A reactive power optimization regulation method for a distributed photovoltaic power station power distribution system

By employing a three-tiered architecture for hierarchical regulation and a margin-aware allocation method, the problem of voltage exceeding limits in distributed photovoltaic power grids was solved, achieving efficient utilization of reactive power resources and stable control of system voltage, thereby improving system fault tolerance and operational economy.

CN123121682APending Publication Date: 2026-11-17XIAN CONGDUAN ELECTRIC CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611640046.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-10-22
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

The existing distribution network suffers from problems such as voltage exceeding limits, power factor exceeding standards, and increased network losses due to distributed photovoltaic access. Existing reactive power control technologies also suffer from problems such as control lag, resource waste, high communication dependence, and equipment malfunction.

Method used

By adopting a three-level architecture for hierarchical control, margin-aware allocation, dynamic threshold adaptation, communication fault backup, and steady-state optimization operation, the system achieves efficient utilization of reactive power resources and stable control of system voltage through the coordinated control of the main controller, regional group control units, distributed photovoltaic inverters, and static var compensators.

Benefits of technology

It achieves efficient utilization of reactive power resources and stable control of system voltage, improves system fault tolerance and operating economy, reduces the risk of equipment overload and communication interruption, and optimizes the power supply quality and equipment life of the power grid.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123121682A_ABST
    Figure CN123121682A_ABST
Patent Text Reader

Abstract

The application provides a kind of distributed photovoltaic power station power distribution system reactive power optimization regulation method, belong to power system reactive power voltage regulation technical field.The method is based on the three-level regulation architecture of main controller, regional group control unit, distributed photovoltaic inverter and static var compensator, including: real-time acquisition of each monitoring point three-phase voltage, current and active power data and time stamp synchronous upload;Calculate node voltage deviation and determine whether it exceeds the preset allowable range, while determining the reactive power shortage of each region based on power flow calculation;When voltage exceeds the limit, the main controller only issues the reactive power shortage of the region corresponding to the limit to the corresponding regional group control unit, and the group control unit generates regulation instructions according to the available reactive power margin and response time delay of the inverter.The method fully utilizes the reactive power regulation capability of the inverter through hierarchical collaborative regulation, reduces system loss, improves voltage stability, and enhances operation reliability under communication failure through local control and gain self-calibration.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of reactive power and voltage regulation technology in power systems, specifically to a method for optimizing reactive power regulation in a distributed photovoltaic power station distribution system. Background Technology

[0002] With the advancement of dual-carbon goals, the scale of distributed photovoltaic (PV) grid integration continues to grow. However, because PV power output is highly random and fluctuating due to the influence of sunlight and temperature, the large-scale integration of distributed PV can lead to problems such as voltage exceeding limits, power factor exceeding standards, and increased network losses, severely impacting power quality and equipment safety.

[0003] Existing reactive power control technologies for distribution networks mostly employ centralized automatic voltage control systems. These systems suffer from the following drawbacks: First, the centralized calculation and communication link latency is long, typically reaching the second level, making it difficult to respond to millisecond-level power fluctuations in distributed photovoltaic systems, easily leading to control lag. Second, they do not fully utilize the reactive power regulation capabilities of distributed photovoltaic inverters themselves, over-relying on centralized static var compensators, resulting in resource waste and high construction costs. Third, some hierarchical control schemes have hierarchical logic conflicts, with mismatched thresholds between backup control and centralized control, easily causing command oscillations. Fourth, they have high communication dependence; once communication is interrupted, terminal equipment becomes uncontrollable, highlighting voltage risks. Fifth, the steady-state operation mode design is unreasonable; blindly pursuing zero reactive power output actually leads to increased system network losses and substandard power factor. Furthermore, most solutions do not consider the impact of temperature derating when calculating the reactive power regulation capacity of the inverter, and estimate the reactive power margin based on the rated capacity. In actual operation, it is easy for the issued commands to exceed the adjustable range of the equipment, triggering overload or overheat protection actions and causing regulation failure. At the same time, the allowable range of voltage deviation often adopts a fixed threshold, which cannot be adaptively adjusted with the fluctuation of load and photovoltaic output. If the threshold is too strict, the regulation will be frequent and affect the life of the equipment. If the threshold is too wide, it will exceed the limit and become uncontrollable, and the power supply quality will deteriorate.

[0004] Therefore, a reactive power optimization and control method for distributed photovoltaic power distribution systems that is hierarchical, logically coordinated, fault-tolerant, and cost-effective is proposed to solve the above-mentioned problems in existing technologies. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing a reactive power optimization and control method for distributed photovoltaic power station distribution systems. Through techniques such as three-level architecture hierarchical control, margin-sensing allocation, dynamic threshold self-adaptation, communication fault backup, and steady-state optimized operation, this method achieves efficient utilization of reactive power resources and stable control of system voltage, while improving system fault tolerance and operational economy.

[0006] The specific technical solution is as follows: A reactive power optimization and control method for a distributed photovoltaic power station distribution system, wherein the distribution system is divided into multiple control zones, each control zone is equipped with a regional group control unit, and is uniformly coordinated by a main controller, forming a three-level control architecture of main controller, regional group control unit, distributed photovoltaic inverter and static var compensator. The method includes the following steps: The system collects three-phase voltage, current, and active power data from each monitoring point in the power distribution system in real time, and uploads them to the main controller after synchronization with timestamps. The voltage deviation value of each node is calculated based on the collected data, and it is determined whether the voltage deviation value exceeds the preset voltage deviation allowable range. At the same time, the reactive power deficit of each area of ​​the power distribution system is determined based on power flow calculation. When the judgment result is that the voltage deviation exceeds the allowable range, the main controller only sends the reactive power deficit corresponding to the voltage over-limit area to the corresponding area group control unit. The area group control unit determines the available reactive power margin of each inverter based on the current active power output and operating temperature of each distributed photovoltaic inverter under its jurisdiction, and generates reactive power control instructions in a hierarchical manner according to the available reactive power margin and the time delay characteristics from instruction reception to response completion, in order of priority from high to low: distributed photovoltaic inverters with response delay less than the first time delay threshold are given priority for allocation, and the reactive power deficit is allocated according to the proportion of the available reactive power margin of each inverter, and the allocation result of each inverter does not exceed its available reactive power margin; the remaining reactive power deficit is allocated by distributed photovoltaic inverters with response delay between the first time delay threshold and the second time delay threshold according to the proportion of available reactive power margin, and the allocation result of each inverter does not exceed its available reactive power margin; the remaining reactive power deficit is supplemented to zero by the output of the static var compensator. The reactive power control command is sent to each distributed photovoltaic inverter and static var compensator through the communication network to adjust its reactive power output and maintain the system voltage within the allowable voltage deviation range. When the distributed photovoltaic inverter detects a communication interruption with the regional group control unit, it automatically switches to local control mode and adjusts the reactive power output in real time using the voltage-reactive power droop characteristic curve based on the measured voltage value of its own grid connection point. Under normal communication conditions, the local control mode is in a locked state.

[0007] This solution constructs a complete technical framework integrating regional collaboration, margin awareness, time delay layering, and fault-tolerant interlocking: reactive power deficits are only issued to areas exceeding voltage limits, avoiding ineffective regulation and communication burdens caused by global unified scheduling; reactive power margins can be combined with active power output and operating temperature calculations to ensure that allocation commands do not exceed equipment physical limits from the source; layering based on response delay and constraining allocation results to not exceed the margin upper limit achieves optimal matching between response speed and equipment safety; local droop control is implemented during communication interruptions, and interlocking occurs during normal communication, forming a fault-tolerant system of primary and backup collaboration. These features work together to solve the technical problems of control lag, command exceeding limits, hierarchical conflicts, and communication failures in existing technologies.

[0008] As a preferred embodiment of the present invention, the method for calculating the available reactive power margin is as follows: Based on the apparent power rating of each distributed photovoltaic inverter under the current active power output, the capacity occupancy corresponding to the current active power is deducted to obtain the theoretical upper limit of reactive power adjustment. The current operating temperature of the inverter is collected, and the theoretical adjustment upper limit is corrected based on the preset temperature-reduction coefficient piecewise function to obtain the available reactive power margin. The temperature-reduction coefficient piecewise function is as follows: when the temperature is below 40℃, the reduction coefficient is 95% to 100%; when the temperature is between 40℃ and 60℃, the reduction coefficient is 85% to 95%; and when the temperature is above 60℃, the reduction coefficient is 70% to 85%.

[0009] This scheme refines the calculation logic of available reactive power margin and adopts a temperature-segmented reduction method to match the actual operating characteristics of the inverter, avoiding the error of fixed coefficients. It fully taps the reactive power regulation capability of the inverter and prevents high-temperature overload operation, thereby improving the feasibility of control commands and equipment safety.

[0010] In a preferred embodiment of the present invention, the first delay threshold is 100ms, and the second delay threshold is 500ms; the hierarchical generation of reactive power control instructions specifically includes: First-level regulation: Select distributed photovoltaic inverters with available reactive power margin greater than 0 and response delay less than 100ms to take priority in reactive power output regulation; Second-level regulation: Select distributed photovoltaic inverters with response delays between 100ms and 500ms, and allocate the remaining reactive power deficit according to the proportion of available reactive power margin; The third level of control: The regional group control unit controls the static var compensator to output supplementary reactive power until the remaining reactive power deficit is zero.

[0011] The scheme clarifies the time delay boundary and allocation rules of the three-layer control, and calls reactive power resources layer by layer according to the response speed from fast to slow. It not only makes full use of the inverter's fast adjustment capability to smooth out instantaneous fluctuations, but also uses the static var compensator to compensate for large capacity shortages, and constrains the allocation results of each inverter to not exceed its physical adjustable upper limit, thus achieving the optimal match between adjustment speed, compensation capacity and equipment safety.

[0012] As a preferred embodiment of the present invention, the reactive power control command includes a target reactive power value and a command execution duration, wherein the command execution duration is set to 500ms to 2000ms; After the command is issued, the system continuously collects data and recalculates the voltage deviation and reactive power deficit. If the voltage deviation is detected to recover to the allowable range before the execution time of the command ends, the execution of the current control command is terminated in advance, and the current reactive power output state of each device is maintained until the next round of control command is issued.

[0013] The scheme incorporates a dynamic termination and status maintenance mechanism, which can terminate the adjustment in advance based on the real-time status of the system, avoiding voltage oscillations caused by over-adjustment. At the same time, it clearly defines the operating status after termination, ensuring continuous and stable system operation and improving the smoothness of the control process.

[0014] As a preferred embodiment of the present invention, the communication network adopts a hierarchical architecture combining wired Ethernet and long-distance low-power wireless communication: Wired Ethernet is used for data transmission and command interaction between the main controller and the group control units in each area; Long-range low-power wireless communication is used for command issuance and operation data feedback between the regional group control unit and its subordinate distributed photovoltaic inverters and static var compensators.

[0015] This scheme adopts a hierarchical heterogeneous communication approach. The core layer ensures high reliability and low latency through wired Ethernet, while the terminal layer achieves wide coverage and low-cost deployment through long-distance low-power wireless communication. The static var compensator is also incorporated into the terminal layer for unified management, which is suitable for the dispersed nature of distributed photovoltaic points and takes into account both real-time control and engineering economy.

[0016] As a preferred embodiment of the present invention, before calculating the voltage deviation and reactive power deficit, the collected three-phase voltage, current and active power data are further filtered. The filtering method uses a moving average filtering method with a sliding window length of 5 to 20 sampling points and a time span of 0.1s to 1s to eliminate instantaneous interference noise.

[0017] This scheme introduces a data preprocessing step, which effectively eliminates sampling noise and instantaneous disturbances by using a sliding window filter that adapts to the sampling period, avoids false triggering of control, improves the accuracy of reactive power deficit calculation and voltage judgment, and enhances the anti-interference capability of the control system.

[0018] As a preferred embodiment of the present invention, each monitoring point in the power distribution system includes a photovoltaic grid connection point, the beginning of the feeder, the end of the feeder, and the low-voltage side busbar of the power distribution transformer; The data acquisition cycle at each monitoring point is 20ms to 50ms. All collected data is uploaded after being synchronized with timestamps through a unified clock.

[0019] The scheme clarifies the key monitoring points and synchronous acquisition mechanism, covers the core characteristic nodes of voltage and power in the distribution network, and the unified time scale ensures the comparability of multi-source data, providing an accurate data foundation for global power flow calculation and reactive power control.

[0020] As a preferred embodiment of the present invention, when the voltage deviation value does not exceed the allowable range of the voltage deviation, the system enters a steady-state optimization mode: With the goal of maintaining the system power factor in the range of 0.95 to 1.0 and minimizing the active power loss of the lines, the reactive power output of each distributed photovoltaic inverter is optimized and the static var compensator is controlled to operate in a low-loss standby state.

[0021] This scheme achieves steady-state optimization with the goals of meeting power factor targets and minimizing network losses. It satisfies the power grid's power factor assessment requirements while reducing line losses caused by reactive circulating currents, thus improving the economy and compliance of system operation.

[0022] As a preferred embodiment of the present invention, after each execution of the reactive power control command, the system records the actual reactive power output value of each distributed photovoltaic inverter and static var compensator at the current moment, compares it with the target reactive power value, calculates the output deviation rate, and when the target reactive power value is zero, calculates the output deviation rate with the rated reactive power capacity of the corresponding device as the denominator. If the absolute value of the output deviation rate exceeds 10% for three consecutive control cycles, the gain calibration process is triggered: the reactive power control gain coefficient of each device is corrected according to historical deviation data so that the actual output tracks the target value.

[0023] The scheme designs a self-calibration mechanism based on deviation statistics, which compensates for the output deviation caused by equipment aging and environmental changes by correcting the control gain coefficient. It does not require modification of the equipment's factory characteristics, has strong engineering feasibility, and can maintain stable control accuracy over a long period of time.

[0024] As a preferred embodiment of the present invention, the preset voltage deviation allowable range is dynamically set within the range of 3% to 7% of the rated voltage; The settings are based on the current system load factor, the predicted photovoltaic output, and the load conditions of adjacent feeders. The setting rules are: the higher the load factor, the greater the fluctuation of photovoltaic output, and the smaller the allowable range of voltage deviation. The dynamic update step is once every 15 minutes to 60 minutes.

[0025] This scheme achieves dynamic adaptive adjustment of the voltage deviation threshold. While ensuring voltage quality, it flexibly adjusts the control sensitivity according to the system operating conditions. During high-risk periods, the threshold is tightened to ensure safety, while during stable periods, the threshold is relaxed to reduce the frequency of control, thus balancing power supply quality and equipment lifespan.

[0026] As a preferred embodiment of the present invention, the dynamic setting rule is specifically as follows: when the system load rate is greater than 70% or the photovoltaic output prediction fluctuation rate is greater than 20% / min, the allowable range of voltage deviation is tightened to 3% of the rated voltage; when the system load rate is between 30% and 70% or the photovoltaic output prediction fluctuation rate is between 10% / min and 20% / min, the allowable range of voltage deviation is set to 5% of the rated voltage; when the system load rate is less than 30% and the photovoltaic output prediction fluctuation rate is less than 10% / min, the allowable range of voltage deviation is widened to 6.5% of the rated voltage.

[0027] This scheme refines the dynamic threshold into three executable rules, using load factor and photovoltaic volatility as dual-dimensional criteria to divide the levels. There are no coverage gaps between the levels, and the sensitivity of the control strategy changes with the operating condition gradient. This ensures voltage safety under high-risk conditions while avoiding frequent adjustments under low-risk conditions, making it highly feasible for engineering projects.

[0028] In the dynamic setting rules, when the load factor and the predicted fluctuation rate of photovoltaic output fall into the ranges corresponding to different levels, the more stringent level (i.e., the smaller allowable range of voltage deviation) shall be used.

[0029] In a preferred embodiment of the present invention, the static var compensator in the third-level control adopts a closed-loop regulation method: The remaining reactive power deficit is used as the control setpoint, and the actual reactive power output of the static var compensator is used as the feedback value. A proportional-integral controller is used for tracking control. The proportional coefficient of the proportional-integral controller is set to 0.5 to 2.0, and the integral time constant is set to 0.1s to 1.0s.

[0030] The scheme clarifies the closed-loop control parameters of the static var compensator, which correspond to the application scenarios of the third-level control. The proportional-integral control can quickly eliminate steady-state errors, and the parameter range is adapted to the typical operating conditions of the power distribution system. It can ensure the compensation response speed without causing overshoot oscillation.

[0031] As a preferred embodiment of the present invention, the voltage-reactive power droop characteristic of the local control mode is specifically as follows: When the grid connection point voltage is higher than the upper limit of the reference voltage, the inverter absorbs inductive reactive power; the higher the voltage, the more inductive reactive power is absorbed. When the grid connection point voltage is lower than the lower limit of the reference voltage, the inverter generates capacitive reactive power. The lower the voltage, the more capacitive reactive power is generated. When the grid connection point voltage is between the upper and lower limits of the reference voltage, the inverter's reactive power output remains at the current value. The change in reactive power output is linearly related to the voltage deviation, and the proportionality coefficient is the droop coefficient.

[0032] This scheme refines the segmented characteristics of local droop control, sets a voltage dead zone to avoid frequent inverter adjustments, and ensures smooth and continuous reactive power output through linear regulation. It can stably maintain the grid connection point voltage in communication failure scenarios, thus achieving autonomous operation without communication.

[0033] The upper and lower limits of the reference voltage in the local control mode are adapted to the allowable range of voltage deviation in the centralized control mode. When the allowable range of voltage deviation in the centralized control mode is tightened to 3%, the reference voltage range of the local control mode is adjusted accordingly to 9.7kV-10.3kV; when the allowable range of voltage deviation in the centralized control mode is widened to 6.5%, the reference voltage range of the local control mode is adjusted accordingly to 9.35kV-10.65kV.

[0034] As a preferred embodiment of the present invention, before calculating the voltage deviation and reactive power deficit, a bad data identification step is also included: if the difference between the current sampled value and the previous filtered value exceeds 20% of the corresponding rated value, the sampled point is determined to be bad data and replaced by the filtered value of the previous time; when the same monitoring point is determined to be bad data for 3 consecutive collection cycles, the data quality of the monitoring point is determined to be abnormal, an alarm is issued and the most recent valid historical data is used for calculation.

[0035] This scheme superimposes a bad data identification and data quality alarm mechanism on the basis of filtering. It can identify abnormal data caused by sensor failure, communication errors, etc., prevent abnormal values ​​from polluting voltage judgment and reactive power deficit calculation, and further improve the reliability of control decisions.

[0036] As a preferred embodiment of the present invention, the multiple control zones are divided according to the power supply range of the feeder or the electrical distance, and each control zone is equipped with one regional group control unit, which is deployed at the box-type transformer or feeder sectionalizing switch within the zone; the regional group control unit only manages the distributed photovoltaic inverters and static var compensators within its own zone.

[0037] The plan clarifies the basis for dividing the control areas and the deployment location of the group control units. By dividing the areas according to electrical coupling, the resources within the areas can be more closely coordinated, and the areas will not affect each other. With the mechanism of issuing control orders only to areas that exceed the limit, the plan can achieve precise delivery of control orders and regional autonomy.

[0038] The present invention has the following beneficial effects: 1. It adopts the physical quantity corresponding logic of voltage over-limit triggering and regional reactive power deficit regulation, avoiding the principle error of directly comparing different dimensions; deficit instructions are only issued to over-limit areas to achieve precise regulation; three-level hierarchical regulation matches resources according to response speed, taking into account both speed and capacity margin.

[0039] 2. Fully leverage the reactive power regulation capabilities of distributed photovoltaic inverters, safely release inverter reserve capacity through temperature reduction algorithms, reduce reliance on centralized reactive power compensation devices, and lower project investment.

[0040] 3. The local droop control is set as a backup after communication interruption. There is no command conflict when the local mode is locked during normal communication. It has a self-calibration function and can maintain the control accuracy for a long time.

[0041] 4. The steady-state mode is optimized with the goal of minimizing network loss and achieving the power factor target, rather than blindly outputting zero reactive power, so as to meet the grid assessment requirements and reduce line loss.

[0042] 5. The hierarchical heterogeneous communication solution is adapted to the characteristics of distributed photovoltaic layout, and the parameter range covers typical power distribution conditions, which can be flexibly configured according to actual scenarios.

[0043] 6. The system utilizes reactive power margin combined with active power output and temperature reduction calculations, working in conjunction with hierarchical allocation based on response delay. The fast layer handles instantaneous fluctuations, while the slow layer handles steady-state deficits, preventing overload of the high-speed layer and lag in the low-speed layer. Dynamic voltage thresholds and gain self-calibration mechanisms ensure control accuracy from the strategy and execution layers respectively; the former adapts to operating condition sensitivity, while the latter cumulatively corrects output deviations. Centralized control and local interlocking switching form a fault-tolerant system with primary and backup coordination, ensuring no command conflicts during normal communication and no risk of loss of control during communication interruptions. These features mutually support each other functionally and enhance each other's effects, forming a complete technical closed loop distinct from existing hierarchical control schemes. Attached Figure Description

[0044] Figure 1 This is a flowchart illustrating the overall process of the reactive power optimization and control method described in this invention. Figure 2 This is a schematic diagram of the three-level hierarchical control logic of the present invention; Figure 3 This is a communication topology diagram of the three-level control architecture of the present invention; Figure 4This is a voltage-reactive power droop characteristic curve of the local control mode of the present invention. Detailed Implementation

[0045] The technical solution of the present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0046] The accompanying drawings are for illustrative purposes only and are schematic diagrams, not actual images. They should not be construed as limiting the scope of this application. To better illustrate the embodiments of the present invention, some parts in the drawings may be omitted, enlarged, or reduced, and do not represent the actual dimensions of the product. It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0047] In the accompanying drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components. In the description of the present invention, it should be understood that if terms such as up, down, left, right, inside, outside, etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the terms used to describe positional relationships in the drawings are only for illustrative purposes and should not be construed as limiting the present application. For those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.

[0048] In the description of this invention, unless otherwise explicitly specified and limited, the terms "connection" or similar designations indicating the connection relationship between components should be interpreted broadly. For example, they may refer to a fixed connection, a detachable connection, or an integral part; a mechanical connection or an electrical connection; a direct connection or an indirect connection through an intermediate medium; or a connection within two components or an interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0049] Example: Reactive power regulation of distributed photovoltaic power in 10kV urban distribution network This embodiment is applied to a 10kV urban distribution network feeder, with a main feeder length of approximately 3.5km and a main line using 240mm² wire. 2The cross-linked polyethylene insulated cable has a resistance of approximately 0.08 Ω / km and a reactance of approximately 0.10 Ω / km per unit length. These line parameters have been entered into the main controller database. Twenty distributed photovoltaic inverters are connected along the feeder, with a total installed capacity of 8MW. Each inverter has a rated active power of 400kW and a rated apparent power of 500kVA. The system is equipped with a 4000kvar static var compensator installed in the middle of the feeder. The system is divided into two control zones, Zone A and Zone B, based on the feeder's power supply range. Zone A contains 12 inverters, and Zone B contains 8 inverters. The system has one main controller deployed at the substation level, and two zone group control units deployed at the box-type transformers in Zones A and B, respectively.

[0050] Step 1: Data Acquisition and Synchronization like Figure 1 As shown, 23 monitoring points are set up at the photovoltaic grid connection points, feeder start-up points, feeder end-up points, and the low-voltage side busbar of the distribution transformer. 20 of these photovoltaic grid connection points correspond to 20 inverters. Each monitoring point uses a synchronous phasor acquisition unit with a acquisition period of 50ms. Timestamp synchronization is achieved via BeiDou satellite time synchronization, with a synchronization error of less than 1ms. Acquired data includes three-phase phase voltage amplitude and phase angle, three-phase line current amplitude and phase angle, active power, and reactive power. The main controller extracts the positive-sequence components of the three-phase measurements as input for subsequent calculations to eliminate interference from three-phase imbalance. Data from each monitoring point is aggregated by the regional group control unit and then uploaded to the main controller via a 100Mbps industrial Ethernet network. The regional group control unit communicates with its assigned inverters and static var compensators using long-distance, low-power wireless communication in the 470MHz to 510MHz frequency band, in a star network configuration. Uplink telemetry and downlink commands are transmitted in a time-division multiplexing manner, with channel priority given to control commands. The communication network topology is as follows: Figure 3 As shown.

[0051] Step 2: Data Filtering and Bad Data Identification The main controller applies a moving average filter to the uploaded raw data, with a moving window length of 10 sampling points, corresponding to a 0.5s time window. The filtering formula is: ; in, Let be the filtered value at time k. These are historical sample values. This represents the length of the sliding window.

[0052] Before filtering, bad data identification is performed: if the difference between the current sampled value x(k) and the previous filtered value exceeds 20% of the corresponding rated value, the sampled point is determined to be bad data, and the previous filtered value is used instead, with the bad data count for that monitoring point being accumulated. When the same monitoring point is determined to be bad data for three consecutive sampling cycles, the system determines that the data quality of that monitoring point is abnormal, issues an alarm, and uses the most recent valid historical data for calculation. Filtering can effectively eliminate sampling noise and power spikes caused by momentary cloud cover, avoiding false triggering of control.

[0053] Step 3: Status assessment and reactive power deficit calculation Calculate the voltage deviation at each node: ; in, Let be the effective value of the filtered positive sequence voltage at node i. The system's rated voltage is 10kV.

[0054] The initial allowable range for voltage deviation is ±5%, corresponding to a voltage range of 9.5kV to 10.5kV. This range is dynamically updated every 30 minutes, with the following dynamic setting rules: the load factor is the ratio of the measured active power at the feeder head end of the area to the rated capacity of the transformer; and the photovoltaic output fluctuation rate is the maximum rate of change of the predicted sequence within 1 minute. When the load factor is greater than 70%, or the photovoltaic output forecast fluctuation rate is greater than 20% / min, the limit will be tightened to ±3%. When the load factor is between 30% and 70%, or the volatility is between 10% / min and 20% / min, maintain ±5%; When the load factor is less than 30% and the volatility is less than 10% / min, the tolerance is relaxed to ±6.5%.

[0055] The network topology and line impedance parameters are pre-stored in the main controller. For load nodes without monitoring points, their active and reactive power data are estimated based on typical daily load curves, and the estimated values ​​are proportionally corrected using the measured load rate of the region. Distribution network power flow calculation is performed based on the forward-backward substitution method, with the convergence criterion being that the change in voltage amplitude at each node is less than 0.01% between two consecutive iterations. The reactive power deficit for each region is obtained as follows: ; in, The reactive power of the total load in this area, This refers to the reactive power losses of power lines and transformers within the region. The current total reactive power output of photovoltaic inverters in the region, This refers to the current reactive power output of the static var compensator within the region.

[0056] When the voltage deviation of any node exceeds the allowable range, the main controller will only send the reactive power deficit of that area to the corresponding area group control unit to start the reactive power regulation process.

[0057] Step 4: The available reactive power margin of the inverters is calculated in real time by the regional group control unit. : ; ; in, Let j be the rated apparent power of the j-th inverter. This represents the current active power output of the j-th inverter. This is the theoretical upper limit for reactive power regulation; This is the temperature reduction factor, with values ​​taken in segments based on the internal temperature of the inverter: temperature At ℃, ; ℃ At ℃, ; At ℃, .

[0058] Taking three inverters in Area A as an example: The first inverter currently has an active power output of 300kW, and its theoretical reactive power limit is calculated by subtracting 300 from the square root of 500, which is approximately 400kvar. The measured temperature is 52℃, so the usable margin after reduction is 360kvar. The second inverter currently has an active power output of 400kW and is operating at full capacity. Its theoretical reactive power limit is 300kvar, and the measured temperature is 45℃. After reduction, the usable margin is 270kvar. The third inverter currently has an active power output of 200kW, and its theoretical reactive power limit is approximately 458kvar. The measured temperature is 35℃, and with a reduction factor of 0.97, the usable margin is approximately 444kvar.

[0059] Step 5: Generate control instructions in layers like Figure 2 As shown, the regional group control unit allocates reactive power deficit in three layers according to response speed priority. The response delay of each inverter is obtained from factory parameters or actual measurements before commissioning, and is periodically checked and updated according to operating conditions. In this embodiment, among the 12 inverters in region A, 8 have a response delay of less than 100ms and are assigned to the first layer; 4 have a response delay between 100ms and 500ms and are assigned to the second layer. The layered control logic is as follows: The first layer, the fast layer: filters response latency. The reactive power deficit is allocated to each inverter according to the proportion of available reactive power margin: ; Here, `min` represents the minimum value operation, meaning the allocation result for each inverter does not exceed its available reactive power margin. If the total available reactive power margin of the first layer is greater than or equal to the reactive power deficit of that region, an instruction is generated directly after allocation, and subsequent layers are not initiated. The allocation result is rounded to the nearest 1 kvar, and the rounded difference is made up by the inverter with the largest margin. This upper limit constraint also applies to the allocation of the second layer.

[0060] Example 1: A scenario where a single layer can meet the requirements: At a certain moment, region A experiences a voltage exceedance. Power flow calculations show a reactive power deficit of 2300 kvar, including 2500 kvar from load reactive power, 300 kvar from line and transformer reactive power losses, 200 kvar from the current reactive power output of the inverters, and 300 kvar from the current output of the static var compensator (SVC). The first layer of 8 fast inverters has a total available margin of 2880 kvar, which is greater than 2300 kvar. The reactive power is allocated proportionally based on this margin. For example, for an inverter with a available margin of 360 kvar, the target reactive power is 360 / 2880 multiplied by 2300, approximately 287 kvar, which does not exceed its margin limit. After allocation in the first layer, the deficit is cleared. The second layer and the SVC do not operate during this control cycle.

[0061] The second layer, the medium-speed layer: if there is still a remaining deficit after the compensation of the first layer. Then filter response latency The inverters are also allocated according to the available margin ratio. .

[0062] The third layer, or the bottom layer: if there is still a remaining shortfall after the first two layers of compensation. Then, a compensation command is sent to the static var compensator (SVC). The SVC uses proportional-integral closed-loop control, with the proportional coefficient set to 1.2 and the integral time constant set to 0.4s. The output is tracked for a given value until the remaining deficit is zero. The regional group control unit also sends commands and sends data back via long-distance low-power wireless communication with the static var compensators under its jurisdiction.

[0063] Example 2: Multi-level compensation scenario: During peak load periods, the reactive power deficit in region A increases to 5000 kvar. The first layer of 8 inverters has a total available margin of 2880 kvar. Because the deficit exceeds the total margin, each inverter outputs its full capacity of 2880 kvar, leaving 2120 kvar remaining. The second layer of 4 inverters has a total available margin of 1440 kvar, which is fully allocated proportionally, meaning each inverter outputs according to its own margin, leaving 680 kvar remaining. The third layer uses a static var compensator (SVC) with a setpoint of 680 kvar for proportional-integral tracking, reaching steady state within approximately 150 ms. Under this condition, the SVC only handles 13.6% of the total deficit, effectively reducing the depth and frequency of its operation.

[0064] Step 6: Instruction Execution and Dynamic Termination The control command frame is sent by the group control unit via a long-range low-power wireless channel, and includes the device address, target reactive power value, command execution duration, command sequence number, and checksum. In this embodiment, the command execution duration is set to 800ms. After the command is sent, the system continuously monitors the voltage status. If the voltage of all nodes recovers to the allowable range within 800ms, the current control is terminated early, and each device maintains its current reactive power output value, waiting for the next control cycle. If the voltage has not recovered to the allowable range after the command execution period expires, the system recalculates the deficit based on the newly acquired data and starts the next control cycle until the voltage recovers or reaches the device's adjustment limit, at which point a device-level alarm is triggered.

[0065] Step 7: Steady-state optimization mode When the system voltage remains within the allowable range, it enters steady-state optimization mode. With the goal of achieving a power factor of 0.98 at the beginning of the system feeder and minimizing active power loss, the reactive power distribution of each inverter is optimized using an optimal power flow algorithm. The static var compensator operates in standby mode, with its output reactive power approaching zero. The optimization objective function is to minimize the active power loss of the entire feeder: ; in Let i be the current in the i-th segment of the line. Let be the resistance of the i-th line segment; constraints include: the voltage deviation at each node does not exceed the allowable range of the current voltage deviation, the reactive power output of each inverter does not exceed its available reactive power margin, and the power factor at the feeder head is between 0.95 and 1.0. The optimization cycle is 1 minute, and the interior point method is used for solution. This mode ensures that the power factor meets the grid assessment standards while avoiding additional losses caused by reactive circulating current.

[0066] Step 8: Self-calibration mechanism After each round of adjustment, the system records the target reactive power value and the actual output value of each device, and calculates the output deviation rate: ; In the formula: Output deviation rate, in % : Actual reactive power output value of the device, kvar; Target reactive power of the equipment, kvar; : Rated reactive power capacity of the corresponding device, kvar.

[0067] If the deviation rate is greater than 10% for three consecutive control cycles, gain calibration is triggered: the reactive power control gain coefficient of the equipment is corrected based on the statistical average of historical deviations. : ; After calibration, control commands are issued according to the new gain coefficient to compensate for output errors caused by equipment aging and parameter drift. After gain calibration is completed, the continuous deviation count of the device is reset to zero; if the deviation rate is still greater than 10% for three consecutive control cycles after calibration, the device is judged to be in control abnormality, the system issues a maintenance alarm, and the device is taken out of automatic control and switched to local control mode.

[0068] Step 9: Local control of communication failures If the inverter fails to receive commands from the group control unit for three consecutive acquisition cycles, it is determined to be a communication interruption and automatically switches to local control mode. Local droop control characteristics are as follows: Figure 4 As shown, local control utilizes the voltage-reactive power droop characteristic: Reference voltage range: 9.4kV to 10.6kV, which is 94% to 106% of the rated voltage; Sag factor: 4%, which is 1% of the rated voltage deviation and 4% of the rated reactive power output adjustment capacity. When the voltage is between 9.4kV and 10.6kV, the reactive power output remains constant; when the voltage is above 10.6kV, it absorbs inductive reactive power, and when it is below 9.4kV, it generates capacitive reactive power. The reactive power change is linearly related to the voltage deviation.

[0069] Specific calculation example: An inverter has a rated apparent power of 500kVA and a droop factor of 4%. Using the rated apparent power of 500kVA as the per-unit benchmark for droop adjustment, the corresponding reactive power capacity is 500kvar. When the measured grid connection voltage is 10.8kV, which is 108% of the rated voltage, it exceeds the upper limit of the reference voltage by 0.2kV, i.e., a deviation of 2%. The inverter absorbs 2 multiplied by 4% multiplied by 500kvar, equaling 40kvar. If the voltage further rises to 11.0kV, which is 110% of the rated voltage, with a deviation of 4%, the absorption increases linearly to 80kvar. When the grid connection voltage drops to 9.2kV, which is 92% of the rated voltage, it is 2% below the lower limit of the reference voltage, and the inverter generates 40kvar of capacitive reactive power. If the voltage further drops to 9.0kV, which is 90% of the rated voltage, with a deviation of 4%, it generates 80kvar of capacitive reactive power, supporting voltage recovery.

[0070] Once communication is restored, the inverter automatically switches back to centralized control mode. At the moment of switching back, the current local output value is used as the initial value of the centralized control command, resulting in a smooth transition without any impact.

[0071] Implementation effect verification Simulation results show that this embodiment can control voltage deviation within the set range under conditions such as sudden changes in sunlight (30% fluctuation in photovoltaic output within 1 second) and sudden load increases. The maximum response delay of the first-layer fast control path is less than 180ms. Compared with the traditional centralized automatic voltage control method, the system network loss is expected to be reduced by 9% to 13%, the number of static var compensator (SVC) operations is expected to be reduced by more than 40%, and there are no voltage limit violations in communication interruption scenarios. The response delay advantage comes from the fast layer priority call and dynamic command termination mechanism. The reduction in network loss comes from the full release of the inverter's reactive power potential and steady-state optimized operation through temperature reduction margin. The reduction in the number of SVC operations comes from the priority call of inverter capacity through time-delay hierarchical compensation. The absence of voltage limit violations in communication interruption scenarios comes from the fault-tolerant design of local interlocking switching.

[0072] It is worth noting that the above performance data are the expected results of simulation calculations, and the actual results will vary depending on the power grid parameters and operating conditions.

[0073] In summary, the working principle of the reactive power optimization and control method for the distributed photovoltaic power station distribution system provided in this embodiment is as follows: The system consists of a main controller, regional group control units, distributed photovoltaic inverters, and static var compensators. The main controller is deployed in the substation and is responsible for global data aggregation, voltage judgment, and power flow calculation. Each control area has a regional group control unit, which is responsible for the management of equipment and command distribution within its area. The inverters and static var compensators serve as terminal execution devices.

[0074] Each monitoring point periodically collects three-phase voltage, current, and active power, which are then uploaded after unified time synchronization. The main controller first filters the data and identifies bad data, then calculates the voltage deviation at each node to determine if it exceeds the limit, and simultaneously calculates the reactive power deficit for each region through power flow calculation. When the voltage exceeds the limit, the main controller only sends the deficit corresponding to the region exceeding the limit to the group control unit of that region. The group control unit first calculates the available reactive power margin for each inverter. The margin is obtained by subtracting the current active power from the apparent power and then reducing it according to the operating temperature. Then, the deficit is allocated hierarchically according to the response delay: the inverter with the fastest response takes the lead, and the allocation is based on the margin ratio, with the allocation not exceeding the margin; if it is still insufficient, the inverter with a slightly slower response makes up the difference according to the same rules; if it is still insufficient, the static var compensator compensates until the deficit is zero. After the command is issued, the system continues to monitor the voltage. Once it recovers to the allowable range, the control ends early and the current state is maintained. When the voltage is normal, the system enters steady-state optimization, adjusting the output with power factor and line loss as the target. After each round of adjustment, the deviation is recorded. If the deviation remains large, the gain is automatically corrected. When communication is interrupted, the inverter switches to local control and automatically adjusts reactive power according to the local voltage. After communication is restored, it returns to centralized control.

[0075] Usage: Follow these steps. First, divide the control area according to the feeder power supply range. Deploy one regional group control unit in each area. The main controller is deployed in the substation, and monitoring points are set at the photovoltaic grid connection point, the beginning and end of the feeder, and the low-voltage busbar of the distribution transformer. Establish a long-distance low-power wireless network between the regional group control unit, the inverter, and the static var compensator. Connect the main controller and the group control unit via wired Ethernet. Before commissioning, measure the response delay of each inverter and record it in the system. Set the allowable voltage deviation range and dynamic update cycle according to the system voltage level. Set two thresholds for response delay stratification and configure the droop coefficient and reference voltage range for local control. After the system is commissioned, it will automatically operate according to the process of data acquisition, judgment, stratified allocation, execution, and verification. Maintenance personnel only need to check the data quality and equipment status when the voltage exceeds the limit for a long time or the equipment alarms, and recalibrate the gain as needed.

[0076] The response delay parameters of each distributed photovoltaic inverter are obtained through step response testing before commissioning and are checked and updated every 3 months during operation. When the inverter control parameters are adjusted or the firmware is upgraded, the response delay needs to be retested and the system database updated.

[0077] In summary, this application has the following beneficial effects: Because the deficit is only issued to areas where the voltage exceeds the limit, while remaining inactive in other areas, the regulation is highly targeted and the communication burden is low. Since the margin is calculated based on active power output and operating temperature reductions, the allocation commands always fall within the adjustable range of the equipment, preventing overload or overheating protection and fully utilizing the inverter's regulation capabilities. Because resources are allocated layer by layer from fastest to slowest response time, fast resources first smooth out instantaneous fluctuations, while slower resources and compensation devices then supplement the steady-state deficit, regulation keeps pace with fluctuations while maintaining capacity margin. Because commands terminate early and remain in place when the voltage recovers, regulation avoids over-adjustment, and voltage oscillations are less likely. The allowable voltage range is dynamically adjusted according to the load factor and photovoltaic volatility, narrowing when the load is heavy and fluctuations are large, and widening when the load is stable, balancing voltage safety and regulation frequency. Since the inverter handles most of the reactive power regulation, the operation of the static var compensator is reduced, thus decreasing equipment wear and maintenance costs. Steady-state optimization aims to achieve the power factor target and minimize line loss, improving operational economy. Gain self-calibration compensates for deviations caused by equipment aging, maintaining long-term operational accuracy. When communication is interrupted, the inverter adjusts the voltage locally; when communication is normal, it controls and blocks the circuit locally. The two modes do not interfere with each other, and the system can still maintain voltage under communication failure.

[0078] The above are merely preferred embodiments of the present invention and are not intended to limit the implementation methods and protection scope of the present invention. Those skilled in the art should recognize that any equivalent substitutions and obvious changes made based on the description and illustrations of the present invention should be included within the protection scope of the present invention.

Claims

1. A reactive power optimization and control method for a distributed photovoltaic power station distribution system, characterized in that, The power distribution system is divided into multiple control zones, each with a regional group control unit, which is uniformly coordinated by the main controller, forming a three-level control architecture consisting of the main controller, regional group control units, distributed photovoltaic inverters, and static var compensators. The method includes the following steps: The system collects three-phase voltage, current, and active power data from each monitoring point in the power distribution system in real time, and uploads them to the main controller after synchronization with timestamps. The voltage deviation value of each node is calculated based on the collected data, and it is determined whether the voltage deviation value exceeds the preset voltage deviation allowable range. At the same time, the reactive power deficit of each area of ​​the power distribution system is determined based on power flow calculation. When the judgment result is that the voltage deviation exceeds the allowable range, the main controller only sends the reactive power deficit corresponding to the voltage over-limit area to the corresponding area group control unit. The area group control unit determines the available reactive power margin of each inverter based on the current active power output and operating temperature of each distributed photovoltaic inverter under its jurisdiction, and generates reactive power control instructions in a hierarchical manner according to the available reactive power margin and the time delay characteristics from instruction reception to response completion, in order of priority from high to low: distributed photovoltaic inverters with response delay less than the first time delay threshold are given priority for allocation, and the reactive power deficit is allocated according to the proportion of the available reactive power margin of each inverter, and the allocation result of each inverter does not exceed its available reactive power margin; the remaining reactive power deficit is allocated by distributed photovoltaic inverters with response delay between the first time delay threshold and the second time delay threshold according to the proportion of available reactive power margin, and the allocation result of each inverter does not exceed its available reactive power margin; the remaining reactive power deficit is supplemented to zero by the output of the static var compensator. The reactive power control command is sent to each distributed photovoltaic inverter and static var compensator through the communication network to adjust its reactive power output and maintain the system voltage within the allowable voltage deviation range. When the distributed photovoltaic inverter detects a communication interruption with the regional group control unit, it automatically switches to local control mode and adjusts the reactive power output in real time using the voltage-reactive power droop characteristic curve based on the measured voltage value of its own grid connection point. Under normal communication conditions, the local control mode is in a locked state.

2. The reactive power optimization and control method for a distributed photovoltaic power station distribution system according to claim 1, characterized in that, The method for calculating the available reactive power margin is as follows: Based on the apparent power rating of each distributed photovoltaic inverter under the current active power output, the capacity occupancy corresponding to the current active power is deducted to obtain the theoretical upper limit of reactive power adjustment. The current operating temperature of the inverter is collected, and the theoretical adjustment upper limit is corrected based on the preset temperature-reduction coefficient piecewise function to obtain the available reactive power margin. The piecewise function of the temperature reduction factor is as follows: when the temperature is below 40℃, the reduction factor is 95% to 100%; when the temperature is between 40℃ and 60℃, the reduction factor is 85% to 95%; and when the temperature is above 60℃, the reduction factor is 70% to 85%.

3. The reactive power optimization and control method for a distributed photovoltaic power station distribution system according to claim 1, characterized in that, The first delay threshold is 100ms, and the second delay threshold is 500ms; the hierarchical generation of reactive power control instructions specifically includes: First-level regulation: Select distributed photovoltaic inverters with available reactive power margin greater than 0 and response delay less than 100ms to take priority in reactive power output regulation; Second-level regulation: Select distributed photovoltaic inverters with response delays between 100ms and 500ms, and allocate the remaining reactive power deficit according to the proportion of available reactive power margin; The third level of control: The regional group control unit controls the static var compensator to output supplementary reactive power until the remaining reactive power deficit is zero.

4. The reactive power optimization and control method for a distributed photovoltaic power station distribution system according to claim 1, characterized in that, The reactive power control command includes a target reactive power value and a command execution duration, wherein the command execution duration is set to 500ms to 2000ms; After the command is issued, the system continuously collects data and recalculates the voltage deviation and reactive power deficit. If the voltage deviation is detected to recover to the allowable range before the execution time of the command ends, the execution of the current control command is terminated in advance, and the current reactive power output state of each device is maintained until the next round of control command is issued.

5. The reactive power optimization and control method for a distributed photovoltaic power station distribution system according to claim 1, characterized in that, The communication network adopts a hierarchical architecture that combines wired Ethernet with long-range low-power wireless communication. Wired Ethernet is used for data transmission and command interaction between the main controller and the group control units in each area; Long-range low-power wireless communication is used for command issuance and operation data feedback between the regional group control unit and its subordinate distributed photovoltaic inverters and static var compensators.

6. The reactive power optimization and control method for a distributed photovoltaic power station distribution system according to claim 1, characterized in that, Before calculating voltage deviation and reactive power deficit, the process also includes filtering the collected three-phase voltage, current and active power data. The filtering method uses a moving average filtering method with a sliding window length of 5 to 20 sampling points and a time span of 0.1s to 1s to eliminate instantaneous interference noise.

7. The reactive power optimization and control method for a distributed photovoltaic power station distribution system according to claim 1, characterized in that, The monitoring points in the power distribution system include the photovoltaic grid connection point, the beginning of the feeder, the end of the feeder, and the low-voltage side busbar of the power distribution transformer; The data acquisition cycle at each monitoring point is 20ms to 50ms. All collected data are uploaded after being synchronized with timestamps through a unified clock.

8. The reactive power optimization and control method for a distributed photovoltaic power station distribution system according to claim 1, characterized in that, When the voltage deviation value does not exceed the allowable range of voltage deviation, the system enters steady-state optimization mode: With the goal of maintaining the system power factor in the range of 0.95 to 1.0 and minimizing the active power loss of the lines, the reactive power output of each distributed photovoltaic inverter is optimized and the static var compensator is controlled to operate in a low-loss standby state.

9. The reactive power optimization and control method for a distributed photovoltaic power station distribution system according to claim 1, characterized in that, After each reactive power control command is executed, the system records the actual reactive power output value of each distributed photovoltaic inverter and static var compensator at the current moment, compares it with the target reactive power value, and calculates the output deviation rate. When the target reactive power value is zero, the output deviation rate is calculated with the rated reactive power capacity of the corresponding device as the denominator. If the absolute value of the output deviation rate exceeds 10% for three consecutive control cycles, the gain calibration process is triggered: the reactive power control gain coefficient of each device is corrected according to historical deviation data so that the actual output tracks the target value.

10. The reactive power optimization and control method for a distributed photovoltaic power station distribution system according to claim 1, characterized in that, The preset allowable voltage deviation range is dynamically set within the range of 3% to 7% of the rated voltage; The settings are based on the current system load factor, the predicted photovoltaic output, and the load conditions of adjacent feeders. The setting rules are: the higher the load factor, the greater the fluctuation of photovoltaic output, and the smaller the allowable range of voltage deviation. The dynamic update step is once every 15 minutes to 60 minutes.