A photovoltaic district integrated protection and self-healing method, system, device and storage medium

Abstract

The application discloses a photovoltaic substation comprehensive protection and self-healing method, system, device and storage medium, comprising: collecting electrical response data before and after disturbance of a photovoltaic substation, and establishing a comprehensive equivalent model of the substation; based on the comprehensive equivalent model of the substation and the electrical response data, updating the operation state of the substation in real time to obtain key indicators of supply-demand balance of the substation operation; calculating passive criteria and active criteria during the operation of the substation, judging the operation state of the substation based on the passive and active criteria, and calculating a consistency measurement value of the passive and active criteria judgment results; when the consistency measurement value of the passive and active criteria judgment results is greater than a preset threshold, triggering a comprehensive protection mechanism, executing substation fault isolation control, and sequentially trying to combine each branch after the main power supply is restored, to realize self-healing recovery of non-fault areas. The application can realize state adaptive identification, intelligent protection and step-by-step self-healing control of the photovoltaic substation, and improve the safety, reliability and intelligent level of the substation operation.

Description

Technical Field

[0001] This invention relates to the technical field of power system relay protection, and in particular to a method, system, equipment and storage medium for integrated protection and self-healing of photovoltaic power distribution areas. Background Technology

[0002] With the large-scale integration of distributed photovoltaic (PV) power generation and energy storage devices into distribution substations, especially the advancement of rooftop PV and county-wide promotion projects, the power supply structure of these substations is gradually shifting from a traditional single, centralized power source to a complex network with multi-source grid connection and bidirectional power flow. This change brings challenges in terms of substation operational status perception, fault identification, and automatic recovery. The short-circuit and fault response characteristics of PV inverters and energy storage inverters differ from those of traditional rotating machines, and the presence of mixed three-phase and single-phase grid connection, load and irradiance fluctuations within the substation makes protection methods based on single static criteria insufficient in terms of sensitivity, selectivity, and robustness to meet practical requirements. To improve discrimination accuracy and online response capability, existing research has proposed an active perturbation approach: applying controllable voltage / current perturbations at grid connection points or transformer taps, collecting voltage, current, and power responses before and after the perturbation, establishing a comprehensive equivalent model, and identifying parameters to obtain the compositional characteristics of power source groups and load groups in the substation and provide a basis for adaptive criterion selection. Meanwhile, the limitations of either a single active or passive method result in weak response of active disturbances in distributed scenarios, and passive criteria are easily affected by operational fluctuations. Therefore, we propose to combine active and passive criteria and use a self-consistent evaluator for consistency verification in order to improve the reliability of islanding and fault determination.

[0003] In terms of fault isolation and recovery, traditional one-time reclosing strategies struggle to accurately recover non-faulty branches due to distributed power generation deployment constraints and multi-branch parallel structures. Corresponding approaches include automatic reclosing processes based on extended voltage-time: first, the fault area is isolated; then, branches are reclosed sequentially by number, and the voltage or current characteristics after reclosing are detected. A joint criterion of voltage / current and duration is used to distinguish between transient and permanent faults, thus achieving step-by-step self-healing and precise fault isolation. Furthermore, energy storage devices, acting as energy buffers, can absorb energy during fault periods and feed it back during recovery. However, their unloading and feedback must be coordinated with protection and reclosing logic to avoid DC-side overvoltage or energy waste. In summary, existing technologies have mature methods for active disturbance identification, fusion of active and passive criteria and self-consistent evaluation, extended voltage-time automatic reclosing, and coordinated energy storage unloading. However, they lack a systematic solution that organically integrates these elements into a comprehensive system encompassing online status identification, reliable judgment, automatic isolation, step-by-step self-healing, and coordinated energy storage, to meet the higher requirements for accuracy, robustness, and automated recovery capabilities in high-penetration photovoltaic areas. Therefore, a new comprehensive protection and self-healing method and system is needed to fill the aforementioned technological gaps. Summary of the Invention

[0004] In view of the aforementioned existing problems, this invention is proposed. Therefore, this invention provides a comprehensive protection and self-healing method, system, device, and storage medium for photovoltaic power grids, addressing the problems of difficulty in identifying the status of power grids in high-penetration photovoltaic power grids, unreliable islanding and fault determination, difficulty in accurately implementing fault isolation and automatic reclosing, and insufficient coordinated control of energy storage unloading and feedback.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, embodiments of the present invention provide a comprehensive protection and self-healing method for photovoltaic power distribution areas, applicable to photovoltaic power distribution networks containing multiple distributed power sources and energy storage units, the method comprising: Collect electrical response data of the photovoltaic power station area before and after disturbance, and establish a comprehensive equivalent model of the power station area; Based on the comprehensive equivalent model of the transformer area and the electrical response data, the operating status of the transformer area is updated in real time to obtain key indicators of supply and demand balance in the transformer area. Calculate the passive and active criteria during the operation of the transformer area, determine the operating status of the transformer area based on the active and passive criteria, and calculate the consistency metric value of the determination result of the active and passive criteria. When the consistency metric value of the active and passive criterion judgment results is greater than the preset threshold, the integrated protection mechanism is triggered, the fault isolation control of the transformer area is executed, and the branch circuits are tested sequentially after the main power supply is restored to achieve self-healing recovery of the non-faulty area.

[0006] As a preferred embodiment of the photovoltaic power distribution area comprehensive protection and self-healing method described in this invention, it further includes: when the operating state of the power distribution area is abnormal, if a drop in the AC voltage of the power grid is detected, the energy storage unit is controlled to enter the energy absorption mode; during the power grid voltage recovery phase, the energy storage unit is controlled to switch to the slow release mode, and the energy feedback speed is dynamically adjusted based on the change rate of the total active power of the power distribution area.

[0007] As a preferred embodiment of the photovoltaic power station integrated protection and self-healing method of the present invention, the integrated protection mechanism is further triggered when the consistency metric value of the active and passive criterion judgment results is greater than a preset threshold: When the consistency metric value of the active and passive criteria judgment results is greater than the preset threshold, the circuit breaker and disconnect switch of the corresponding branch are disconnected; after the main circuit breaker is closed, starting from the smallest numbered branch, the reclosing operation is performed on each branch in sequence; after each reclosing, the voltage and current of the branch are detected in real time. If a low voltage or overcurrent is detected in a branch and the duration exceeds a preset time threshold, the branch is determined to have a permanent fault, and the branch is disconnected again. If the voltage and current return to normal, the circuit is kept closed, and the reclosing operation is performed on the next branch until all branches have been detected, thus achieving step-by-step self-healing recovery.

[0008] As a preferred embodiment of the photovoltaic power distribution area comprehensive protection and self-healing method described in this invention, the calculation of passive and active criteria during the operation of the distribution area, and the determination of the operating status of the distribution area based on the active and passive criteria, includes: continuously sampling the inverter output current during the normal operation of the distributed photovoltaic power distribution area, and storing it according to a preset time period to form a memory current reference value for the corresponding time period. When a disturbance is detected in the transformer area, the current sampling value for the current period is obtained, and the current change rate is calculated based on the memory current reference value. If the current change rate is greater than the active criterion threshold, the transformer area is determined to be in an abnormal operating state; otherwise, the transformer area is in a normal operating state. After active disturbance, the current is sampled again to calculate the new current change rate. If the new current change rate is greater than the passive criterion threshold, the transformer area is determined to be in abnormal operation. If the new current change rate is not greater than the passive criterion threshold, active scrambling is performed until the passive criterion is met or the maximum number of scrambling operations is reached. If the current change rate is still not greater than the passive criterion threshold after the maximum number of scrambling operations is reached, the transformer area is determined to be in normal operation.

[0009] As a preferred embodiment of the photovoltaic power station integrated protection and self-healing method described in this invention, the calculation of the consistency metric value of the active and passive criterion judgment results includes: Acquire current and voltage data of active and passive photovoltaic power distribution areas during normal and abnormal operation, and extract key features of current and voltage data to build a deep learning model for learning the relationship between current and voltage characteristics of photovoltaic power distribution areas and islanding status. The current and voltage data of the current transformer area during operation are input into the trained deep learning model, and the self-consistency score output by the deep learning model is compared with the results of active and passive anomaly protection. If the self-consistency score of the deep learning model is consistent with or similar to the results of active and passive protection, the decision system continues to maintain its current state; if the self-consistency score is inconsistent with the results of active and passive protection, the comprehensive protection mechanism is triggered.

[0010] As a preferred embodiment of the photovoltaic power station comprehensive protection and self-healing method of the present invention, the operation status of the power station is updated in real time based on the comprehensive equivalent model of the power station and the electrical response data, and the key indicators of the supply and demand balance of the power station operation are obtained by: feature extraction of the operation status of the power station. Calculate the composition ratio, equivalent impedance, and dynamic power distribution of the power supply group and load group in the transformer area based on the electrical response data. Using the calculation results as input, the operating status of the transformer area is iteratively updated in real time based on the parameter identification model of particle swarm optimization, and the supply and demand balance parameters are output.

[0011] As a preferred embodiment of the photovoltaic power station integrated protection and self-healing method described in this invention, the electrical response data includes voltage, current and power before and after the disturbance.

[0012] Secondly, this invention provides a comprehensive protection and self-healing system for photovoltaic power distribution areas, applicable to photovoltaic power distribution networks containing multiple distributed power sources and energy storage units. The system includes: The data acquisition module is used to collect electrical response data of the photovoltaic power station area before and after disturbance, and to establish a comprehensive equivalent model of the power station area; The update and optimization module is used to update the operating status of the transformer area in real time based on the comprehensive equivalent model of the transformer area and the electrical response data, so as to obtain the key indicators of supply and demand balance in the operation of the transformer area. The judgment module is used to calculate the passive and active criteria during the operation of the transformer area, and to judge the operating status of the transformer area based on the active and passive criteria, and to calculate the consistency metric value of the judgment result of the active and passive criteria. The fault isolation and self-healing control module is used to trigger the integrated protection mechanism when the consistency metric value of the judgment result of the active and passive criteria is greater than the preset threshold, execute the fault isolation control of the transformer area, and test each branch in sequence after the main power is restored to achieve self-healing recovery of the non-faulty area.

[0013] Thirdly, the present invention provides an electronic device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the integrated protection and self-healing method for the photovoltaic area are implemented.

[0014] Fourthly, the present invention provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the photovoltaic area integrated protection and self-healing method.

[0015] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention achieves dynamic characterization of the power supply and load groups in a photovoltaic distribution area by actively scrambling and obtaining equivalent parameters of the distribution area through online identification. This provides a basis for adaptive setting of protection criteria and thresholds, thereby improving the specificity and sensitivity of protection judgments. Furthermore, by performing consistency verification of active and passive criteria through a self-consistent evaluator, the probability of false alarms and failures to operate is significantly reduced, improving the reliability and robustness of islanding detection and fault identification.

[0016] Based on an extended voltage-time automatic reclosing and branch-by-branch recovery strategy, dynamic differentiation and step-by-step self-healing of transient and permanent faults can be achieved without significantly increasing the number of measuring devices, improving the automation efficiency and recovery speed of fault handling. Energy storage-coordinated unloading and feedback control provide an effective means for energy absorption during faults and energy balance during recovery, which helps to suppress overvoltage on the DC and grid sides, improve equipment safety, and enhance energy utilization efficiency. This invention organically combines active scrambling identification, self-consistent evaluation and judgment, extended voltage-time self-healing strategy, and energy storage unloading coordination to form a comprehensive protection and self-healing solution for high-penetration photovoltaic power stations, enhancing the safety, reliability, and automated recovery capabilities of power stations under complex operating conditions. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments 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. Wherein: Figure 1 This is a schematic diagram of the process flow of a comprehensive protection and self-healing method for a photovoltaic power station area according to an embodiment of the present invention; Figure 2 This is a logic diagram of a self-consistent evaluation and judgment unit for a photovoltaic power station integrated protection and self-healing method according to an embodiment of the present invention. Figure 3 This is a logic diagram of the fault isolation and self-healing control unit of a photovoltaic power station integrated protection and self-healing method according to an embodiment of the present invention. Detailed Implementation

[0018] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0019] Example 1, referring to Figures 1-3 This is one embodiment of the present invention, which provides a comprehensive protection and self-healing method for photovoltaic power distribution areas, applicable to photovoltaic power distribution networks containing multiple distributed power sources and energy storage units, such as... Figure 1 As shown, the method includes: S100: Collect electrical response data of the photovoltaic power station area before and after disturbance, and establish a comprehensive equivalent model of the power station area; S200: Based on the comprehensive equivalent model of the transformer area and the electrical response data, the operating status of the transformer area is updated in real time to obtain key indicators of supply and demand balance in the transformer area. S300: Calculate the passive and active criteria during the operation of the transformer area, determine the operating status of the transformer area based on the active and passive criteria, and calculate the consistency metric value of the determination result of the active and passive criteria. S400: When the consistency metric value of the judgment result of the active and passive criteria is greater than the preset threshold, the integrated protection mechanism is triggered, the fault isolation control of the transformer area is executed, and the branch circuits are tested sequentially after the main power supply is restored to achieve self-healing recovery of the non-faulty area.

[0020] It should be noted that this invention addresses the challenges of state identification, inaccurate fault identification, and insufficient reliability of islanding protection in high-penetration photovoltaic (PV) distribution areas due to the mixed operation of distributed power sources and loads. The invention involves actively scrambling the PV distribution area, collecting electrical parameters before and after the disturbance, establishing a comprehensive equivalent model, and identifying the supply and demand status of the area. Then, based on a self-consistent evaluation module, consistency analysis is performed on the outputs of active and passive criteria to determine abnormal states such as islanding or short circuits. Next, the system performs fault isolation and disconnection operations based on the determination results, and gradually reconnects branches according to an extended voltage-time strategy after grid recovery, achieving segmented self-healing. Finally, the energy storage device absorbs excess energy to suppress overvoltage during faults and releases energy to balance voltage fluctuations after system recovery. This method enables adaptive state identification, intelligent anomaly protection, and step-by-step self-healing control of PV distribution areas, significantly improving the safety, reliability, and intelligence level of the area's operation.

[0021] Furthermore, it also includes: when the operating state of the transformer area is abnormal, if a drop in the AC voltage of the power grid is detected, the energy storage unit is controlled to enter the energy absorption mode; during the power grid voltage recovery phase, the energy storage unit is controlled to switch to the slow release mode, and the energy feedback speed is dynamically adjusted based on the change rate of the total active power of the transformer area.

[0022] In an optional embodiment, the active scrambling signal in step S100 includes any one or a combination of voltage pulse injection, current pulse injection, or tap changer voltage regulation.

[0023] It should be noted that by controlling the periodic abrupt changes in the amplitude or phase of the photovoltaic power station current or voltage, under grid-connected conditions, the damping effect of the grid makes the current change caused by the active disturbance weak. However, when in an islanded or locally abnormal voltage state, the disturbance signal produces a significant response in the amplitude or phase of the power station current, thereby identifying islanded or locally abnormal voltage drops. To prevent excessive disturbance amplitude from causing system fluctuations, this method adopts a successive incremental disturbance strategy: when the current change rate does not meet the main criterion, the disturbance amplitude is gradually increased and sampling and discrimination are repeated until the auxiliary criterion is met or the maximum disturbance limit is reached, thus controlling the disturbance within a safe range.

[0024] In this embodiment of the invention, the electrical response data in step S100 includes: voltage, current and power before and after the disturbance.

[0025] Furthermore, during the active scrambling and data acquisition phase, it is necessary to analyze and determine the parameters that need to be identified in the integrated load model. Among these, the electromotive force of the equivalent Thevenin circuit of the motor is crucial. E S , motor reactance X S / resistance R S The ratio k1, photovoltaic power reactive power Q V / meritorious P V Ratio k2; Load reactance X / Load resistance R The ratio k3 is a specified empirical value. Parameters that need to be identified include: the electromotive force of the equivalent Thevenin circuit of the motor. E S,x and E S,y (Its modulus is a specified Es), motor reactance X S Photovoltaic power supply active power P V Load reactance X Four parameters. Multiple sets of measured values ​​for parameters other than the one to be identified are obtained by adjusting the perturbation, including the power at the measurement point. P L , Q L Voltage at measurement point L. .

[0026] The specific circuit equation is as follows: Photovoltaic power sources are constant power sources, and their output current is affected by the grid connection point voltage. The loop equations for the circuit shown in the equivalent model of the system are as follows: Considering that traditional power supply side nodes have: In a preferred embodiment, the particle swarm optimization algorithm is used to identify the model parameters. The specific identification process is as follows: Considering the photovoltaic power source as a power source, and incorporating the electromotive force of the equivalent Thevenin circuit of the motor with a specified empirical value. E S , motor reactance X S / resistance R S The ratio k1, photovoltaic power reactive power Q V / meritorious P V Ratio k2; Load reactance X / Load resistance R The ratio k3 gives the real and imaginary parts as follows: exist T Within each measurement group, find a set of parameter values: the electromotive force of the equivalent Thevenin circuit of the motor. E S,x and E S,y (Its modulus is a specified Es), motor reactance X S Photovoltaic power supply active power P V Load reactance X Four parameters. X Making the above two equations hold true within the allowable range during actual operation is a typical optimization problem.

[0027] The parameters to be determined can be expressed as the optimization problem shown in the following formula.

[0028] The upper and lower limits of each parameter are the maximum and minimum values ​​that may occur during actual operation.

[0029] The specific application method of the Particle Swarm Optimization (PSO) algorithm is as follows: First, initialize a group of random particles with a dimension of 4. Each dimension of the particle represents a different parameter to be solved. The upper and lower limits of the particles should be within a reasonable range, and each particle represents a feasible solution. All particles have their own "velocity" and "position". During the solution process, the particles change their velocity and position based on the current optimal solution of the entire population and their own optimal solution. The update formula is as follows: in, v id （k+1） Represents the k-th iteration. i The first particle d The speed of each parameter; x id （k+1） No. k During the nth iteration i The first particle d The position of each parameter; α , ω , c 1, c 2, r 1, r 2 indicates the variable parameters of the algorithm. pbest and gbest These represent the historical best value of the particle and the population best value, respectively. After multiple iterations, the final population best solution is the parameter value to be determined.

[0030] In this embodiment of the invention, step S200 involves updating the operating status of the transformer substation in real time based on the comprehensive equivalent model of the substation and the electrical response data, and obtaining key indicators for the supply and demand balance of the substation operation, including: extracting features from the operating status of the substation. Calculate the composition ratio, equivalent impedance, and dynamic power distribution of the power supply group and load group in the transformer area based on the electrical response data. Using the calculation results as input, the operating status of the transformer area is iteratively updated in real time based on the parameter identification model of particle swarm optimization, and the supply and demand balance parameters are output.

[0031] Reference Figure 2 In the self-consistency assessment and comprehensive judgment phase, the implementation steps include steps one through nine: Step 1: Sampling and storing memory current values. This involves continuously sampling the current between the distributed photovoltaic (PV) power inverter and the distribution network during normal operation, and storing the sampled values ​​as memory current values. These memory current values ​​represent the current characteristics of the PV power distribution area under normal operating conditions.

[0032] Step 2: Calculate the rate of change of current. This involves comparing the sampled current value after the disturbance with the corresponding memory current value to calculate the rate of change of current. The rate of change of current indicates the relative change in current within the photovoltaic power distribution area after the disturbance.

[0033] Step 3: Determine the primary criterion. This involves comparing the calculated rate of change of current with a preset threshold. If the rate of change of current exceeds the set threshold, it indicates that the current has changed significantly, satisfying the primary criterion condition, meaning that the transformer area may be isolated.

[0034] Step 4: Active Scrambling. If the main criterion is not met, the transformer area is actively scrambled. This can be achieved by increasing or decreasing the output current or adjusting the output voltage to elicit a response from the transformer area.

[0035] Step 5: Calculate the new rate of change of current. That is, after active scrambling, the current is sampled again, and the new rate of change of current is calculated.

[0036] Step Six: Determine the Auxiliary Criterion. This involves comparing the newly calculated rate of change of current with the threshold set by the auxiliary criterion. If the new rate of change of current exceeds the threshold, the auxiliary criterion condition is met, meaning the scrambled current change is significant, and islanding may exist.

[0037] Step 7: If the auxiliary criterion is not met, continue the active scrambling operation and repeat steps 4 to 6 until the auxiliary criterion is met or the maximum number of scrambling times is reached.

[0038] Step 8: End Judgment. That is, if the auxiliary criterion is not met even after reaching the maximum number of scrambling attempts, it is determined that the station area has not formed an island and normal operation is restored.

[0039] Step 9: Input the islanding status judgment results of active and passive anti-islanding protection into the self-consistent evaluator, comprehensively analyze the information of active and passive anti-islanding, and then make a decision about the circuit breaker status based on this information.

[0040] Specifically, current memory comparison is used as the primary criterion to establish the current characteristics of the photovoltaic power station under normal operating conditions. This includes the different characteristics of the sampled current under the influence of various factors such as solar radiation intensity and load demand at different times of the day. Current values ​​within different time periods are memorized and categorized for storage, thereby identifying the output current patterns under different conditions and establishing a more representative current benchmark. Specifically, the average value of different time periods is memorized to construct a benchmark used to assess whether the current sampled current value deviates from the normal operating condition.

[0041] Specifically, the comparison refers to calculating the rate of change of current and comparing it with a pre-set threshold. This rate of change of current indicates the relative magnitude of the change in current and is a key indicator for identifying the degree of impact of disturbances.

[0042] The calculation process for the current change rate is as follows: in, The reference value of the memory current during normal operation in a certain period of time. This represents the current sample value after the disturbance during this period.

[0043] It should be noted that the calculation involves the percentage change in current value after disturbance relative to the current value under normal operating conditions. When the percentage exceeds a pre-set threshold, it indicates a significant change in current, suggesting that the transformer area may be becoming isolated. Conversely, if the rate of change in current is below the threshold, it indicates that the current change is not significant and the transformer area is operating normally.

[0044] Specifically, the pre-set threshold is obtained by the following method: The method for calculating the normal maximum current change rate is characterized by the fact that, under different normal operating conditions, the maximum current value obtained through continuous sampling is... Maximum rate of change of current Calculation formula: Furthermore, the introduced coefficients ,and This is a dimensionless constant used to multiply the maximum rate of change of current during normal system operation to determine an appropriate value for the threshold. This coefficient... This reflects the system's stability requirements and its tolerance for abnormal conditions. Among these, for The value should be selected appropriately based on the system's needs and stability requirements. Larger coefficient values ​​provide greater margin but reduce system sensitivity, while smaller coefficient values ​​may increase the risk of malfunctions. By adjusting the coefficients, a balance can be struck between system response speed and stability, thereby ensuring the accuracy of islanding detection.

[0045] Island detection threshold Calculated by multiplying the maximum rate of change of current by a coefficient: When the rate of change of current Greater than the island determination threshold If so, the primary criterion infers the possible existence of an island phenomenon.

[0046] Specifically, the aforementioned current boosting or voltage regulation method employs active scrambling as an auxiliary criterion, namely, using current pulse injection or voltage pulse injection to control the amplitude or phase of the photovoltaic power station current or voltage to periodically change abruptly. When islanding does not occur, the entire system is still affected by the power grid, and the current changes caused by active scrambling under the grid's regulation are weak. However, when islanding occurs, the amplitude or phase of the sampled current in the system also changes abruptly, thus detecting the islanding state.

[0047] Specifically, self-consistency is verified based on active and passive anti-islanding protection signals, and the final circuit breaker tripping signal is output based on the degree of self-consistency.

[0048] In an optional embodiment, a deep learning-based self-consistency detection method can ensure consistency between active and passive anti-islanding protection decisions, thereby determining the on / off state of the grid-connected circuit breaker. Specific details include: Current and voltage data acquired during both active and passive anti-islanding protection are collected and preprocessed. The data includes data from normal operation and islanding operation, and the preprocessing includes denoising, normalization, and time series reconstruction.

[0049] Extract key features from current and voltage data, including the frequency, amplitude, and phase of the current waveform. Construct a deep learning model to learn the relationship between the current and voltage characteristics of the photovoltaic substation and the islanding state.

[0050] In real-time operation, the current current and voltage data are input into the trained deep learning model. The model will output a consistency score, which represents the probability that the current system is in an islanded state.

[0051] The self-consistency score output by the deep learning model is compared with the results of active and passive anti-islanding protection. If the self-consistency score of the deep learning model is consistent with or close to the results of active and passive protection, the decision system maintains its current state. If the self-consistency score is inconsistent with the results of active and passive protection, the system can trigger further detection or circuit breaker control to ensure the safe operation of the power grid.

[0052] In another alternative embodiment, a deep learning model is trained using the collected data via a convolutional neural network, including the construction of convolutional layers, pooling layers, and fully connected layers.

[0053] In an optional embodiment, data preprocessing can use Markov transform fields to convert the one-dimensional time-series current and voltage data input to the convolutional neural network into spatial image data. The generated potential spatial image data is then passed as input to the convolutional neural network.

[0054] In this embodiment of the invention, step S300, which calculates the passive and active criteria for the operation of the distribution area and determines the operating status of the distribution area based on the active and passive criteria, includes: during the normal operation of the distributed photovoltaic power distribution area, continuously sampling the inverter output current and storing it according to a preset time period to form a memory current reference value for the corresponding time period. When a disturbance is detected in the transformer area, the current sampling value for the current period is obtained, and the current change rate is calculated based on the memory current reference value. If the current change rate is greater than the active criterion threshold, the transformer area is determined to be in an abnormal operating state; otherwise, the transformer area is in a normal operating state. After active disturbance, the current is sampled again to calculate the new current change rate. If the new current change rate is greater than the passive criterion threshold, the transformer area is determined to be in abnormal operation. If the new current change rate is not greater than the passive criterion threshold, active scrambling is performed until the passive criterion is met or the maximum number of scrambling operations is reached. If the current change rate is still not greater than the passive criterion threshold after the maximum number of scrambling operations is reached, the transformer area is determined to be in normal operation.

[0055] In this embodiment of the invention, step S300, calculating the consistency metric value of the active / passive criterion judgment result, includes... Acquire current and voltage data of active and passive photovoltaic power distribution areas during normal and abnormal operation, and extract key features of current and voltage data to build a deep learning model for learning the relationship between current and voltage characteristics of photovoltaic power distribution areas and islanding status. The current and voltage data of the current transformer area during operation are input into the trained deep learning model, and the self-consistency score output by the deep learning model is compared with the results of active and passive anomaly protection. If the self-consistency score of the deep learning model is consistent with or similar to the results of active and passive protection, the decision system continues to maintain its current state; if the self-consistency score is inconsistent with the results of active and passive protection, the comprehensive protection mechanism is triggered.

[0056] Reference Figure 3 In the fault isolation and self-healing control phase, the system adopts a novel automatic reclosing and branch-by-branch recovery strategy based on extended voltage time, including steps one through five: Step 1: When a short circuit fault occurs in the line connected to a certain disconnector in the photovoltaic area, first quickly disconnect the circuit breaker upstream of the disconnector and all disconnectors on the busbar connected to it to isolate the faulty line.

[0057] Step 2: Enter the automatic reclosing stage, characterized in that, after closing the circuit breaker, reclosing begins sequentially from the smallest numbered disconnector connected to the circuit breaker. Assume the currently reclosing disconnector is the i-th disconnector.

[0058] Step 3: After each reclosing of a disconnect switch, simultaneously detect the voltage or current signal of that line.

[0059] Step 4: Determine if the voltage or current signal is abnormal. If low voltage or overcurrent is detected, it indicates a non-transient fault on the line. In this case, disconnect the disconnect switch again without reclosing it, and continue with subsequent disconnect switch reclosing operations. If the voltage or current signal is normal, i.e., no abnormality is detected, the line is restored to normal power supply.

[0060] Step 5: Determine if the circuit breaker has been reclosed to the last disconnector switch n. If it has, the fault has been isolated. If it has not been reclosed to the last disconnector switch, reclose the (i+1)th disconnector switch and return to Step 4.

[0061] Furthermore, the characteristics of voltage or current signals detected by sensors in the photovoltaic area; Based on the detected information, the voltage or current changes are analyzed and the occurrence of a fault is identified. Then, the power supply to the faulty line is cut off. The feature is that the circuit breaker upstream of the fault disconnecting switch and all disconnecting switches connected to the circuit breaker are disconnected.

[0062] The system performs automatic reclosing operations, controlling the actions of corresponding disconnecting switches, circuit breakers, or other isolation devices to isolate the faulty area and restore normal power supply to other areas. Its key feature is that it closes the circuit breaker and then sequentially delays the closing of the disconnecting switches in the power supply network, identifying the faulty line by detecting the characteristics of the voltage or current on the reclosed line.

[0063] In an optional embodiment, an automatic reclosing and branch-by-branch recovery strategy based on extended voltage time is adopted. The feature is that, based on a certain reclosing procedure, the disconnect switches are closed one by one after a delay. By judging the magnitude of the voltage or current of each line segment, the normal power supply of non-faulty lines can be restored, and the specific location of the faulty line in the photovoltaic area can be accurately identified and isolated.

[0064] It should be noted that the extended voltage-time strategy is essentially based on comparing the duration of voltage anomalies in the photovoltaic area with a preset time threshold, which represents the upper limit of the allowable duration of voltage anomalies in order to identify faults.

[0065] In this embodiment of the invention, step S400, when the consistency metric value of the active and passive criterion judgment result is greater than a preset threshold, further includes triggering the comprehensive protection mechanism as follows: When the consistency metric value of the active and passive criteria judgment results is greater than the preset threshold, the circuit breaker and disconnect switch of the corresponding branch are disconnected; after the main circuit breaker is closed, starting from the smallest numbered branch, the reclosing operation is performed on each branch in sequence; after each reclosing, the voltage and current of the branch are detected in real time. If a low voltage or overcurrent is detected in a branch and the duration exceeds a preset time threshold, the branch is determined to have a permanent fault, and the branch is disconnected again. If the voltage and current return to normal, the circuit is kept closed, and the reclosing operation is performed on the next branch until all branches have been detected, thus achieving step-by-step self-healing recovery.

[0066] It should be noted that the preset time threshold in the embodiments of the present invention can be set to different values ​​under different usage scenarios. For example, when the proportion of residential / photovoltaic is high and the impact load is low, it can be set to about 0.2s (about 10 cycles); when there is a certain amount of motor / impact load, it can be set to 0.4s–0.6s (about 20–30 cycles); when the start-up of industrial and commercial motors is obvious, it can be set to 0.8s–1.2s (about 40–60 cycles).

[0067] It should be noted that the final time threshold needs to be set in conjunction with the line length of the transformer area, load type, inverter control characteristics and protection coordination, in order to avoid misjudging short-term closing inrush current / start-up drop as a permanent fault, while ensuring that permanent faults can be quickly identified and disconnected.

[0068] It is worth noting that this invention achieves dynamic characterization of the power supply and load groups in a photovoltaic distribution area by actively scrambling and obtaining equivalent parameters of the distribution area through online identification. This provides a basis for adaptive setting of protection criteria and thresholds, thereby improving the pertinence and sensitivity of protection judgment. By performing consistency verification of active and passive criteria through a self-consistent evaluator, the probability of false operation and failure to operate is significantly reduced, improving the reliability and robustness of islanding detection and fault identification. Based on the automatic reclosing and branch-by-branch recovery strategy using extended voltage time, dynamic differentiation and step-by-step self-healing of transient and permanent faults can be achieved without significantly increasing the number of measuring devices, improving the automation efficiency and recovery speed of fault handling. The energy storage-coordinated unloading and feedback control provides an effective means for energy absorption during faults and energy balance during recovery, which is beneficial for suppressing overvoltage on the DC side and grid-connected side, improving equipment safety, and enhancing energy utilization efficiency.

[0069] Example 2: The above example is an illustrative scheme of a comprehensive protection and self-healing method for photovoltaic power distribution areas. It should be noted that the technical solution of this comprehensive protection and self-healing system for photovoltaic power distribution areas belongs to the same concept as the technical solution of the aforementioned comprehensive protection and self-healing method for photovoltaic power distribution areas. Details not described in detail in this example can be found in the description of the technical solution of the aforementioned comprehensive protection and self-healing method for photovoltaic power distribution areas.

[0070] This embodiment presents a comprehensive protection and self-healing system for photovoltaic (PV) distribution areas, applicable to PV distribution networks containing multiple distributed power sources and energy storage units. The system includes: The data acquisition module is used to collect electrical response data of the photovoltaic power station area before and after disturbance, and to establish a comprehensive equivalent model of the power station area; The update and optimization module is used to update the operating status of the transformer area in real time based on the comprehensive equivalent model of the transformer area and the electrical response data, so as to obtain the key indicators of supply and demand balance in the operation of the transformer area. The judgment module is used to calculate the passive and active criteria during the operation of the transformer area, and to judge the operating status of the transformer area based on the active and passive criteria, and to calculate the consistency metric value of the judgment result of the active and passive criteria. The fault isolation and self-healing control module is used to trigger the integrated protection mechanism when the consistency metric value of the judgment result of the active and passive criteria is greater than the preset threshold, execute the fault isolation control of the transformer area, and test each branch in sequence after the main power is restored to achieve self-healing recovery of the non-faulty area.

[0071] This embodiment also provides an electronic device applicable to the integrated protection and self-healing methods for photovoltaic power distribution areas, including: The system includes a memory and a processor. The memory stores computer-executable instructions, and the processor executes these instructions to implement the comprehensive protection and self-healing method for photovoltaic power distribution areas as proposed in the above embodiments.

[0072] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the method for comprehensive protection and self-healing of photovoltaic power distribution areas as proposed in the above embodiments.

[0073] The storage medium proposed in this embodiment belongs to the same inventive concept as the method for realizing comprehensive protection and self-healing of photovoltaic power stations proposed in the above embodiments. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0074] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0075] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A photovoltaic grid integrated protection and self-healing method, characterized in that, The method, applicable to photovoltaic distribution networks containing multiple distributed power sources and energy storage units, includes: Collect electrical response data of the photovoltaic power station area before and after disturbance, and establish a comprehensive equivalent model of the power station area; Based on the comprehensive equivalent model of the transformer area and the electrical response data, the operating status of the transformer area is updated in real time to obtain key indicators of supply and demand balance in the transformer area. Calculate the passive and active criteria for the operation of the transformer area, determine the operating status of the transformer area based on the active and passive criteria, and calculate the consistency metric value of the determination result of the active and passive criteria. When the consistency metric value of the active and passive criterion judgment results is greater than the preset threshold, the integrated protection mechanism is triggered, the fault isolation control of the transformer area is executed, and the branch circuits are tested sequentially after the main power supply is restored to achieve self-healing recovery of the non-faulty area.

2. The comprehensive protection and self-healing method for photovoltaic power distribution areas as described in claim 1, characterized in that, Also includes: When the operating status of the transformer substation is abnormal, if a drop in the AC voltage of the power grid is detected, the energy storage unit is controlled to enter the energy absorption mode; during the power grid voltage recovery phase, the energy storage unit is controlled to switch to the slow release mode, and the energy feedback speed is dynamically adjusted based on the change rate of the total active power of the transformer substation.

3. The comprehensive protection and self-healing method for photovoltaic power distribution areas as described in claim 2, characterized in that, When the consistency metric value of the active and passive criterion judgment results is greater than a preset threshold, the comprehensive protection mechanism is further triggered by: When the consistency metric value of the active and passive criteria judgment results is greater than the preset threshold, the circuit breaker and disconnect switch of the corresponding branch are disconnected; after the main circuit breaker is closed, starting from the smallest numbered branch, the reclosing operation is performed on each branch in sequence; after each reclosing, the voltage and current of the branch are detected in real time. If a low voltage or overcurrent is detected in a branch and the duration exceeds a preset time threshold, the branch is determined to have a permanent fault, and the branch is disconnected again. If the voltage and current return to normal, the circuit is kept closed, and the reclosing operation is performed on the next branch until all branches have been detected, thus achieving step-by-step self-healing recovery.

4. The comprehensive protection and self-healing method for photovoltaic power distribution areas as described in claim 3, characterized in that, The calculation of passive and active criteria during the operation of the distribution area, and the determination of the operating status of the distribution area based on the active and passive criteria, includes: during the normal operation of the distributed photovoltaic power distribution area, continuously sampling the inverter output current and storing it according to a preset time period to form a memory current reference value for the corresponding time period. When a disturbance is detected in the transformer area, the current sampling value for the current period is obtained, and the current change rate is calculated based on the memory current reference value. If the current change rate is greater than the active criterion threshold, the transformer area is determined to be in an abnormal operating state; otherwise, the transformer area is in a normal operating state. After active disturbance, the current is sampled again to calculate the new current change rate. If the new current change rate is greater than the passive criterion threshold, the transformer area is determined to be in abnormal operation. If the new current change rate is not greater than the passive criterion threshold, active scrambling is performed until the passive criterion is met or the maximum number of scrambling operations is reached. If the current change rate is still not greater than the passive criterion threshold after the maximum number of scrambling operations is reached, the transformer area is determined to be in normal operation.

5. The comprehensive protection and self-healing method for photovoltaic power distribution areas as described in claim 4, characterized in that, The calculation of the consistency metric value of the active and passive criterion judgment results includes: Acquire current and voltage data of active and passive photovoltaic power distribution areas during normal and abnormal operation, and extract key features of current and voltage data to build a deep learning model for learning the relationship between current and voltage characteristics of photovoltaic power distribution areas and islanding status. The current and voltage data of the current transformer area during operation are input into the trained deep learning model, and the self-consistency score output by the deep learning model is compared with the results of active and passive anomaly protection. If the self-consistency score of the deep learning model is consistent with or similar to the results of active and passive protection, the decision system continues to maintain its current state; if the self-consistency score is inconsistent with the results of active and passive protection, the comprehensive protection mechanism is triggered.

6. The comprehensive protection and self-healing method for photovoltaic power distribution areas as described in claim 5, characterized in that, Based on the comprehensive equivalent model of the transformer substation and the electrical response data, the operating status of the transformer substation is updated in real time, and the key indicators of supply and demand balance in the operation of the transformer substation are obtained, including feature extraction of the operating status of the transformer substation. Calculate the composition ratio, equivalent impedance, and dynamic power distribution of the power supply group and load group in the transformer area based on the electrical response data. Using the calculation results as input, the operating status of the transformer area is iteratively updated in real time based on the parameter identification model of particle swarm optimization, and the supply and demand balance parameters are output.

7. The comprehensive protection and self-healing method for photovoltaic power distribution areas as described in claim 6, characterized in that, The electrical response data includes voltage, current, and power before and after the disturbance.

8. A comprehensive protection and self-healing system for photovoltaic power distribution areas, applied to the method described in any one of claims 1-7, characterized in that, The system is applicable to photovoltaic distribution networks containing multiple distributed power sources and energy storage units, and includes: The data acquisition module is used to collect electrical response data of the photovoltaic power station area before and after disturbance, and to establish a comprehensive equivalent model of the power station area; The update and optimization module is used to update the operating status of the transformer area in real time based on the comprehensive equivalent model of the transformer area and the electrical response data, so as to obtain the key indicators of supply and demand balance in the operation of the transformer area. The judgment module is used to calculate the passive and active criteria during the operation of the transformer area, and to judge the operating status of the transformer area based on the active and passive criteria, and to calculate the consistency metric value of the judgment result of the active and passive criteria. The fault isolation and self-healing control module is used to trigger the integrated protection mechanism when the consistency metric value of the judgment result of the active and passive criteria is greater than the preset threshold, execute the fault isolation control of the transformer area, and test each branch in sequence after the main power is restored to achieve self-healing recovery of the non-faulty area.

9. An electronic device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, they implement the steps of the photovoltaic area integrated protection and self-healing method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the photovoltaic area integrated protection and self-healing method according to any one of claims 1 to 7.

Images