Island detection and protection method and device applied to photovoltaic grid-connected intelligent measurement and control switch, and computer program product

Abstract

The invention discloses an island detection and protection method and device applied to a photovoltaic grid-connected intelligent measurement and control switch and a computer program product, the intelligent measurement and control switch is connected in series between an inverter output end of a distributed photovoltaic system and a grid-connected point of a power grid, and the island detection and protection method is characterized by comprising the following steps: acquiring electrical quantity of the grid-connected point in real time; dynamically updating an adaptive reference model based on the historical data of the electrical quantity; comparing the electrical quantity acquired in real time with the adaptive reference model, and performing anomaly detection through multi-criterion fusion; and when the islanding is detected to be suspected, executing a rapid protection action through a switch execution mechanism, so that the photovoltaic system is separated from the power grid. The method is accurate and rapid in detection, has adaptive capability and communication cooperation capability, and can rapidly and accurately identify the island state, execute a tripping command and realize off-grid protection, thereby effectively preventing various potential safety hazards caused by island operation.

Description

Technical Field

[0001] This invention relates to the field of power system technology, specifically to a method, device, and computer program product for islanding detection and protection of photovoltaic grid-connected intelligent monitoring and control switches. Background Technology

[0002] With the rapid popularization of distributed photovoltaic (PV) power generation technology and its widespread integration into distribution networks, the power grid is transforming from a traditional passive unidirectional radiation network to an active bidirectional interactive system. While this transformation brings clean energy, it also poses serious challenges to the safe and stable operation of the power grid, with the islanding effect being a particularly prominent issue. The islanding effect refers to the situation where, when the main power source of the grid is interrupted due to a fault or maintenance, the distributed PV power generation system fails to detect the power outage in time and continues to supply power to local loads, forming a power island. Islanding operation brings a series of serious risks, including threatening the personal safety of grid maintenance personnel, causing damage to grid equipment, and negatively impacting the power quality of local loads.

[0003] Currently, the closest existing technologies mainly fall into two categories: one relies on the local islanding protection function of the photovoltaic inverter itself, which often employs passive detection methods (such as over / under protection by monitoring voltage and frequency) or active detection methods (such as introducing small disturbances). However, this type of method has a detection blind zone (NDZ), which may fail when the load power and photovoltaic output power are close to balance, and the active disturbance method may negatively impact power quality. The other category is a system-level protection scheme based on a smart grid-connected terminal for the distribution area. This scheme monitors the distribution area topology through the terminal, and after the inverter protection fails, the fusion terminal issues a trip command to the smart photovoltaic grid-connected switch. Although this method improves reliability, its response speed depends on the communication between the terminal and the switch. In the event of communication interruption or delay, the timeliness of the protection action is difficult to guarantee.

[0004] In addition, while some existing anti-islanding methods based on multi-switch collaboration have improved the detection success rate through communication and voting mechanisms between the bus grid-connected switch and the photovoltaic low-voltage grid-connected switch, their decision-making process is relatively complex, relies on reliable communication between switches, and may not respond quickly enough to rapidly changing islanding conditions. Summary of the Invention

[0005] The technical problem to be solved by the embodiments of the present invention is to provide a photovoltaic grid-connected intelligent monitoring and control switch and its reactive power regulation method, device and computer program product, so as to realize the rapid and accurate identification of islanding status and the execution of tripping commands without relying entirely on external communication, thereby effectively preventing various safety hazards caused by islanding operation.

[0006] To address the aforementioned technical problems, this invention provides a method for islanding detection and protection of a photovoltaic grid-connected intelligent monitoring and control switch. The intelligent monitoring and control switch is connected in series between the inverter output terminal of a distributed photovoltaic system and the grid connection point. The method includes: Step S1: Collect electrical quantities at the grid connection point in real time, and dynamically update the adaptive reference model based on the historical data of the electrical quantities; Step S2: Compare the real-time collected electrical quantities with the adaptive reference model, and perform anomaly detection through multi-criteria fusion; Step S3: When islanding is suspected, a fast protection action is performed through the switch actuator to disconnect the photovoltaic system from the grid.

[0007] Preferably, the electrical quantities include voltage amplitude, frequency, phase, and total harmonic distortion (THD), wherein: The adaptive reference model for voltage amplitude is established using a first-order Gaussian function, as shown in the following equation: N ( , ) in, This is the average value of the voltage amplitude; The standard deviation of the voltage amplitude; The adaptive reference model for frequency is established using a first-order Gaussian function, as shown in the following equation: N

[0008] in, The mean of the frequencies; The standard deviation of the frequency; Phase model based on the rate of change of phase difference between adjacent periods Establish as shown in the following formula:

[0009] in, The phase difference between adjacent periods; For the current period index, This is the sampling interval time; The harmonic model is established using a first-order Gaussian function, as shown in the following equation: N

[0010] in, The mean of THD; denoted as the standard deviation of THD.

[0011] Preferably, the adaptive reference model is dynamically updated based on historical data within a sliding time window, wherein the mean voltage amplitude is... and standard deviation The dynamic update calculation formula is as follows:

[0012]

[0013] in, This represents the number of sampling points within the time window; the frequency, phase, and harmonic model parameters are updated using the same calculation method.

[0014] Preferably, the anomaly detection through multi-criteria fusion specifically includes: The real-time collected voltage amplitude, frequency, phase, and total voltage harmonic distortion are input into the corresponding adaptive reference model; Determine whether the voltage amplitude exceeds the low probability range, whether the frequency exceeds the low probability range, whether the phase change value exceeds the adaptive threshold, and whether the total harmonic distortion rate deviates significantly from the normal range. If any judgment result is yes, the "island suspected" flag is triggered.

[0015] Preferably, when a "suspected islanding" flag is detected, if the communication module is normally connected to the smart converged terminal in the distribution area or adjacent switches, the islanding status is confirmed through a collaborative verification mechanism, specifically including: The intelligent monitoring and control switch sends alarm signals to the fusion terminal and adjacent switches. The fusion terminal returns the power grid status verification result based on global topology data, and the adjacent switches return the local detection status. The island is confirmed if any of the following conditions are met: the fusion terminal returns "main grid power outage" and this switch and ≥1 adjacent switch report an anomaly, or ≥2 adjacent switches including this switch report an anomaly simultaneously.

[0016] Preferably, the collaborative verification mechanism further includes: if the collaborating party fails to respond within a preset time, then a protection action is performed based on the detection result of step S2.

[0017] Preferably, when communication is interrupted, the autonomous verification mechanism includes: A small characteristic current signal of a specific frequency is injected into the power grid, and the voltage component at the injection frequency point is extracted by Fourier analysis to calculate the equivalent impedance. An islanded state is determined when the equivalent impedance meets one of the following criteria: The equivalent impedance increased sharply to 10 times or more of the historical baseline impedance; The equivalent impedance is greater than the dynamic judgment threshold, which is the sum of the historical impedance statistical mean and three times the historical impedance statistical standard deviation.

[0018] Preferably, the update cycle of the adaptive reference model is 1 second, and the sliding time window is the historical data of the most recent 15 minutes.

[0019] This invention also provides an islanding detection and protection device for a photovoltaic grid-connected intelligent monitoring and control switch, comprising: One or more processors; Memory; One or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, and the one or more applications are configured to perform the islanding detection and protection method applied to photovoltaic grid-connected smart monitoring and control switches.

[0020] The present invention also provides a computer program product, including computer instructions that instruct a computer device to perform an operation corresponding to the method.

[0021] Implementing this invention offers the following advantages: First, by using a local reference model that integrates and adaptively updates multiple parameters (voltage, frequency, phase, and harmonics) for judgment, it effectively solves the detection blind zone problem in existing technologies, significantly improving detection accuracy and reliability, especially under conditions where load and power generation are nearly balanced. Second, by delegating the core detection algorithm to the local execution of the intelligent control switch, it avoids delays caused by communication transmission and upper-level decision-making, ensuring rapid protection action with a total action time not exceeding 100 milliseconds, meeting relevant standard requirements. Third, this invention combines autonomy and collaboration, enabling it to independently perform high-precision islanding detection and protection when communication is interrupted, and to collaboratively verify with the intelligent fusion terminal of the distribution area or adjacent switches when communication is normal. This achieves the best combination of local rapid protection and system collaborative optimization, greatly improving protection redundancy and reliability. Finally, this invention not only achieves the core safety function of "controllability", but its comprehensive data acquisition, adaptive learning and event logging functions also provide a solid data foundation and control execution capability for the "observable, measurable and adjustable" nature of distributed photovoltaics, effectively supporting the national "four-fold" management requirements for distributed photovoltaics and providing a reliable guarantee for the large-scale safe grid connection of distributed photovoltaic power generation. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1This is a flowchart illustrating an islanding detection and protection method for a photovoltaic grid-connected intelligent monitoring and control switch, according to Embodiment 1 of the present invention.

[0024] Figure 2 This is a schematic diagram of the application scenario and intelligent measurement and control switch functional module of the method in the embodiment of the present invention. Detailed Implementation

[0025] The following description of the embodiments is taken with reference to the accompanying drawings, which illustrate specific embodiments in which the invention can be implemented.

[0026] Please refer to Figure 1 As shown, Embodiment 1 of the present invention provides a method for islanding detection and protection of a photovoltaic grid-connected intelligent monitoring and control switch. The intelligent monitoring and control switch is connected in series between the inverter output terminal of the distributed photovoltaic system and the grid connection point. The method includes: Step S1: Collect electrical quantities at the grid connection point in real time, and dynamically update the adaptive reference model based on the historical data of the electrical quantities; Step S2: Compare the real-time collected electrical quantities with the adaptive reference model, and perform anomaly detection through multi-criteria fusion; Step S3: When islanding is suspected, a fast protection action is performed through the switch actuator to disconnect the photovoltaic system from the grid.

[0027] As can be seen from the above steps, unlike the passive detection method of the prior art which uses a fixed threshold to detect blind spots, the embodiments of the present invention dynamically update the adaptive reference model based on historical data, which can adapt to changes in power grid operating conditions, significantly reduce the misjudgment rate and blind spots of single parameter detection, and especially improve the detection reliability when the load and power generation are close to balanced.

[0028] Specifically, please combine Figure 2As shown, the physical carrier of this embodiment of the invention is a photovoltaic grid-connected intelligent monitoring and control switch, which is installed in series between the output terminal of the photovoltaic inverter and the grid connection point (PCC point). Its hardware foundation includes a main control MCU module, a data acquisition module, a communication module, and a switch actuator. The high-precision data acquisition module includes a high-precision voltage transformer (PT), a current transformer (CT), an anti-aliasing filter, a signal conditioning circuit (including amplification and level shifting), and a high-speed, high-resolution ADC (analog-to-digital converter). Its function is to synchronously acquire the AC voltage and current signals at the grid connection point and convert them into digital quantities for processing by the main control MCU. The accuracy and bandwidth of the PT and CT must meet the requirements of harmonic measurement and phase measurement. The main control MCU module is the core processing unit and can use a high-performance 32-bit microcontroller (such as an ARM Cortex-M4 / M7 core), equipped with a floating-point unit (FPU) and sufficient storage space. The islanding detection and protection method of this invention is embedded within it. The switch actuator is typically a fast magnetic latching relay or a high-performance AC contactor, possessing high breaking capacity and fast operating characteristics (operating time is usually required to be less than 30-50ms) to ensure that breaking is completed within a total time limit of 100ms. The communication module enables communication between the main control MCU and external systems, supporting high-speed carrier communication (HPLC) or wireless communication (such as 4G / 5G or low-power wireless), and conforming to specific protocols such as IEC 61850, DL / T 860, or those formulated by various provincial power grid companies. It is used for data interaction and collaborative diagnostics with the intelligent converged terminal of the distribution area or adjacent switches.

[0029] The island detection and protection method of the present invention will be described in detail below.

[0030] Step S1 involves real-time data acquisition and local model building.

[0031] The high-precision data acquisition module continuously and synchronously acquires the voltage amplitude (U), frequency (f), phase (θ), and total harmonic distortion (THD) of the grid connection point (PCC point).

[0032] The main control MCU establishes adaptive reference models for each parameter under normal non-islanded operation based on historical data within a set time window (e.g., the most recent 15 minutes). These models are dynamically updated based on historical data within the set time window (e.g., the most recent 15 minutes) to establish a reference benchmark under normal non-islanded operation. All models rely solely on locally acquired data, such as voltage amplitude (U), frequency (f), phase (θ), and total harmonic distortion (THD), and are adaptively updated through the main control MCU. The specific forms and parameter meanings of each model are as follows: 1. Voltage Model Model format: Based on statistical distribution modeling, using a first-order Gaussian function (i.e., normal distribution): N ( , ) in, This represents the average voltage amplitude, indicating the average voltage level under normal operating conditions. The standard deviation of the voltage amplitude represents the degree of dispersion of voltage fluctuations.

[0033] Adaptive update mechanism: The main control MCU dynamically recalculates based on historical voltage data within a sliding time window (such as the most recent 15 minutes). and The calculation formula is:

[0034]

[0035] in, The number of sampling points within the time window (e.g., if 15 minutes of data are sampled once per second, then...). =900).

[0036] The purpose of this model is to: in real-time detection, if the current voltage value Falling into a low probability range (e.g., | - |>3 If the probability is less than 1%, it is considered abnormal.

[0037] 2. Frequency Model Model format: Also based on statistical distribution modeling, using a first-order Gaussian function (normal distribution): N

[0038] in, This is the average frequency, representing the average frequency under normal operating conditions (usually the power frequency of 50Hz). The standard deviation of the frequency represents the stability of frequency fluctuations.

[0039] Adaptive update mechanism: The main control MCU dynamically updates based on historical frequency data within the same time window. and The calculation formula is consistent with the voltage model:

[0040]

[0041] The purpose of this model is to: in real-time detection, if the current frequency value Exceeding the low probability range (e.g., | - |>3 If it is, then it is considered abnormal.

[0042] 3. Phase Model Model format: Used to calculate the rate of change of phase difference between adjacent periods (i.e., the rate of phase abrupt change):

[0043] in, The rate of change of the phase difference; The phase difference between adjacent periods; For the current period index, The sampling interval is typically one power grid cycle, 20ms for 50Hz. Phase difference between adjacent periods The calculation method is as follows:

[0044] in, and These are the phase values ​​for the current period and the previous period, respectively.

[0045] The model is based on the rate of change of phase difference. Calculate the adaptive threshold using the following formula. :

[0046] in, For history The mean value represents the average rate of phase change; For history The standard deviation represents the amplitude of the phase change fluctuation; and Adjustment coefficient (typical value) =1, =3, which can be optimized based on historical data.

[0047] Adaptive update mechanism: The main control MCU dynamically updates based on historical phase data within the time window. and :

[0048]

[0049] in, This represents the number of samples of the phase difference change rate within the time window.

[0050] The purpose of this model is to monitor phase abrupt changes; if the current... > This indicates a possible phase anomaly (such as phase loss due to islanding).

[0051] 4. Harmonic Model Model format: Used to monitor abnormal deviations in total harmonic distortion (THD), modeled based on statistical distribution: N

[0052] in, The mean of THD represents the average level of waveform distortion under normal operating conditions. The standard deviation of THD represents the range of waveform distortion fluctuations.

[0053] Adaptive update mechanism: The main control MCU dynamically updates based on historical THD data within the time window. and :

[0054]

[0055] The purpose of this model is: in real-time detection, if the current... Significant deviation from the normal range (e.g., | |> ,in If the value is a constant (usually taken as 3 to cover 99.7% of the normal range), then it is considered an abnormal waveform distortion.

[0056] It should be noted that all the above models use historical data within a sliding time window (e.g., 15 minutes) for periodic updates (e.g., every second or every power grid cycle) to ensure the models adapt to changes in power grid operation. Among the parameters of the above models, the mean... The standard deviation represents the long-term average value of the parameter under non-islanded conditions, reflecting the steady-state level. This indicates the fluctuation range of the parameter, used to quantify the sensitivity of anomaly detection; the larger the value, the greater the allowable fluctuation; threshold coefficient (e.g.) This is a constant optimized based on historical data, used to define "extremely low probability intervals" or "significant deviations." Typical values ​​are derived from statistical confidence levels (e.g., 99.7% corresponds to...). ).

[0057] These models provide an adaptive reference benchmark for the main control MCU, used for multi-criteria fusion analysis in subsequent step S2 (such as determining whether parameters fall into a low-probability interval). Model updates and calculations are performed locally, without relying on external input.

[0058] Voltage and frequency models can be built based on their statistical distributions (such as a first-order Gaussian function) and their mean and standard deviation can be updated dynamically; phase models are used to calculate the rate of change of phase difference between adjacent periods; harmonic models are used to monitor anomalies in the degree of waveform distortion.

[0059] Step S2 performs anomaly detection using multi-criteria fusion.

[0060] The main control MCU inputs the real-time collected parameter values ​​into the corresponding adaptive reference model established in step S1 for analysis. Comprehensive criteria are set, such as: voltage or frequency values ​​appearing in the current model within a very low probability (e.g., 1%) range, phase abrupt changes exceeding the adaptive threshold, or THD significantly deviating from the normal range. When any preset primary anomaly condition is met, the "suspicious islanding" flag is triggered.

[0061] Step S3 involves executing a rapid protection action.

[0062] Once the islanded state is confirmed via step S2, the main control MCU immediately sends a trip command to the switch actuator (circuit breaker / contactor) to disconnect the photovoltaic system from the power grid. The total time for the entire detection and protection action, from the occurrence of the abnormality to the complete disconnection of the switch, is required to be no more than 100 milliseconds (100ms).

[0063] After the protection action is executed, the intelligent monitoring and control switch uploads the event record (including trigger criteria, action time, electrical parameters before the action, etc.) to the distribution area fusion terminal or the superior monitoring system through the communication module, providing operators with detailed and observable data.

[0064] As a further improvement to this embodiment of the invention, in addition to anomaly detection through local multi-criteria in step S2, this invention also establishes a dual-mode collaborative verification mechanism, namely, collaborative verification when communication is normal, and active perturbation verification when communication is interrupted. Specific details are as follows: Upon detecting a suspected islanding situation, if the communication module of the intelligent monitoring and control switch (such as an HPLC carrier communication or wireless module) is connected normally to the intelligent fusion terminal of the distribution area or other adjacent intelligent switches, the abnormal information will be uploaded immediately. The main control MCU can receive topology verification information from the fusion terminal or collaborative diagnostic information from adjacent switches for joint judgment. Under specific strategies or during communication interruptions, this switch can also autonomously inject a small characteristic current signal of a specific frequency into the power grid and monitor changes in grid impedance. If the injected signal characteristics are significantly amplified or reflected, the islanding judgment is strengthened. The specific joint judgment logic and implementation method are as follows: I. Joint Judgment When Communication is Normal (Based on Topology Verification and Collaborative Diagnosis) (a) Information exchange process: (1) When the local detection triggers the “island suspected” flag, the intelligent measurement and control switch immediately sends an alarm signal to the intelligent fusion terminal of the distribution area and the adjacent intelligent switch of the same level through the communication module (such as HPLC or wireless), including the abnormal parameter type (such as voltage drop, frequency offset), timestamp and location identifier.

[0065] (2) The integrated distribution terminal feeds back the power grid status verification results based on global topology data: If the upstream power grid circuit breaker has been tripped (for planned maintenance or power outage due to fault), a "main grid power outage" confirmation signal will be returned; otherwise, a "power grid normal" status will be returned.

[0066] (3) Adjacent smart switches synchronously return local detection status (such as whether the same type of abnormality has been triggered), forming a multi-node diagnostic matrix.

[0067] (ii) Joint decision-making rules: (1) Scene 1: Confirming the true isolated island An island is confirmed if any of the following conditions are met: a. The converged terminal returns "Main network power failure" and this switch and ≥1 adjacent switch report abnormalities; b. ≥2 adjacent switches (including the switch itself) report anomalies simultaneously (even if the fusion terminal does not respond).

[0068] (2) Scenario 2: Eliminating pseudo-isolated islands If the fusion terminal returns "Power grid normal" and there are no abnormalities in adjacent switches, it is determined to be a local false alarm, and the "island suspected" flag is reset.

[0069] (3) Scenario 3: Timeout Degradation Handling If the collaborating party does not respond within the set time window (e.g., 20ms), the decision will be made independently by local criteria (depending on the result of step S2).

[0070] II. Autonomous Verification During Communication Interruption (Based on Active Perturbation Method) (a) Disturbance injection and monitoring mechanism: (1) Signal generation: The main control MCU controls the inverter circuit to inject a micro current of a specific frequency (such as 55Hz / 65Hz, amplitude ≤0.5% of rated current, lasting 100~200ms) to avoid the power quality sensitive frequency band.

[0071] (2) Impedance calculation: Real-time acquisition of grid connection point voltage response, and extraction of voltage component at injection frequency point through Fourier analysis. ), combined with injected current ( ) Calculate the equivalent impedance ( ):

[0072] (II) Determination of Isolated Island Characteristics: (1) Normal grid connection: The power grid exhibits low impedance characteristics ( The injected signal is absorbed, resulting in a weak voltage response. Impedance determination threshold:

[0073] in, Historical impedance statistical mean The standard deviation of historical impedance statistics.

[0074] (2) Islanding state: After the power grid is disconnected, the load impedance dominates. If the following is detected: Equivalent impedance It surged to more than 10 times the historical baseline, that is , Historical baseline impedance; or ; This is then judged as "significantly amplified impedance", strengthening the confidence level of the islanding problem.

[0075] III. Output and Action Logic of Joint Judgment

[0076] The working principle of this invention will be briefly explained below: After the system is powered on, the data acquisition module and the main control MCU continue to operate. The main control MCU continuously performs multi-parameter monitoring and model adaptive updates. When the power supply from the grid side is normal, the acquired electrical parameters are all within the normal range of its adaptive reference model. Once the grid side loses power and forms an island, the electrical parameters at the grid connection point (especially voltage, frequency, and phase) will change due to the disruption of the power balance between photovoltaic power generation and local loads. This invention significantly improves the sensitivity to such changes through multi-criteria fusion analysis. For example, even in extreme cases of high power balance, load switching or the operation of nonlinear loads may cause harmonic content or phase transients, and these subtle changes can be captured by the comprehensive criteria of this method. When an islanding state is determined, the main control MCU directly drives the switch actuator to quickly trip. This local decision-making and rapid execution mechanism is the fundamental guarantee of its high reliability and speed. At the same time, the collaborative verification function realized through the communication module provides additional judgment basis for complex operating conditions, further reducing the risk of false tripping, and making the protection action "observable" by the upper-level system, thus forming a complete protection system integrating detection, decision-making, execution, and reporting.

[0077] The known and potential application areas and methods of this invention include: 1. Distributed photovoltaic grid-connected system: As the core technology of the new generation of intelligent photovoltaic grid-connected switches with "four functions", it is widely used in the grid connection points of residential rooftop photovoltaic and industrial and commercial distributed photovoltaic projects, providing local and fast islanding protection and meeting grid access requirements.

[0078] 2. Microgrid system: Applied to microgrids composed of multiple energy sources such as photovoltaic, energy storage, diesel and charging, it is deployed at the grid connection point with the main grid to ensure that it can quickly and reliably disconnect from the main grid in the event of a grid failure, thus ensuring the safety of the microgrid itself and the main grid.

[0079] 3. Low-voltage distribution substations in rural and remote areas: Given the relatively weak power grid structure and potentially poor communication conditions in these areas, the local adaptive detection capability of this method is particularly important, which can effectively improve the power supply security and quality of rural power grids.

[0080] 4. Upgrading of existing photovoltaic power plants to meet the "four requirements": For early-built distributed photovoltaic projects that do not yet meet the "four requirements", adding or replacing them with the intelligent monitoring and control switches described in this invention is an effective technical path to achieve functional upgrades such as "controllability" and "observability".

[0081] 5. Virtual Power Plant (VPP) Safety Boundary Control: As the underlying safety barrier for distributed photovoltaic units aggregated in the virtual power plant, it ensures that individual units can reliably disconnect under abnormal circumstances when participating in grid ancillary services, thus preventing the escalation of accidents.

[0082] Corresponding to the islanding detection and protection method for photovoltaic grid-connected intelligent monitoring and control switches described in Embodiment 1 of the present invention, Embodiment 3 of the present invention also provides an islanding detection and protection device for photovoltaic grid-connected intelligent monitoring and control switches, comprising: One or more processors; Memory; One or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, and the one or more applications are configured to perform the islanding detection and protection method for photovoltaic grid-connected smart monitoring and control switches as described in Embodiment 1 of the present invention.

[0083] Corresponding to the islanding detection and protection method for photovoltaic grid-connected intelligent monitoring and control switches described in Embodiment 1 of the present invention, Embodiment 4 of the present invention also provides a computer program product, including computer instructions, which instruct computer equipment to perform the operations corresponding to the islanding detection and protection method for photovoltaic grid-connected intelligent monitoring and control switches described in Embodiment 1 of the present invention.

[0084] Preferably, the processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor, or the processor can be any conventional processor. The processor is the control center of the device, connecting various parts of the device through various interfaces and lines.

[0085] The memory mainly includes a program storage area and a data storage area. The program storage area can store the operating system, applications required for at least one function, etc., while the data storage area can store related data, etc. Furthermore, the memory can be a high-speed random access memory, or a non-volatile memory, such as a plug-in hard drive, a SmartMedia Card (SMC), a Secure Digital (SD) card, and a Flash Card, or other volatile solid-state storage devices.

[0086] It should be noted that the above-mentioned devices may include, but are not limited to, processors and memory, as will be understood by those skilled in the art.

[0087] As can be seen from the above description, compared with the prior art, the beneficial effects of the present invention are as follows: First, by using a local reference model that integrates and adaptively updates multiple parameters such as voltage, frequency, phase, and harmonics, the detection blind zone problem in the prior art is effectively solved, especially under the condition that the load and power generation are nearly balanced, significantly improving the detection accuracy and reliability. Second, by delegating the core detection algorithm to the local execution of the intelligent measurement and control switch, the delay caused by communication transmission and upper-level decision-making is avoided, ensuring the speed of protection action, with a total action time of no more than 100 milliseconds, meeting the relevant standard requirements. Third, the present invention has both autonomous and collaborative capabilities. It can independently complete high-precision islanding detection and protection when communication is interrupted, and can also perform collaborative verification with the intelligent fusion terminal of the distribution area or adjacent switches when communication is normal. This achieves the best combination of local rapid protection and system collaborative optimization, greatly improving the redundancy and reliability of protection. Finally, this invention not only achieves the core safety function of "controllability", but its comprehensive data acquisition, adaptive learning and event logging functions also provide a solid data foundation and control execution capability for the "observable, measurable and adjustable" nature of distributed photovoltaics, effectively supporting the national "four-fold" management requirements for distributed photovoltaics and providing a reliable guarantee for the large-scale safe grid connection of distributed photovoltaic power generation.

[0088] The above description is merely a preferred embodiment of the present invention and should not be construed as limiting the scope of the invention. Therefore, any equivalent variations made in accordance with the claims of the present invention are still within the scope of the present invention.

Claims

1. A method for islanding detection and protection of a photovoltaic grid-connected intelligent monitoring and control switch, wherein the intelligent monitoring and control switch is connected in series between the inverter output terminal of a distributed photovoltaic system and the grid connection point, characterized in that, The method includes: Step S1: Collect electrical quantities at the grid connection point in real time, and dynamically update the adaptive reference model based on the historical data of the electrical quantities; Step S2: Compare the real-time collected electrical quantities with the adaptive reference model, and perform anomaly detection through multi-criteria fusion; Step S3: When islanding is suspected, a fast protection action is performed through the switch actuator to disconnect the photovoltaic system from the grid.

2. The method according to claim 1, characterized in that, The electrical quantities include voltage amplitude, frequency, phase, and total harmonic distortion (THD), wherein: The adaptive reference model for voltage amplitude is established using a first-order Gaussian function, as shown in the following equation: N ( , ) in, This is the average value of the voltage amplitude; The standard deviation of the voltage amplitude; The adaptive reference model for frequency is established using a first-order Gaussian function, as shown in the following equation: N in, The mean of the frequencies; The standard deviation of the frequency; Phase model based on the rate of change of phase difference between adjacent periods Establish as shown in the following formula: in, The phase difference between adjacent periods; For the current period index, This is the sampling interval time; The harmonic model is established using a first-order Gaussian function, as shown in the following equation: N in, The mean of THD; denoted as the standard deviation of THD.

3. The method according to claim 1, characterized in that, The adaptive reference model is dynamically updated based on historical data within a sliding time window, wherein the mean voltage amplitude is... and standard deviation The dynamic update calculation formula is as follows: in, This represents the number of sampling points within the time window; the frequency, phase, and harmonic model parameters are updated using the same calculation method.

4. The method according to claim 1, characterized in that, The anomaly detection through multi-criteria fusion specifically includes: The real-time collected voltage amplitude, frequency, phase, and total voltage harmonic distortion are input into the corresponding adaptive reference model; Determine whether the voltage amplitude exceeds the low probability range, whether the frequency exceeds the low probability range, whether the phase change value exceeds the adaptive threshold, and whether the total harmonic distortion rate deviates significantly from the normal range. If any judgment result is yes, the "island suspected" flag is triggered.

5. The method according to claim 4, characterized in that, When a "suspected islanding" flag is detected, if the communication module is normally connected to the smart converged terminal in the distribution area or adjacent switches, the islanding status is confirmed through a collaborative verification mechanism, which specifically includes: The intelligent monitoring and control switch sends alarm signals to the fusion terminal and adjacent switches. The fusion terminal returns the power grid status verification result based on global topology data, and the adjacent switches return the local detection status. The island is confirmed if any of the following conditions are met: the fusion terminal returns "main grid power outage" and this switch and ≥1 adjacent switch report an anomaly, or ≥2 adjacent switches including this switch report an anomaly simultaneously.

6. The method according to claim 5, characterized in that, The collaborative verification mechanism further includes: if the collaborating party fails to respond within a preset time, a protection action is performed based on the detection result of step S2.

7. The method according to claim 4, characterized in that, When communication is interrupted, the autonomous verification mechanism includes: A small characteristic current signal of a specific frequency is injected into the power grid, and the voltage component at the injection frequency point is extracted by Fourier analysis to calculate the equivalent impedance. An islanded state is determined when the equivalent impedance meets one of the following criteria: The equivalent impedance increased sharply to 10 times or more of the historical baseline impedance; The equivalent impedance is greater than the dynamic judgment threshold, which is the sum of the historical impedance statistical mean and three times the historical impedance statistical standard deviation.

8. The method according to claim 3, characterized in that, The adaptive reference model is updated every 1 second, and the sliding time window is the historical data of the most recent 15 minutes.

9. An islanding detection and protection device applied to a photovoltaic grid-connected intelligent monitoring and control switch, characterized in that, include: One or more processors; Memory; One or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, the one or more applications being configured to perform the islanding detection and protection method for photovoltaic grid-connected smart monitoring and control switches as described in any one of claims 1 to 8.

10. A computer program product, characterized in that, Includes computer instructions that instruct a computer device to perform an operation corresponding to the method as described in any one of claims 1 to 8.

Images