Data acquisition and control system of photovoltaic grid-connected inverter

Abstract

The invention relates to the technical field of photovoltaic grid-connected control, and discloses a data acquisition and control system of a photovoltaic grid-connected inverter. The system comprises an operation environment modeling module which constructs an irradiance discrete state matrix by associating the real-time irradiance of a photovoltaic array, the temperature of a component and voltage and current sampling values at the output end of an inverter; the electrical topology generation module divides a grid-connected node cluster and generates a topology path set based on the identifier and the monitoring point coordinate in the matrix; the fluctuation characteristic analysis module calls the same cluster data and quantifies the response characteristics of the inverter under the fluctuation of the power grid; the strategy control module identifies abnormal points of cross-cluster groups to form an abnormal operation point set; and the coordination output module marks nodes needing to be controlled and generates a cluster coordination control instruction set. The system can comprehensively analyze the influence of environment and power grid factors, achieves the precise coordination control of the inverter cluster, and improves the operation performance of a photovoltaic power generation system.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic grid-connected control technology, specifically to a data acquisition and control system for a photovoltaic grid-connected inverter. Background Technology

[0002] As the global energy structure shifts towards clean energy, photovoltaic (PV) power generation, as an important form of renewable energy utilization, has seen its installed capacity grow rapidly. As a key device connecting PV arrays to the power grid, the operating status of the PV grid-connected inverter directly affects the efficiency, stability, and grid security of the PV power generation system.

[0003] In actual operation, photovoltaic arrays face complex and variable environmental conditions, such as drastic fluctuations in irradiance and uneven distribution of module temperature. These factors cause nonlinear changes in the inverter's input power. Simultaneously, frequency fluctuations and voltage distortions on the grid side also have a feedback effect on the inverter, further exacerbating its operational instability. Traditional photovoltaic grid-connected inverter control systems often employ single-parameter monitoring and fixed control strategies, making it difficult to comprehensively capture the coupled influence of environmental and grid factors, resulting in insufficient control accuracy and response lag.

[0004] The coordinated operation of inverter clusters in large-scale photovoltaic (PV) power plants faces challenges. Due to the different locations of the inverters within the PV array, their received irradiance and module temperatures vary significantly, resulting in inconsistent output characteristics. When grid disturbances occur, inverter clusters lacking coordination mechanisms may exhibit uncoordinated responses, leading to power surges, harmonic amplification, and other problems, affecting the safe and stable operation of the grid. Current technologies often limit inverter operation monitoring to local parameter acquisition, lacking comprehensive modeling and analysis of environmental factors and grid conditions. This hinders precise coordinated control of inverter clusters and restricts the overall performance improvement of PV power generation systems. Summary of the Invention

[0005] The purpose of this invention is to provide a data acquisition and control system for a photovoltaic grid-connected inverter to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides a data acquisition and control system for a photovoltaic grid-connected inverter, the system comprising:

[0007] The runtime environment modeling module acquires real-time irradiance data and component temperature data at each monitoring point of the photovoltaic array, correlates the voltage and current sampling values ​​at the inverter output terminal, divides the discrete state intervals of irradiance and matches the corresponding temperature intervals, and constructs the discrete state matrix of irradiance.

[0008] The electrical topology generation module extracts the discrete state identifiers and monitoring point coordinates from the irradiance discrete state matrix, identifies the electrical distance relationship between adjacent monitoring points, divides grid-connected nodes into clusters based on voltage phase differences, and generates a set of inverter grid-connected point topology paths.

[0009] The fluctuation characteristic analysis module calls the monitoring points in the same cluster topology path of the inverter grid connection point topology path set, collects historical output power time series and grid frequency data, compares the fluctuation deviation with the inverter rated capacity, quantifies the inverter response characteristics under grid fluctuations, and generates fluctuation response characteristic analysis results.

[0010] The strategy control module identifies abnormal points in the monitoring points whose response intensity is greater than the grid tolerance threshold and are located in the cross-cluster topology path in the fluctuation response characteristic analysis results, extracts the power fluctuation characteristics and frequency offset data of the abnormal points, and forms a set of abnormal operation points of the inverter.

[0011] The coordination output module acquires all monitoring points and corresponding electrical parameters from the set of abnormal inverter operation points, marks the inverter nodes that need to be coordinated and controlled, and generates a set of coordinated control instructions for the inverter cluster.

[0012] Preferably, the runtime environment modeling module includes:

[0013] The multi-dimensional parameter acquisition submodule obtains the geographical coordinates and timestamps of the photovoltaic array monitoring points, and simultaneously collects the real-time irradiance value, module temperature value, inverter output voltage value and output current value at the coordinate location, and integrates them into a four-dimensional parameter data group.

[0014] The state discretization submodule, based on the irradiance data and temperature data in the four-dimensional parameter data group, divides the state intervals into preset order of magnitude, maps the voltage and current sampled values ​​to the corresponding state intervals, calculates the average voltage and current values ​​in the same state interval, and generates an irradiance discrete state matrix.

[0015] Preferably, the electrical topology generation module includes:

[0016] The phase difference calculation submodule extracts the discrete state identifier and monitoring point location information from the irradiance discrete state matrix, sorts adjacent monitoring point pairs based on the shortest electrical distance, measures the output voltage phase difference between each pair of monitoring points, and generates a phase difference sequence of adjacent monitoring points.

[0017] The topology cluster partitioning submodule sets a phase difference tolerance threshold based on the phase difference sequence of adjacent monitoring points. Monitoring point pairs below the threshold are marked as nodes in the same cluster, and monitoring point pairs exceeding the threshold are marked as nodes crossing the cluster boundary. The topology path is divided according to the cluster boundary to generate a set of inverter grid connection point topology paths.

[0018] Preferably, the fluctuation characteristic analysis module includes:

[0019] The power sequence extraction submodule filters the monitoring points of the same cluster topology path in the inverter grid connection point topology path set, extracts the inverter output active power sequence and grid frequency sampling sequence within a continuous time window, and generates a power frequency time series dataset by aligning them in time.

[0020] The response feature quantization submodule calculates the power fluctuation rate and frequency offset rate per unit time based on the power frequency time series dataset, compares them with the allowable fluctuation range under the inverter's rated capacity, identifies fluctuation periods that exceed the allowable range and records the response over-limit intensity, and generates fluctuation response feature analysis results.

[0021] Preferably, the strategy control module includes:

[0022] The over-limit node identification submodule detects monitoring points whose response intensity is greater than the grid tolerance threshold in the fluctuation response feature analysis results, as well as monitoring points located in cross-cluster topology paths, extracts the extreme values ​​of power fluctuation and frequency offset during abnormal periods, and generates an abnormal operation feature dataset.

[0023] The dynamic adjustment submodule calculates the consistency between the power fluctuation direction and the frequency offset direction based on the abnormal operation feature dataset. When the changes in both directions are inconsistent, it is marked as a coordination control priority node. All priority node numbers and parameters are integrated to form a set of abnormal operation points of the inverter.

[0024] Preferably, the coordination output module includes:

[0025] The control strategy generation submodule obtains the monitoring point number and electrical parameters of the inverter abnormal operation point set, calculates the weight ratio of the power fluctuation of each node to the total capacity of the cluster, and allocates the output power adjustment amount according to the weight ratio.

[0026] The instruction synthesis submodule generates an inverter output power correction instruction set based on the allocated power adjustment amount, combined with grid dispatch instructions and local load requirements.

[0027] Preferably, the runtime environment modeling module further includes:

[0028] The irradiance fluctuation feature submodule obtains the temporal distribution pattern of historical irradiance data, identifies the occurrence periods of sudden increases and decreases in irradiance, and establishes a mapping relationship table between time periods and irradiance fluctuation intensity.

[0029] The fluctuation characteristic analysis module calls the mapping table to perform fluctuation characteristic quantization on the inverter output power sequence corresponding to the time period.

[0030] Preferably, the electrical topology generation module further includes:

[0031] The voltage phase calibration submodule acquires the voltage phase reference value at the power grid point of common coupling and calculates the deviation between the output voltage phase of each monitoring point and the reference value.

[0032] The topology cluster partitioning submodule uses the deviation value instead of the output voltage phase difference to perform cluster partitioning.

[0033] Preferably, the strategy control module further includes:

[0034] The wave propagation analysis submodule detects the direction of power wave fluctuations at monitoring points across cluster boundaries and tracks the propagation path of waves in the topology.

[0035] The dynamic adjustment submodule adds a preventative adjustment coefficient to the downstream nodes based on the transmission path.

[0036] Preferably, the coordination output module further includes:

[0037] The optimization strategy feedback submodule collects actual power fluctuation convergence data after command execution and updates the power adjustment weight ratio under the same scenario.

[0038] The instruction synthesis submodule generates an inverter output power correction instruction set based on the updated weight ratio.

[0039] Compared with the prior art, the beneficial effects of the present invention are:

[0040] By using the operating environment modeling module, the real-time irradiance and component temperature of the photovoltaic array are correlated with the voltage and current sampling values ​​at the inverter output. Discrete state intervals are divided and a matrix is ​​constructed, enabling a comprehensive capture of the coupling relationship between environmental factors and electrical parameters. This provides multi-dimensional foundational information for subsequent analysis and control. This multi-parameter correlation modeling method overcomes the limitations of traditional single-parameter monitoring and can more accurately reflect the complex operating environment of the inverter.

[0041] The electrical topology generation module identifies electrical distance relationships and clusters grid-connected nodes based on the identifiers and monitoring point coordinates in the discrete state matrix, generating a set of topology paths. This helps clarify the electrical connection characteristics and spatial distribution patterns of each node in the inverter cluster. Clustering allows inverters with similar operating characteristics to be grouped together, creating conditions for subsequent targeted analysis and control, and enabling the system to better adapt to the complex structure of large-scale inverter clusters.

[0042] The fluctuation characteristic analysis module calls upon monitoring point data from the same cluster topology path, combining historical output power time series, grid frequency, and inverter rated capacity to quantify response characteristics under fluctuations. This allows for in-depth analysis of the operating patterns of inverters from different clusters during grid fluctuations. This cluster-based analysis method avoids analytical biases caused by individual inverter differences, more accurately grasps the response characteristics of similar inverters, and provides a reliable basis for identifying abnormal responses.

[0043] The strategy control module identifies anomalies where the response intensity exceeds a threshold and is located on a cross-cluster path. It extracts power fluctuation characteristics and frequency offset data to form a set of anomalous operating points, enabling precise location of key nodes affecting system stability. By focusing on cross-cluster anomalies, potential incoordination issues between different clusters can be detected in a timely manner, clarifying the objectives for subsequent coordinated control.

[0044] The coordination output module acquires the monitoring points and electrical parameters of the anomaly point set, marks the nodes that need to be controlled, and generates a cluster coordination control command set, realizing targeted regulation of the inverter cluster. This anomaly-based coordination control enables the inverter cluster to form a coordinated response when the grid fluctuates or the environment changes, reducing system fluctuations caused by individual incoordination, improving the overall operational stability, and adapting to complex operational needs under different environmental and grid conditions, thus enhancing the compatibility of the photovoltaic power generation system with the grid. Attached Figure Description

[0045] Figure 1 This is a timing diagram of the data acquisition and control system for the photovoltaic grid-connected inverter described in this invention;

[0046] Figure 2 A flowchart for the runtime environment modeling module;

[0047] Figure 3 A flowchart for the electrical topology generation module;

[0048] Figure 4 A flowchart for coordinating the output module;

[0049] Figure 5 A flowchart for modeling and extending the runtime environment. Detailed Implementation

[0050] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] Please see Figure 1This invention provides a data acquisition and control system for a photovoltaic grid-connected inverter, the system comprising: an operating environment modeling module, an electrical topology generation module, a fluctuation characteristic analysis module, a strategy control module, and a coordination output module.

[0052] The operating environment modeling module collects real-time data such as irradiance, component temperature, and inverter output voltage and current from photovoltaic array monitoring points, divides and matches discrete state intervals of irradiance and temperature, and constructs a discrete state matrix of irradiance.

[0053] Based on the discrete state identifiers and monitoring point coordinates in the matrix, the electrical topology generation module identifies the electrical distance relationship between adjacent monitoring points, divides grid-connected nodes into clusters according to voltage phase differences, and generates a set of inverter grid-connected point topology paths.

[0054] The fluctuation characteristic analysis module calls the historical output power and grid frequency data of the monitoring points on the same cluster topology path, combines the inverter rated capacity to quantify the fluctuation deviation, and generates the fluctuation response characteristic analysis results; the strategy control module identifies the abnormal points in the analysis results whose response intensity exceeds the threshold and are located on cross-cluster paths, extracts their power fluctuation and frequency offset characteristics, and forms an abnormal operation point set;

[0055] The coordination output module marks the inverter nodes that need to be coordinated and controlled based on the monitoring points and electrical parameters of the abnormal operation point set, and generates a cluster coordination control instruction set.

[0056] Example 1: See Figure 2 The operating environment modeling module comprises a multi-dimensional parameter acquisition submodule and a state discretization submodule. Its implementation revolves around the acquisition, processing, and state classification of environmental and electrical parameters at the photovoltaic array monitoring points. The multi-dimensional parameter acquisition submodule utilizes a sensor network deployed at each monitoring point of the photovoltaic array to synchronously acquire geographic, temporal, and environmental electrical parameters. Geographic coordinate information is acquired by a GPS module mounted on the monitoring point support, achieving sub-meter accuracy, and is used to identify the specific location of each monitoring point. Timestamps are generated by a high-precision clock synchronization unit built into the module, aligned with the satellite timing system to ensure time consistency of data from different monitoring points, achieving millisecond-level time resolution. Irradiance data is acquired through a total radiation meter mounted on the component plane. This sensor employs the thermopile principle and can measure the solar radiation power received per unit area in real time, with a range covering 0 to 1500 W / m². 2The output frequency is 1Hz. Component temperature data is acquired via thermocouple sensors attached to the back of the photovoltaic module. The thermocouples are of type K material, with a response time of less than 10 seconds and a measurement range of -40℃ to 120℃. The output frequency is consistent with that of the irradiance sensor. Inverter output voltage and current data are collected by electronic current transformers installed on the DC side of the inverter. The voltage sensor has a range of 0 to 1500V, and the current sensor has a range of 0 to 200A. Both use fiber optic signal transmission to avoid electromagnetic interference, with a sampling frequency of 10kHz. After being converted into digital signals by a high-speed analog-to-digital converter (ADC) inside the module, the signals are averaged and filtered at a frequency of 1Hz to obtain valid sampled values. All collected data is transmitted to the module's data processing unit via wired or wireless means. Geographic coordinates, timestamps, and irradiance and temperature data are integrated into one group, while inverter output voltage and current data are integrated into another group, ultimately forming a four-dimensional parameter data set. Each data set contains information in four dimensions: timestamp, geographic coordinates, irradiance-temperature correlation value, and voltage-current sampled value.

[0057] The state discretization submodule, based on irradiance and temperature data from the multidimensional parameter data set, completes the division of state intervals and the statistical analysis of corresponding electrical parameters. According to the historical operating data of the photovoltaic array and local climate characteristics, the number and range of irradiance and temperature state intervals are preset. For example, the irradiance state intervals can be divided into three levels: [0, 200), [200, 500), and [500, 1000], corresponding to "low irradiance," "medium irradiance," and "high irradiance" states, respectively; the temperature state intervals can be divided into three levels: [0, 25), [25, 40), and [40, 60], corresponding to "low temperature," "medium temperature," and "high temperature" states, respectively. Each state interval is uniquely identified by a start value and an end value, and the interval boundaries are dynamically adjusted according to the distribution characteristics of the actual data. For example, if the irradiance frequently occurs between 300-400 W / m² within a certain period... 2 Within this range, the interval can be further refined to [200, 300), [300, 400), or [400, 500) to improve the accuracy of the division. For real-time acquired irradiance and temperature data, by comparing their values ​​with the boundary values ​​of the preset interval, they are mapped to the corresponding state interval. For example, if the irradiance value is 850 W / m², the interval can be further refined to [200, 300), [300, 400), or [400, 500) to improve the accuracy of the division. 2 When the temperature is 32℃, it matches the range [500, 1000], and when the temperature is 32℃, it matches the range [25, 40]. The combination forms the status range identifier "high irradiance - medium temperature".

[0058] After completing the state interval mapping, the submodule performs statistical processing on the inverter output voltage and current sampling values ​​of all monitoring points under the same state. It filters out all monitoring point data belonging to the same state interval, such as all monitoring points with irradiance between [500, 1000] and temperature between [25, 40]. The arithmetic mean of the voltage sampling values ​​of these monitoring points is calculated to eliminate the influence of random noise. For example, if the voltage values ​​of 10 monitoring points under a certain state are 305V, 308V, 312V, etc., the average voltage is approximately 309V. Similarly, the average current is calculated; for example, if the current values ​​of 10 monitoring points are 5.1A, 5.3A, 5.0A, etc., the average current is approximately 5.15A. The state interval identifier is associated with the corresponding average voltage and average current to generate an irradiance discrete state matrix. This matrix has a row-column structure, where rows represent different combinations of state intervals, and columns represent information such as the state interval identifier, average voltage, and average current. Each element in the matrix completely records the typical environmental and electrical parameters of the photovoltaic array under a certain state, providing a basis for state division for the subsequent electrical topology generation module.

[0059] Throughout the process, the multidimensional parameter acquisition submodule and the state discretization submodule communicate via a data bus to ensure real-time acquisition and processing. The acquisition submodule performs real-time detection of abnormal data, such as when the irradiance sensor output value exceeds the upper limit of its range (e.g., 1500W / m). 2 When a missing value is found, it is marked as invalid, and a supplementary sampling mechanism is triggered to fill the missing value through linear interpolation of adjacent time points. The state discretization submodule verifies the validity of the divided state intervals. If there is no monitoring point data in a certain state interval (e.g., no "high irradiance" state under extreme weather), the interval is marked as empty in the matrix to avoid invalid data interfering with subsequent analysis. Through the above process, the operating environment modeling module finally constructs an irradiance discrete state matrix containing information on the correlation between environmental and electrical states, realizing a quantitative description of the photovoltaic array's operating environment.

[0060] Example 2: See Figure 3 The electrical topology generation module, a key component of the system, is responsible for constructing a topology that reflects the electrical connection characteristics of the photovoltaic array based on the discrete irradiance state matrix generated by the operating environment modeling module. This provides a topological basis for subsequent fluctuation characteristic analysis and strategy control. The module comprises a phase difference calculation submodule and a topology cluster partitioning submodule, which work together to complete the conversion from state data to topology paths.

[0061] The phase difference calculation submodule extracts key input information from the discrete state matrix of irradiance, including discrete state identifiers and monitoring point location information. The discrete state identifier is a combination of irradiance and temperature ranges, used to identify the environmental state category of the monitoring point; the monitoring point location information is geographic coordinates (latitude and longitude), accurately reflecting the physical distribution of each monitoring point in the photovoltaic array. Based on this information, the submodule needs to identify adjacent monitoring point pairs, where "adjacent" is defined as a direct electrical connection or geographical proximity. For example, in a series-connected photovoltaic module string on the DC side, adjacent string monitoring points are connected through a DC combiner box; in a parallel-connected inverter cluster on the AC side, adjacent inverter monitoring points are connected through an AC bus. For geographically adjacent monitoring points, even if not directly electrically connected, they must be included in the evaluation scope of adjacent monitoring point pairs to comprehensively reflect the electrical coupling characteristics of the array.

[0062] After identifying adjacent monitoring point pairs, the submodule needs to measure the phase difference of the output voltage of these pairs. To achieve high-precision phase measurement, the system adopts synchronous phasor measurement technology: the voltage signals of all monitoring points are acquired through high-precision electronic current transformers and transmitted to the synchronous clock unit via optical fiber, ensuring that the sampling time of each monitoring point is strictly aligned (time error less than 1 microsecond). Digital signal processing algorithms are used to extract the amplitude and phase angle of the voltage phasor (the phase angle is referenced to the grid's point of common coupling voltage, ranging from -180° to +180°), and the phase difference of the output voltage of each pair of adjacent monitoring points (i.e., the absolute value of the difference in phase angles between the two points) is calculated. For example, if the voltage phase of monitoring point A is 30° and the phase of monitoring point B is 25°, then the phase difference between them is 5°. For multiple pairs of adjacent monitoring points, the submodule sorts them according to the principle of shortest electrical distance, generating a sequence of phase differences between adjacent monitoring points. This sequence contains the monitoring point pair identifier and the corresponding phase difference value, comprehensively recording the electrical phase correlation characteristics between nodes in the photovoltaic array.

[0063] The topology clustering submodule, based on the phase difference sequence output by the phase difference calculation submodule, completes the clustering of the photovoltaic array. To achieve reasonable clustering, the system pre-sets a phase difference tolerance threshold. The submodule traverses each pair of monitoring points in the phase difference sequence and determines the relative magnitude of its phase difference value to the threshold: if the phase difference value of a pair of monitoring points is less than or equal to the threshold, it is determined that the two belong to the same cluster; if it is greater than the threshold, it is determined to be a cross-cluster boundary node pair. For example, if the phase difference between monitoring points A and B is 5° (≤ threshold), the phase difference between B and C is 7° (≤ threshold), and the phase difference between C and D is 12° (> threshold), then A, B, and C belong to the same cluster, D belongs to another cluster, and the connection between C and D is a cross-cluster boundary.

[0064] After completing the cluster labeling of all monitoring point pairs, the submodule divides the photovoltaic array into regions based on the location across cluster boundaries. Monitoring points within each cluster region are connected via low-phase-difference nodes, forming a tightly coupled electrical subsystem; clusters are separated by high-phase-difference nodes, exhibiting weaker electrical correlation. For example, on the DC side of the photovoltaic array, multiple string monitoring points are grouped into cluster 1 due to their small phase difference, while another string monitoring point is grouped into cluster 2 due to its larger phase difference. The two clusters are electrically isolated by disconnecting switches or fuses in the combiner box. Finally, the topology cluster division submodule generates a set of inverter grid connection point topology paths. This set contains detailed topology information for each cluster (such as the number of monitoring points within the cluster, connection hierarchy, and key electrical parameters) and connection paths between clusters (such as cross-cluster monitoring point pairs and impedance parameters of connecting lines). For example, the topology path set may record that cluster 1 contains monitoring points A, B, and C, and the connection method is string A → combiner box 1 → string B → inverter X; cluster 2 contains monitoring points D and E, and the connection method is string D → combiner box 2 → string E → inverter Y; cluster 1 and cluster 2 are connected through the isolation switch of the DC bus to form the grid connection point topology of the entire array.

[0065] Throughout the process, the phase difference calculation submodule and the topology clustering submodule communicate via a high-speed data bus to ensure real-time transmission and processing of phase difference data. To address measurement errors in complex environments, the phase difference calculation submodule incorporates an error correction mechanism, reducing the impact of random errors by averaging multiple samples. The topology clustering submodule possesses adaptive adjustment capabilities, dynamically updating the clustering results based on changes in the phase difference of monitoring points during actual operation. Through this process, the electrical topology generation module ultimately constructs a set of topology paths that accurately reflects the electrical connection characteristics of the photovoltaic array. This provides the subsequent fluctuation characteristic analysis module with clear cluster boundaries and node association information, ensuring that fluctuation characteristic analysis can focus on the electrical coupling characteristics within the same cluster and the propagation characteristics across clusters.

[0066] Example 3: See Figure 4 The fluctuation characteristic analysis module, as a core component of the system, is primarily responsible for quantifying the inverter's response characteristics under grid fluctuations using historical and real-time data from monitoring points within the same cluster, providing a characteristic basis for subsequent strategy control. This module includes a power sequence extraction submodule and a response characteristic quantification submodule, which work together to complete the entire process from data screening to feature quantification.

[0067] The primary task of the power sequence extraction submodule is to screen target monitoring points and extract valid data. The system first identifies monitoring points belonging to the same cluster based on the inverter grid connection point topology path set output by the electrical topology generation module. Monitoring points within a cluster exhibit strong electrical coupling during grid fluctuations due to their smaller electrical phase difference (less than the phase difference tolerance threshold), resulting in higher consistency in power changes. Therefore, selecting monitoring points from the same cluster can more accurately reflect the impact of local grid fluctuations on the inverter.

[0068] After selecting monitoring points, the submodule needs to extract the inverter output active power sequence and the grid frequency sampling sequence within a continuous time window. The length of the time window is dynamically adjusted according to the scale and fluctuation characteristics of the photovoltaic array, and is usually set to 1 hour (containing 60 1-minute sampling points) to cover typical fluctuation cycles (such as the rapid change cycle of irradiance caused by cloud cover, which is about 10-30 minutes). For the inverter output active power, the data comes from the real-time sampling values ​​of the inverter output voltage and current in the operating environment modeling module. It is calculated in real time using the power calculation formula P(t)=U(t)×I(t) (where U(t) is the inverter output voltage at time t, and I(t) is the inverter output current at time t). The sampling frequency is 1Hz, and finally, it is downsampled according to the minute-level average within the time window to form an active power sequence of length 60, P=[P(t1),P(t2),...,P(t... 60 )], where t i Represents the timestamp of the i-th minute (e.g., from "2025-08-11 10:00:00" to "2025-08-11 11:00:00").

[0069] The acquisition of the grid frequency sampling sequence relies on a synchronous phasor measurement unit (PMU), which is deployed at the grid connection point of the photovoltaic array. Using a synchronous clock (such as BeiDou satellite timing) as a reference, it acquires the grid frequency signal in real time at a sampling frequency of 1 Hz, and outputs the real-time grid frequency f(t) (in Hz). To align with the power sequence time, the submodule also downsamples the frequency sequence by a minute-level average, generating a frequency sequence F = [f(t1), f(t2), ..., f(t...] corresponding one-to-one with the power sequence timestamps. 60 )).

[0070] After data extraction and alignment are completed, the response feature quantization submodule begins calculating the power fluctuation rate and frequency offset rate per unit time. The power fluctuation rate is used to quantify the degree of change in inverter output power over time, and its calculation formula is defined as:

[0071]

[0072] Where: ρ(t) i) represents the power fluctuation rate at minute i (i≥2), t i The time point refers to the timestamp of the i-th minute, P(t) i Let P(t) be the average active power output of the inverter at the i-th minute. i-1 Let P(t) be the power average over the (i-1)th minute, and let the denominator be the arithmetic mean of the power over the two minutes. This is used to eliminate the influence of the absolute value of power on volatility, making volatility comparable across different power levels. For example, if the power increases from 300kW to 320kW within a minute, then P(t) i )=320kW,P(t i-1 ) = 300kW, substituting into the formula, we get (i.e., 6.45% / min).

[0073] The frequency offset rate is used to quantify the degree of deviation of the power grid frequency from the rated frequency (50Hz), and its calculation formula is as follows:

[0074]

[0075] Where: δ(t) i ) represents the frequency offset rate at minute i, t i The time point refers to the timestamp of the i-th minute, f(t) i Let f be the average power grid frequency at the i-th minute. ref =50Hz is the rated frequency. For example, if the power grid frequency is 49.8Hz in a certain minute, then (i.e., 0.4% / min).

[0076] The response feature quantization submodule calculates the power volatility sequence ρ=[ρ(t2),ρ(t3),...,ρ(t)]. 60 )] and the frequency offset sequence δ=[δ(t2),δ(t3),...,δ(t 60 The fluctuation range is compared with the preset allowable fluctuation range. The allowable fluctuation range is determined by the inverter's rated capacity and the grid dispatch requirements. For example, for an inverter with a rated capacity of 1000kW, the upper limit of the allowable power fluctuation rate is ±5% / min (i.e., ρ≤0.05), and the upper limit of the frequency offset rate is ±0.2% / min (i.e., δ≤0.002).

[0077] When the power volatility or frequency offset exceeds the allowable range for a given minute, the submodule determines that period as an abnormal volatility period and records the start time, end time, and maximum volatility (e.g., ρ) of that period. max =0.08), maximum frequency offset (e.g., δ) max=0.005) and the corresponding power value (e.g., P = 350kW) and frequency value (e.g., f = 49.5Hz). Finally, the response characteristic quantification submodule generates the fluctuation response characteristic analysis results, which record the abnormal fluctuation period and the corresponding over-limit intensity (i.e., the difference between the actual volatility and the allowable volatility, e.g., 0.08-0.05 = 0.03) for each monitoring point in each time window in tabular or time series form.

[0078] Throughout the process, the power sequence extraction submodule and the response feature quantization submodule interact with each other via memory sharing or message queues to ensure strict alignment of the extracted power and frequency data. To address data gaps or anomalies, the submodule incorporates a data completion mechanism: for short-term gaps (less than 5 minutes), linear interpolation is used to fill them; for long-term gaps, the data for that period is marked as invalid and noted in the analysis results to avoid affecting the accuracy of subsequent strategy control. Through this process, the fluctuation feature analysis module ultimately achieves a quantitative description of the inverter's response characteristics under grid fluctuations, providing crucial fluctuation characteristic basis for the strategy control module to identify abnormal operating points.

[0079] Example 4: See Figure 5 The operation of the irradiance fluctuation characteristic submodule relies on long-term accumulated historical irradiance data of photovoltaic arrays. This data is typically stored hourly, covering irradiance variations under different seasons and weather conditions. For example, a photovoltaic power station recorded irradiance values ​​at each monitoring point daily from January 2024 to July 2025, forming a historical database containing approximately 18,000 records. The submodule performs time series analysis on this data, identifying abrupt irradiance events using a sliding window method (window length set to 10 minutes). Specifically, for each time point t, it calculates the irradiance difference between it and the previous time point t-1:

[0080] ΔI=I(t)-I(t-1)

[0081] If the absolute value of ΔI exceeds the set threshold (e.g., 100W / m), 2 If the irradiance suddenly increases (ΔI>0) or suddenly decreases (ΔI<0), the event is determined as a sudden increase (ΔI>0) or a sudden decrease (ΔI<0), and the start time of the event is recorded as t-5 minutes (center of the window), the end time is t+5 minutes, and the fluctuation intensity is |ΔI|.

[0082] Taking data from a week in June 2025 as an example, the submodule analysis yielded the following typical event: On June 10th, from 10:00 AM to 10:10 AM, the irradiance increased from 450 W / m². 2 Rapidly rises to 580W / m 2 ΔI=130W / m 2 Marked as "strong surge"; from 14:30 to 14:40 on June 12, the irradiance increased from 800W / m2 The power dropped sharply to 620 W / m 2 ΔI=-180W / m 2 Marked as "extremely strong sudden drop"; from 9:15 to 9:25 AM on June 15th, the irradiance dropped from 300 W / m². 2 Increased to 420W / m 2 ΔI=120W / m 2 These events are marked as "strong surges". The time periods of these events and their corresponding fluctuation intensities are compiled into a mapping table, see Table 1.

[0083] Table 1: Mapping Relationship Table.

[0084]

[0085] When the fluctuation characteristic analysis module processes the real-time power sequence of a monitoring point, if the time period overlaps with an event time period in the mapping table (e.g., time overlap exceeds 80%), it calls the fluctuation intensity level of that event and performs joint analysis with the power fluctuation rate. For example, the power sequence of a monitoring point during the period from 10:00 to 10:10 on June 10, 2025, shows that its output power increased from 200kW to 280kW, with a power fluctuation rate of (280-200) / [(200+280) / 2] = 0.2857 (i.e., 28.57% / min). The corresponding irradiance fluctuation intensity level for this period is "strong surge". The module will correlate the power fluctuation rate with the irradiance fluctuation amount and calculate the correlation coefficient between the two (e.g., 0.92), thereby quantifying that the power fluctuation during this period is mainly caused by a surge in irradiance. This correlation analysis makes the fluctuation response characteristic analysis results more consistent with the actual physical process and avoids the one-sidedness of relying solely on the power fluctuation rate.

[0086] The voltage phase calibration submodule of the electrical topology generation module obtains the voltage phase reference value through the synchronous phasor measurement unit (PMU) at the power grid's point of common coupling (PCC). The PMU uses BeiDou satellite timing as its time reference and outputs PCC voltage phasor information once per second. The phase angle is referenced to the zero-crossing point of the sine wave at the grid's rated frequency (50Hz), ranging from -180° to +180°. For example, if the PMU measures a PCC voltage phase of 30° at a certain moment, this value is broadcast as the reference value to all monitoring points. The voltage sensors at each monitoring point acquire the local output voltage signal in real time, extract the phase angle after synchronous sampling, and calculate the deviation from the reference value. For example, the local voltage phase at monitoring point A is 35°, with a deviation of 35° - 30° = +5°; the local voltage phase at monitoring point B is 28°, with a deviation of 28° - 30° = -2°; and the local voltage phase at monitoring point C is 42°, with a deviation of 42° - 30° = +12°.

[0087] The topology clustering submodule performs clustering based on these deviations. Unlike the original phase difference calculation, the submodule no longer directly uses the phase difference between adjacent monitoring points, but instead compares the difference in deviations between adjacent points. For example, the deviations between monitoring points A and B are +5° and -2°, respectively, with a difference of (-2°) - (+5°) = -7° (absolute value 7°); the deviations between monitoring points B and C are -2° and +12°, respectively, with a difference of (+12°) - (-2°) = +14° (absolute value 14°). The system's preset tolerance threshold for deviation differences remains 10°. Therefore, the difference of 7° between monitoring points A and B is less than the threshold, and they are classified as being in the same cluster; the difference of 14° between monitoring points B and C is greater than the threshold, and they are classified as being across cluster boundaries.

[0088] Based on this, the clustering results of a photovoltaic array are as follows: Cluster 1 includes monitoring points A, B, and D (the deviation difference between monitoring point D and A is 3°, less than the threshold), with local voltage phases of 35°, 28°, and 32°, and deviations of +5°, -2°, and +2°, respectively. The deviation difference between all points within the cluster is less than 10°, indicating tight electrical coupling. Cluster 2 includes monitoring points C and E (the deviation difference between monitoring points C and E is 5°, less than the threshold), with local voltage phases of 42° and 38°, and deviations of +12° and +8°, ​​respectively. The cluster is tightly coupled, but the deviation difference from cluster 1 is 14° (greater than the threshold), forming a clear cluster boundary. This clustering method based on deviation can more accurately reflect the impact of the deviation between the monitoring points and the grid reference phase on electrical connections, avoiding misjudgments caused by local phase fluctuations.

[0089] In practical applications, if the local voltage phase of a monitoring point slowly shifts due to component aging or changes in line impedance (e.g., the phase gradually increases by 5° over 24 hours), the voltage phase calibration submodule will update its deviation in real time (from the initial +5° to +10°), and the topology clustering submodule will reassess the cluster affiliation based on the new deviation difference. For example, monitoring point F, originally belonging to cluster 1, gradually increases its deviation from +3° to +11°, and the difference between it and other points in cluster 1 (deviations of +5°, -2°) gradually exceeds 10°, eventually being reclassified as cluster 3, ensuring the dynamic adaptability of cluster division.

[0090] By analyzing historical data and establishing a mapping table in the irradiance fluctuation characteristic submodule, the correlation between fluctuation characteristics and physical events was quantified. By acquiring the reference phase and calculating the deviation in the voltage phase calibration submodule, and by dynamically dividing the topology clusters in the topology cluster division submodule, the adaptability of the topology structure to grid phase fluctuations was improved, providing a more reliable topology and characteristic basis for the subsequent strategy control module to accurately identify abnormal operating points.

[0091] Example 5: After identifying the out-of-limit nodes, the dynamic adjustment submodule of the strategy control module needs to further analyze the fluctuation propagation characteristics of the abnormal points in order to formulate targeted coordinated control strategies. The fluctuation propagation analysis submodule acquires historical and real-time power data of cross-cluster boundary monitoring points and determines the fluctuation direction by comparing the power value changes at adjacent time points. For example, if the power of a cross-cluster boundary monitoring point C is 200kW at time t and 210kW at time t+1, its fluctuation direction is determined to be upward; if the power drops to 190kW at time t+1, it is determined to be downward. For power data at multiple consecutive time points, the submodule eliminates instantaneous noise interference by calculating the average rate of change within a sliding window (e.g., taking the average power difference of the most recent 5 minutes), ensuring the accuracy of fluctuation direction judgment.

[0092] After determining the direction of the fluctuation, the submodule tracks the transmission path of the fluctuation between clusters based on the inverter grid connection point topology path set output by the electrical topology generation module. The topology path set records the electrical connection relationships between each monitoring point (e.g., node C belongs to cluster A and is directly connected to node D in cluster B; node D is connected to node E in cluster B). Therefore, the fluctuation originating from node C will preferentially propagate to the directly connected node D, and then from node D to node E, forming a propagation path of "C→D→E". To verify the accuracy of the path, the submodule can call historical data from the same period (e.g., whether the power change of node C in the same fluctuation event causes the power of nodes D and E to change synchronously). If the historical data shows that the correlation coefficient between the power change of node D and node C exceeds 0.8, the validity of the path is confirmed.

[0093] Based on the tracked propagation path, the dynamic adjustment submodule adds a preventative adjustment coefficient to the downstream nodes along the path. The magnitude of the adjustment coefficient is determined based on the electrical distance between nodes (determined by line resistance and reactance) and the historical fluctuation propagation efficiency (i.e., the ratio of power change at the downstream node to power change at the upstream node). For example, if the line resistance from node C to node D is low (low impedance) and the historical propagation efficiency is 0.9 (i.e., the power change at node D is 90% of that at node C), then the preventative adjustment coefficient for node D is set to 0.8 (i.e., 80% of the original adjustment). If the line resistance from node D to node E is high (high impedance) and the historical propagation efficiency is 0.6, then the preventative adjustment coefficient for node E is set to 0.5. In this way, the adjustment amount at downstream nodes is corrected in advance, effectively suppressing the propagation of fluctuations to other clusters and reducing the probability of large-area power fluctuations.

[0094] After receiving the set of abnormal operating points, the control strategy generation submodule of the coordinated output module calculates the weight ratio of each node's power fluctuation to the total cluster capacity. The total cluster capacity is the sum of the rated capacities of all nodes in the abnormal operating point set. For example, if node A's rated capacity is 500kW, node B's is 300kW, and the total cluster capacity is 800kW, then node A's weight ratio is 500 / 800 = 62.5%, and node B's is 37.5%. Based on the weight ratios, the submodule allocates the total adjustment amount to each node. For example, if the total adjustment amount is -100kW (requiring a 100kW reduction in output), then node A's initial adjustment amount is -62.5kW, and node B's is -37.5kW.

[0095] The instruction synthesis submodule, based on the allocated adjustment amount, combines grid dispatch instructions and local load demand to generate the final power correction instruction set. Grid dispatch instructions may include target power values ​​(e.g., requiring the cluster output power to not exceed 800kW) or adjustment rate limits (e.g., adjustment amount not exceeding 20kW per minute); local load demand reflects the actual electricity demand of the grid to which the photovoltaic array is connected (e.g., real-time load of 750kW). For example, if the grid dispatch requires a cluster output power limit of 800kW, while the current total output power is 900kW and the total adjustment amount is -100kW, the instruction synthesis submodule must ensure that the adjusted output power of each node does not exceed the dispatch limit while simultaneously meeting local load demand. If the initial adjustment of node A is -62.5kW, causing its output power to drop to 437.5kW (original output 500kW), and node B to drop to 262.5kW (original output 300kW), the total output will be 700kW, which is lower than the local load of 750kW. In this case, the instruction synthesis submodule needs to fine-tune the adjustment amount, appropriately reducing the adjustment amounts of nodes A and B (e.g., adjusting node A to -50kW and node B to -30kW) to raise the total output to 720kW, which satisfies the scheduling requirements and is close to the load demand.

[0096] After the correction command is executed, the optimization strategy feedback submodule analyzes the effectiveness of the adjustment by collecting actual output power data from each monitoring point. For example, after executing the correction command, the power of node C decreases from 210kW to 195kW (volatility decreases from +10kW / min to -15kW / min), the power of node D decreases from 205kW to 190kW (originally predicted to rise to 215kW), and the power of node E decreases from 200kW to 185kW (originally predicted to rise to 210kW), indicating that the preventative adjustment coefficient effectively suppresses the propagation of volatility. The submodule compares the actual volatility convergence data (such as the actual power change rate of each node and the deviation of the final output power from the target value) with the prediction results before adjustment to evaluate the rationality of the weight ratio and adjustment coefficient. If it is found that the weight ratio of node A is too high, resulting in an excessive adjustment (e.g., only -40kW is needed to meet the demand), then the weight ratio of node A is updated to 50%, and the weight ratio of node B is increased to 50% accordingly. If the preventive adjustment coefficient of 0.8 for node D is insufficient to suppress fluctuations (e.g., the power of node D still increases by 5kW / min), then the adjustment coefficient of the path is increased to 0.9.

[0097] Through the aforementioned feedback mechanism, the coordinated output module can continuously optimize the control strategy and improve the accuracy of subsequent commands. For example, in subsequent similar cross-cluster fluctuation events, the weight ratio of node A is adjusted to 50%, node B to 50%, and the preventative adjustment coefficient of node D is increased to 0.9. The system can more accurately allocate adjustment amounts, reducing the impact of power fluctuations on the power grid. This dynamic optimization mechanism enables the entire data acquisition and control system to be adaptive, continuously improving the control strategy based on actual operating conditions to ensure the stable operation of the photovoltaic grid-connected inverter.

[0098] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0099] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A data acquisition and control system for a photovoltaic grid-connected inverter, characterized in that, The system includes: The runtime environment modeling module acquires real-time irradiance data and component temperature data at each monitoring point of the photovoltaic array, and constructs a discrete state matrix of irradiance. The electrical topology generation module extracts the discrete state identifiers and monitoring point coordinates from the irradiance discrete state matrix to generate a set of inverter grid connection point topology paths. The fluctuation characteristic analysis module calls the monitoring points in the same cluster topology path of the inverter grid connection point topology path set to generate fluctuation response characteristic analysis results; The strategy control module identifies abnormal points in the monitoring points whose response intensity is greater than the grid tolerance threshold and are located in the cross-cluster topology path in the fluctuation response characteristic analysis results, extracts the power fluctuation characteristics and frequency offset data of the abnormal points, and forms a set of abnormal operation points of the inverter. The coordination output module acquires all monitoring points and corresponding electrical parameters from the set of abnormal inverter operation points, marks the inverter nodes that need to be coordinated and controlled, and generates a set of coordinated control instructions for the inverter cluster.

2. The data acquisition and control system for a photovoltaic grid-connected inverter according to claim 1, characterized in that, The runtime environment modeling module includes: The multi-dimensional parameter acquisition submodule obtains the geographical coordinates and timestamps of the photovoltaic array monitoring points, and simultaneously collects the real-time irradiance value, module temperature value, inverter output voltage value and output current value at the coordinate location, and integrates them into a four-dimensional parameter data group. The state discretization submodule, based on the irradiance data and temperature data in the four-dimensional parameter data group, divides the state intervals into preset order of magnitude, maps the voltage and current sampled values ​​to the corresponding state intervals, calculates the average voltage and current values ​​in the same state interval, and generates an irradiance discrete state matrix.

3. The data acquisition and control system for a photovoltaic grid-connected inverter according to claim 2, characterized in that, The electrical topology generation module includes: The phase difference calculation submodule extracts the discrete state identifier and monitoring point location information from the irradiance discrete state matrix, sorts adjacent monitoring point pairs based on the shortest electrical distance, measures the output voltage phase difference between each pair of monitoring points, and generates a phase difference sequence of adjacent monitoring points. The topology cluster partitioning submodule sets a phase difference tolerance threshold based on the phase difference sequence of adjacent monitoring points. Monitoring point pairs below the threshold are marked as nodes in the same cluster, and monitoring point pairs exceeding the threshold are marked as nodes crossing the cluster boundary. The topology path is divided according to the cluster boundary to generate a set of inverter grid connection point topology paths.

4. The data acquisition and control system for a photovoltaic grid-connected inverter according to claim 3, characterized in that, The fluctuation characteristic analysis module includes: The power sequence extraction submodule filters the monitoring points of the same cluster topology path in the inverter grid connection point topology path set, extracts the inverter output active power sequence and grid frequency sampling sequence within a continuous time window, and generates a power frequency time series dataset by aligning them in time. The response feature quantization submodule calculates the power fluctuation rate and frequency offset rate per unit time based on the power frequency time series dataset, compares them with the allowable fluctuation range under the inverter's rated capacity, identifies fluctuation periods that exceed the allowable range and records the response over-limit intensity, and generates fluctuation response feature analysis results.

5. The data acquisition and control system for a photovoltaic grid-connected inverter according to claim 4, characterized in that, The policy control module includes: The over-limit node identification submodule detects monitoring points whose response intensity is greater than the grid tolerance threshold in the fluctuation response feature analysis results, as well as monitoring points located in cross-cluster topology paths, extracts the extreme values ​​of power fluctuation and frequency offset during abnormal periods, and generates an abnormal operation feature dataset. The dynamic adjustment submodule calculates the consistency between the power fluctuation direction and the frequency offset direction based on the abnormal operation feature dataset. When the changes in both directions are inconsistent, it is marked as a coordination control priority node. All priority node numbers and parameters are integrated to form a set of abnormal operation points of the inverter.

6. The data acquisition and control system for a photovoltaic grid-connected inverter according to claim 5, characterized in that, The coordination output module includes: The control strategy generation submodule obtains the monitoring point number and electrical parameters of the inverter abnormal operation point set, calculates the weight ratio of the power fluctuation of each node to the total capacity of the cluster, and allocates the output power adjustment amount according to the weight ratio. The instruction synthesis submodule generates an inverter output power correction instruction set based on the allocated power adjustment amount, combined with grid dispatch instructions and local load requirements.

7. The data acquisition and control system for a photovoltaic grid-connected inverter according to claim 6, characterized in that, The runtime environment modeling module also includes: The irradiance fluctuation feature submodule obtains the temporal distribution pattern of historical irradiance data, identifies the occurrence periods of sudden increases and decreases in irradiance, and establishes a mapping relationship table between time periods and irradiance fluctuation intensity. The fluctuation characteristic analysis module calls the mapping table to perform fluctuation characteristic quantization on the inverter output power sequence corresponding to the time period.

8. The data acquisition and control system for a photovoltaic grid-connected inverter according to claim 7, characterized in that, The electrical topology generation module also includes: The voltage phase calibration submodule acquires the voltage phase reference value at the power grid point of common coupling and calculates the deviation between the output voltage phase of each monitoring point and the reference value. The topology cluster partitioning submodule uses the deviation value instead of the output voltage phase difference to perform cluster partitioning.

9. A data acquisition and control system for a photovoltaic grid-connected inverter according to claim 8, characterized in that, The policy control module also includes: The wave propagation analysis submodule detects the direction of power wave fluctuations at monitoring points across cluster boundaries and tracks the propagation path of waves in the topology. The dynamic adjustment submodule adds a preventative adjustment coefficient to the downstream nodes based on the transmission path.

10. A data acquisition and control system for a photovoltaic grid-connected inverter according to claim 9, characterized in that, The coordination output module also includes: The optimization strategy feedback submodule collects actual power fluctuation convergence data after command execution and updates the power adjustment weight ratio under the same scenario. The instruction synthesis submodule generates an inverter output power correction instruction set based on the updated weight ratio.

Images