity coefficient, and negatively correlated with the virtual responsibility coefficient before adjustment.

[0113] In one specific embodiment, the photovoltaic grid-connected unit performs the following calculation process to update its virtual liability factor:

[0114] First, the pre-stored local adjustment sensitivity coefficient Si is read from the local non-volatile memory.

[0115] Then, based on the neighborhood pressure factor, the local adjustment sensitivity coefficient, and the virtual responsibility coefficient of the current period, the updated virtual responsibility coefficient is calculated according to the following formula:

[0116] K i[new] =K i[old] +γ×S i ×P locali / K i[old] ;

[0117] in:

[0118] K i[new] Let be the virtual responsibility coefficient of the i-th photovoltaic grid-connected unit after this adjustment, which will be used for broadcasting in the next cycle. It is dimensionless.

[0119] K i[old] Let be the virtual responsibility coefficient of the i-th grid-connected photovoltaic unit before the adjustment in the current cycle, which is dimensionless. In the first adjustment cycle, its value is the initial virtual responsibility coefficient K calculated in step S1. i ;

[0120] P = The neighborhood pressure factor of the i-th cell calculated in step S3 is expressed as a percentage (%).

[0121] S i Let be the local regulation sensitivity coefficient of the i-th unit, expressed as a percentage per percent (% / %). Physically, it represents the expected change in local voltage deviation rate caused by a change in the unit's virtual responsibility factor. This coefficient is a pre-stored positive number characterizing the regulation capability of this node.

[0122] γ is the convergence step size coefficient, a small positive constant, dimensionless, used to control the magnitude of each adjustment and ensure the stability of the iteration process.

[0123] It is understandable that in the formula (S) i ×P locali This constitutes the main driving force for adjustment, with dimensions of %, where % represents the combined intensity of "pressure-capacity". To ensure consistency with the dimensionless K... i[old] The product needs to be divided by a reference value with the same "intensity" dimension; in this implementation, we choose to divide by K. i[old] The local regulation sensitivity coefficient Si can be pre-calibrated through offline power flow calculation, field tests, or system identification. For a relatively stable power grid topology, its value can be considered a constant over a period of time. The value of the convergence step size coefficient γ is crucial to the convergence speed and stability of the algorithm. The value range is 0.01 to 0.2, and preferably, γ = 0.05.

[0124] Understandably, the adjustment is driven by two factors: first, external factors, namely the "neighborhood pressure factor," which quantifies the severity of the overall high voltage problem in neighboring units and the extent of their self-identified responsibility. Greater pressure means a higher expectation that the local unit will assume more regulation tasks. Second, internal factors, namely the "local regulation sensitivity coefficient," which objectively reflects the unit's ability to regulate voltage at its electrical location. Higher sensitivity means that the same power adjustment is more effective in alleviating the voltage problem, thus requiring more regulation responsibility. The formula multiplies these two factors to obtain a "regulation driving force." However, to prevent some units from experiencing unlimited growth in their responsibility coefficient due to initial conditions or excessive sensitivity, and to reflect the synergy of "relatively balanced responsibility," the formula introduces a negative correlation term with the current responsibility coefficient as the denominator. This means that the greater the responsibility a unit currently identifies, the more the same driving force will suppress its effect on increasing its responsibility coefficient. This self-suppressive feedback mechanism ensures that, in multiple iterations, the responsibility coefficients of each unit will not diverge but will tend towards a stable equilibrium point that matches their respective "pressure contribution" and "regulation effectiveness." This invention is based entirely on local information and limited neighborhood information. By having all units execute this update rule in parallel and periodically, it ultimately guides the entire distributed system to spontaneously form a reasonable, fair, and effective virtual responsibility allocation scheme without a central coordinator, laying a consensus foundation for subsequent precise power regulation.

[0125] Step S5: Repeat steps S2 to S4. Based on the comparison results of each photovoltaic grid-connected unit determining that its virtual responsibility coefficient is less than or equal to the preset responsibility coefficient convergence threshold, and the rate of change of voltage deviation rate of key photovoltaic grid-connected units in the neighborhood is less than or equal to the preset voltage deviation rate convergence threshold, a consensus is reached and each photovoltaic grid-connected unit enters parameter steady state.

[0126] Please continue reading. Figure 4 The diagram shows a flowchart of step S5 in the distributed photovoltaic grid-connected flexible voltage regulation method of the present invention. Specifically, step S5 includes:

[0127] Step S51: After each completion of step S4, calculate the instantaneous rate of change of the current virtual responsibility coefficient compared to the previous calculation result;

[0128] In one specific embodiment, the photovoltaic grid-connected unit calculates the new virtual responsibility factor K according to step S4. i[new] Then, immediately calculate this value relative to the virtual responsibility factor K used in the previous period. i[old] instantaneous relative rate of change ΔK iThe rate of change is calculated as follows: take the absolute value of the difference between the old and new coefficients, and then divide it by the coefficient value of the previous period. That is, the rate of change is equal to the ratio of the absolute value of the difference between the new virtual responsibility coefficient and the old virtual responsibility coefficient to the old virtual responsibility coefficient.

[0129] Understandably, the purpose of calculating this rate of change is to quantify the fluctuation range of the virtual responsibility coefficient between two adjacent control cycles, which is a key indicator for judging whether the state of the unit itself tends to stabilize.

[0130] Step S52: Traverse all received neighborhood coordination messages, and by comparing the voltage deviation rate in each neighborhood coordination message, obtain at least one photovoltaic grid-connected unit with the largest voltage deviation rate, which is then used as the key photovoltaic grid-connected unit.

[0131] In one specific embodiment, after receiving the message in step S2, the photovoltaic grid-connected unit iterates through all records stored in its local neighborhood status table. It reads the voltage deviation rate field from each record and identifies the highest voltage deviation rate by comparison. One or more photovoltaic grid-connected units that report the highest voltage deviation rate are designated as "critical photovoltaic grid-connected units" for the current period. If multiple units report the same highest voltage deviation rate, they are all considered critical photovoltaic grid-connected units.

[0132] Understandably, the purpose of this step is to dynamically identify the node with the most prominent voltage problem from the neighborhood. The key photovoltaic grid-connected unit represents the point in the local power grid where voltage exceedances are most severe or the voltage level is highest; changes in its state directly reflect the effectiveness of coordinated voltage regulation measures in improving the most critical problem. Using it as an observation point for the overall system convergence ensures that the goal of coordination is to genuinely solve the voltage problem, rather than merely achieving mathematical equilibrium among the algorithm's internal variables.

[0133] Step S53: Monitor the voltage deviation rate of the key photovoltaic grid-connected unit and calculate its instantaneous change rate;

[0134] In one specific embodiment, the photovoltaic grid-connected unit extracts the neighbor records of those identified as key photovoltaic grid-connected units in step S52 from the local neighborhood state table. For each key unit, it obtains the voltage deviation rate reported in the current cycle and the voltage deviation rate reported in the previous cycle from the local cache. Then, it calculates the instantaneous rate of change of the voltage deviation rate of the key unit, denoted as ΔUr; ΔUr is calculated by taking the absolute value of the difference between the voltage deviation rate of the current cycle and the voltage deviation rate of the previous cycle, and then dividing it by the voltage deviation rate of the previous cycle.

[0135] Understandably, the purpose of this step is to monitor the dynamic changes in the voltage state of the most critical nodes. Calculating the rate of change is to determine whether adjustments made to address the most severe voltage issues have produced a stabilizing effect. Cached data from the previous cycle forms the basis for calculating the rate of change, and this falls under the category of routine data storage and retrieval operations.

[0136] Step S54: When the instantaneous rate of change of the current virtual responsibility coefficient is lower than the first convergence threshold and the instantaneous rate of change of the voltage deviation rate of the key photovoltaic grid-connected unit is lower than the second convergence threshold, it is determined that the system has reached a consensus and the current photovoltaic grid-connected unit enters the parameter steady state.

[0137] In one specific embodiment, the photovoltaic grid-connected unit will use the rate of change ΔK of its own virtual responsibility coefficient calculated in step S51. i With a preset first convergence threshold The comparison is performed; simultaneously, the voltage deviation rate change ΔU of the key photovoltaic grid-connected unit calculated in step S53 is compared. r (If multiple key units exist, the maximum rate of change is taken), along with a preset second convergence threshold. Comparison. If and only if and When both conditions are met simultaneously, the photovoltaic grid-connected unit determines that the distributed collaborative system has reached a consensus and enters a parameter steady state. First convergence threshold. With the second convergence threshold All are dimensionless small positive numbers, typically ranging from one-thousandth in value. Preferably, The value is 0.005. The value is 0.01.

[0138] Understandably, the dual-threshold comparison is a logical decision-making step in determining convergence. The first convergence threshold is used to determine whether the unit's own "responsibility perception" has been basically fixed, that is, whether the internal negotiation process has stopped. The second convergence threshold is used to determine whether the "outstanding problems" of the external power grid have been basically stabilized, that is, whether the effects of the adjustment actions have been manifested and no longer fluctuate drastically. Both must be satisfied simultaneously to declare a steady state reached, which ensures the reliability and effectiveness of the coordinated results.

[0139] Step S55: If the determination condition of step S54 is not met, return to step S2 to perform message broadcasting, pressure factor calculation and virtual responsibility coefficient adjustment for the next cycle.

[0140] In a specific embodiment, if the photovoltaic grid-connected unit's determination result in step S54 is "not satisfied," meaning either its own rate of change of responsibility coefficient or the rate of change of voltage of its critical neighbor exceeds its corresponding convergence threshold, then the unit will not enter a steady state. It will then use the new virtual responsibility coefficient K calculated in step S4. i[new] Update the value to the current value for the next broadcast, then wait to enter the next control cycle, and automatically start from step S2 to continue to execute a new round of collaborative message broadcasting, receiving, pressure factor calculation and responsibility coefficient adjustment until the convergence condition of step S54 is met.

[0141] Understandably, this step constitutes the iterative loop mechanism of the entire cooperative algorithm. The failure to meet the judgment condition means that the system has not yet found a stable responsibility allocation scheme or that the voltage problem is still dynamically changing, therefore iteration must continue. Returning to step S2 means starting a new cooperative cycle, using the latest local and neighborhood information for the next round of calculation and adjustment.

[0142] Understandably, step S5 defines the achievement of "cooperative consensus" through a dual criterion: the first criterion focuses on the internal state, namely whether the responsibility coefficient of each unit is stable. This reflects that all participating units no longer have significant disagreements on the question of "who should bear how much responsibility," and local negotiation has reached a stable point similar to Nash equilibrium. The second criterion focuses on the external effect, namely whether the voltage of the critical node with the most severe voltage problem in the region is stable. This ensures that the achievement of internal consensus truly leads to the effective suppression of the most pressing grid problem, and its state no longer deteriorates or fluctuates drastically. Satisfying only internal stability may simply be the algorithm spinning at a local optimum, while satisfying only external stability may ignore the situation where the responsibility of some units is still slightly oscillating. Combining the two means that the system has not only formed a stable responsibility allocation scheme, but this scheme is also having a stable and continuous regulatory effect on the critical fluctuation points of the grid.

[0143] Step S6: After the photovoltaic grid-connected unit enters the parameter steady state, the corresponding power regulation amount is determined according to the virtual responsibility coefficient at the parameter steady state through a preset mapping relationship, and the connected photovoltaic inverter is controlled to adjust based on the power regulation amount to complete the coordinated voltage regulation.

[0144] In a specific embodiment, after the photovoltaic grid-connected unit determines that it has entered a parameter steady state according to step S5, it uses the virtual responsibility coefficient finally determined in the steady state as input to query or calculate a preset mapping relationship to obtain the corresponding power regulation amount. This mapping relationship is usually represented as a function or lookup table, and its construction principle is: the larger the virtual responsibility coefficient, the larger the corresponding power reduction amount (in overvoltage scenarios). Subsequently, the unit generates a specific power regulation command, the target value of which is the difference between the current inverter output power and the power regulation amount. Finally, the photovoltaic grid-connected unit sends the power reference value command or a direct power limiting command to the photovoltaic inverter it is connected to through its standard communication interface (such as RS485 or Ethernet) using a common photovoltaic inverter monitoring protocol (such as Modbus TCP / RTU). The inverter receives and executes the command, adjusting its actual output power, thereby achieving flexible regulation of the grid connection point voltage.

[0145] Step S7: When each photovoltaic grid-connected unit detects that the local voltage deviation rate is continuously lower than the recovery threshold, the virtual responsibility coefficient is adjusted in a negative iterative manner until the photovoltaic inverter recovers to normal output.

[0146] Specifically, S7 includes:

[0147] Step S71: When the photovoltaic grid-connected unit detects that the local voltage deviation rate is continuously less than the recovery threshold, the first adjustment amount is calculated based on the current virtual responsibility coefficient and the current local voltage deviation rate.

[0148] In one specific embodiment, after the photovoltaic grid-connected unit enters a parameter steady state and performs voltage regulation, it continuously monitors the local voltage deviation rate. When the deviation rate ΔU is detected... i The value is less than the preset recovery threshold ΔU for N consecutive control cycles. ref If the voltage problem is basically resolved, then this recovery process can be initiated. The first adjustment amount is calculated using the following formula:

[0149] ;

[0150] Wherein, δK i The first adjustment is dimensionless and has a negative value, indicating a reduction in the virtual liability coefficient.

[0151] η is the recovery damping coefficient, a small, dimensionless positive constant used to control the speed and stability of the recovery process. Its value ranges from 0.05 to 0.3. Preferably, η is 0.1 to achieve a balance between recovery speed and stability.

[0152] K i[cur] The virtual liability coefficient at the current moment (which is the steady-state liability coefficient Ki at the initial moment of starting the recovery process) is dimensionless.

[0153] ΔU i The current monitored local voltage deviation rate (%).

[0154] ΔU ref The preset recovery threshold (%) is usually set to be slightly higher than the rated voltage but significantly lower than the voltage over-limit protection value, for example, 2%.

[0155] Understandably, the first adjustment amount δK in the formula... i The value is negative, and its magnitude is mainly determined by three parts: First, it is proportional to the current responsibility coefficient, meaning that the greater the responsibility of the unit, the greater the reduction in the initial recovery phase, which conforms to the fair logic of "those who bear more responsibility should withdraw first"; second, it is related to (ΔU) ref ΔU i This is directly proportional to the voltage deviation rate, meaning that the lower the current voltage deviation rate (i.e., the better the voltage recovery), the larger the absolute value of the adjustment amount, and the faster the recovery speed; thirdly, divide by ΔU ref Normalization is performed, and the overall scaling is done using the recovery damping coefficient η to ensure a smooth adjustment. Recovery threshold ΔU ref This setting is to establish a "safety redundancy zone" to avoid frequent switching between regulation and recovery states caused by voltage fluctuations near the critical value.

[0156] Step S72: Reduce the current virtual responsibility coefficient according to the first adjustment amount to obtain the updated virtual responsibility coefficient;

[0157] Step S73: Based on the updated virtual responsibility coefficient, determine the corresponding updated power regulation command based on the preset mapping relationship, and control the connected photovoltaic inverter to execute the updated power regulation command;

[0158] In one specific embodiment, the execution logic of this step is exactly the same as that of step S6. The photovoltaic grid-connected unit takes the updated virtual responsibility factor as input and determines the corresponding new power regulation amount through the same preset mapping relationship described in step S6. Subsequently, a new power regulation command is generated and issued to the photovoltaic inverter. As the virtual responsibility factor decreases, the new power regulation amount also decreases, meaning that the inverter's power limit can be relaxed, and the output power can be increased.

[0159] Step S74: After executing the updated power adjustment command, monitor the local voltage deviation rate again. If it is still lower than the recovery threshold, repeat steps S71 to S73 to perform the next round of adjustment.

[0160] Step S75: Repeat the above adjustment until the virtual responsibility coefficient is reduced to the virtual responsibility coefficient threshold, and the photovoltaic inverter returns to normal output state.

[0161] In a specific embodiment, the negative iterative loop consisting of steps S71 to S74 described above will continue indefinitely. In each loop, the virtual responsibility coefficient is reduced by a first adjustment amount determined by the current voltage and the coefficient. When the coefficient is reduced to the virtual responsibility coefficient threshold through iteration, the loop terminates. The virtual responsibility coefficient threshold is zero or a pre-set, negligible small positive number (e.g., 0.01), which can be adjusted by those skilled in the art according to actual conditions, and will not be elaborated further here. At this time, according to the mapping relationship, the power regulation amount is also zero, and the instruction received by the photovoltaic inverter is to cancel the power limit (or set it to the rated capacity), thereby fully restoring to the normal output state.

[0162] Understandably, step S7 defines the termination conditions of the recovery process to enable the system to autonomously, smoothly, and safely exit after the abnormal operating conditions are eliminated. When the voltage is detected to have stabilized below the safety threshold, it indicates that the external stress (such as overvoltage caused by a surge in photovoltaic power generation) has weakened or disappeared. At this point, continuing to operate under derating would result in unnecessary power generation losses. Therefore, the system needs a mechanism to automatically release the temporary control. This step uses negative iteration instead of a one-step zeroing to avoid the voltage potentially exceeding the limit again due to a sudden full recovery of power, i.e., "oscillation". By linking the reduction in the responsibility factor to the degree to which the current voltage deviates from the recovery threshold and to the current responsibility factor itself, an intelligent exit logic is achieved: "the better the voltage recovery, the faster the derating unit exits." This gradual recovery allows the grid voltage to maintain a smooth transition as photovoltaic output gradually recovers, effectively preventing secondary shocks.

[0163] The distributed photovoltaic grid-connected flexible voltage regulation system provided by this invention includes:

[0164] The status monitoring unit is deployed in each photovoltaic grid-connected unit to measure the current voltage of the local grid connection point in real time and calculate the voltage deviation rate, as well as to obtain the current output power of the connected photovoltaic inverter in real time and calculate the output percentage.

[0165] A responsibility coefficient determination unit, which is connected to the status monitoring unit, is used to determine the initial virtual responsibility coefficient based on the voltage deviation rate and the output percentage;

[0166] The communication unit, which is deployed in each photovoltaic grid-connected unit, is used to periodically broadcast a coordination message including its current voltage deviation rate and current virtual responsibility coefficient to other photovoltaic grid-connected units in the neighborhood, and to receive coordination messages from other photovoltaic grid-connected units in the neighborhood.

[0167] A pressure factor calculation unit, which is connected to the communication unit, is used to calculate the neighborhood pressure factor based on all received neighborhood cooperation messages.

[0168] The responsibility coefficient adjustment unit is connected to the pressure factor calculation unit, the responsibility coefficient determination unit, and the pre-stored local adjustment sensitivity coefficient, respectively, and is used to dynamically adjust the virtual responsibility coefficient based on the neighborhood pressure factor and the local adjustment sensitivity coefficient.

[0169] A consensus determination unit, which is connected to the responsibility coefficient adjustment unit and the communication unit, is used to determine whether a collaborative consensus has been reached and whether the parameters have entered a steady state based on the instantaneous change rate of the adjusted virtual responsibility coefficient and the instantaneous change rate of the voltage deviation rate of the key photovoltaic grid-connected unit determined from the received message.

[0170] A power regulation unit, which is connected to the consensus determination unit and the photovoltaic inverter respectively, is used to determine the corresponding power regulation amount according to the virtual responsibility coefficient of the steady state through a preset mapping relationship after entering the parameter steady state, and control the photovoltaic inverter to adjust based on the power regulation amount;

[0171] The recovery control unit is connected to the status monitoring unit, the responsibility coefficient adjustment unit, and the power regulation unit respectively. When the local voltage deviation rate is detected to be continuously lower than the recovery threshold, it initiates a negative iterative adjustment of the virtual responsibility coefficient and updates the power regulation amount synchronously until the photovoltaic inverter recovers to normal output state.

[0172] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A distributed photovoltaic grid-connected flexible voltage regulation method, characterized in that, include: Step S1: Each photovoltaic grid-connected unit monitors the voltage deviation rate and its own output percentage at the local grid connection point in real time to obtain the initial virtual responsibility coefficient of each local grid connection point. Step S1 includes: Step S11: Measure the current voltage at the local grid connection point in real time, and calculate the voltage deviation rate based on the current voltage and the rated voltage. The specific formula for calculating the voltage deviation rate is as follows: DU i =(U i -U N ) / U N 100%; Wherein, ΔU i U represents the voltage deviation rate of the i-th photovoltaic grid-connected unit, expressed as a percentage (%). i The current voltage measured in real time by this unit, in volts (V); U N The rated voltage specified for the power distribution network, in volts (V). Step S12: Obtain the current output power of the connected photovoltaic inverter in real time, and calculate the output percentage based on the current output power and the rated capacity of the photovoltaic inverter; Step S13: Determine the initial virtual responsibility coefficient based on the voltage deviation rate and the output percentage; Step S2: Each photovoltaic grid-connected unit periodically broadcasts a coordination message, including its current voltage deviation rate and the current initial virtual responsibility coefficient, to other photovoltaic grid-connected units in the neighborhood, and receives coordination messages from other photovoltaic grid-connected units in the neighborhood. Step S3: Each photovoltaic grid-connected unit calculates a neighborhood pressure factor based on all received neighborhood coordination messages. The neighborhood pressure factor is a weighted average of the voltage deviation rates of other photovoltaic grid-connected units in the neighborhood, with their respective initial virtual responsibility coefficients as weights. Step S3 includes: Step S31: Extract the voltage deviation rate and the initial virtual responsibility coefficient of each sending photovoltaic grid-connected unit from all received neighborhood cooperative messages; Step S32: Determine the weight of the photovoltaic grid-connected unit in the weighted calculation based on the initial virtual responsibility coefficient of each transmitting photovoltaic grid-connected unit; In step S32, the weight of the photovoltaic grid-connected unit in the weighted calculation is the ratio of the initial virtual responsibility coefficient of a single transmitting photovoltaic grid-connected unit to the sum of the initial virtual responsibility coefficients of all transmitting photovoltaic grid-connected units. Step S33: Based on the voltage deviation rate of each transmitting photovoltaic grid-connected unit and its corresponding weight, calculate the weighted average value to obtain the neighborhood pressure factor; Step S4: Each photovoltaic grid-connected unit obtains the pre-stored local regulation sensitivity coefficient, and adjusts the initial virtual responsibility coefficient based on the neighborhood pressure factor and the local regulation sensitivity coefficient to obtain the virtual responsibility coefficient. The local regulation sensitivity coefficient reflects the voltage regulation capability of the electrical location where the photovoltaic grid-connected unit is located. Step S5: Repeat steps S2 to S4. Based on the comparison results of each photovoltaic grid-connected unit determining that its virtual responsibility coefficient is less than or equal to the preset responsibility coefficient convergence threshold, and the rate of change of voltage deviation rate of key photovoltaic grid-connected units in the neighborhood is less than or equal to the preset voltage deviation rate convergence threshold, a consensus is reached and each photovoltaic grid-connected unit enters parameter steady state. Step S6: After the photovoltaic grid-connected unit enters the parameter steady state, the corresponding power regulation amount is determined according to the virtual responsibility coefficient at the parameter steady state through a preset mapping relationship, and the connected photovoltaic inverter is controlled to adjust based on the power regulation amount to complete the coordinated voltage regulation. Step S7: When each photovoltaic grid-connected unit detects that the local voltage deviation rate is continuously lower than the recovery threshold, the virtual responsibility coefficient is adjusted in a negative iterative manner until the photovoltaic inverter recovers to normal output.

2. The distributed photovoltaic grid-connected flexible voltage regulation method according to claim 1, characterized in that, In step S13, the first component of the initial virtual responsibility coefficient is positively correlated with the voltage deviation rate; the second component of the initial virtual responsibility coefficient is positively correlated with the output percentage; and the initial virtual responsibility coefficient is the sum of the first component and the second component.

3. The distributed photovoltaic grid-connected flexible voltage regulation method according to claim 2, characterized in that, Step S5 includes: Step S51: After each completion of step S4, calculate the instantaneous rate of change of the current virtual responsibility coefficient compared to the previous calculation result; Step S52: Traverse all received neighborhood coordination messages, and by comparing the voltage deviation rate in each neighborhood coordination message, obtain at least one photovoltaic grid-connected unit with the largest voltage deviation rate, which is then used as the key photovoltaic grid-connected unit. Step S53: Monitor the voltage deviation rate of the key photovoltaic grid-connected unit and calculate its instantaneous change rate; Step S54: When the instantaneous rate of change of the current virtual responsibility coefficient is lower than the first convergence threshold and the instantaneous rate of change of the voltage deviation rate of the key photovoltaic grid-connected unit is lower than the second convergence threshold, it is determined that the system has reached a consensus and the current photovoltaic grid-connected unit enters the parameter steady state. Step S55: If the determination condition of step S54 is not met, return to step S2 to perform message broadcasting, pressure factor calculation and virtual responsibility coefficient adjustment for the next cycle.

4. The distributed photovoltaic grid-connected flexible voltage regulation method according to claim 3, characterized in that, S7 includes: Step S71: When the photovoltaic grid-connected unit detects that the local voltage deviation rate is continuously less than the recovery threshold, the first adjustment amount is calculated based on the current virtual responsibility coefficient and the current local voltage deviation rate. Step S72: Reduce the current virtual responsibility coefficient according to the first adjustment amount to obtain the updated virtual responsibility coefficient; Step S73: Based on the updated virtual responsibility coefficient, determine the corresponding updated power regulation command based on the preset mapping relationship, and control the connected photovoltaic inverter to execute the updated power regulation command; Step S74: After executing the updated power adjustment command, monitor the local voltage deviation rate again. If it is still lower than the recovery threshold, repeat steps S71 to S73 to perform the next round of adjustment. Step S75: Repeat the above adjustment until the virtual responsibility coefficient is reduced to the virtual responsibility coefficient threshold, and the photovoltaic inverter returns to normal output state.

5. The distributed photovoltaic grid-connected flexible voltage regulation method according to claim 1, characterized in that, Other photovoltaic grid-connected units within the neighborhood refer to all other photovoltaic grid-connected units that are connected to the same transformer area as the current photovoltaic grid-connected unit.

6. The distributed photovoltaic grid-connected flexible voltage regulation method according to claim 1, characterized in that, In step S4, the adjustment range of the virtual responsibility coefficient is positively correlated with the product of the neighborhood pressure factor and the local adjustment sensitivity coefficient, and negatively correlated with the virtual responsibility coefficient before adjustment.

7. A distributed photovoltaic grid-connected flexible voltage regulation system, applied to the distributed photovoltaic grid-connected flexible voltage regulation method according to any one of claims 1-6, characterized in that, include: The status monitoring unit is deployed in each photovoltaic grid-connected unit to measure the current voltage of the local grid connection point in real time and calculate the voltage deviation rate, as well as to obtain the current output power of the connected photovoltaic inverter in real time and calculate the output percentage. A responsibility coefficient determination unit, which is connected to the status monitoring unit, is used to determine the initial virtual responsibility coefficient based on the voltage deviation rate and the output percentage; The communication unit, which is deployed in each photovoltaic grid-connected unit, is used to periodically broadcast a coordination message including its current voltage deviation rate and current virtual responsibility coefficient to other photovoltaic grid-connected units in the neighborhood, and to receive coordination messages from other photovoltaic grid-connected units in the neighborhood. A pressure factor calculation unit, which is connected to the communication unit, is used to calculate the neighborhood pressure factor based on all received neighborhood cooperation messages. The responsibility coefficient adjustment unit is connected to the pressure factor calculation unit, the responsibility coefficient determination unit, and the pre-stored local adjustment sensitivity coefficient, respectively, and is used to dynamically adjust the virtual responsibility coefficient based on the neighborhood pressure factor and the local adjustment sensitivity coefficient. A consensus determination unit, which is connected to the responsibility coefficient adjustment unit and the communication unit, is used to determine whether a collaborative consensus has been reached and whether the parameters have entered a steady state based on the instantaneous change rate of the adjusted virtual responsibility coefficient and the instantaneous change rate of the voltage deviation rate of the key photovoltaic grid-connected unit determined from the received message. A power regulation unit, which is connected to the consensus determination unit and the photovoltaic inverter respectively, is used to determine the corresponding power regulation amount according to the virtual responsibility coefficient of the steady state through a preset mapping relationship after entering the parameter steady state, and control the photovoltaic inverter to adjust based on the power regulation amount; The recovery control unit is connected to the status monitoring unit, the responsibility coefficient adjustment unit, and the power regulation unit respectively. When the local voltage deviation rate is detected to be continuously lower than the recovery threshold, it initiates a negative iterative adjustment of the virtual responsibility coefficient and updates the power regulation amount synchronously until the photovoltaic inverter recovers to normal output state.

Images