Intelligent identification and smooth switching control method, system and equipment for isolated island operation mode of power distribution network, and medium

Abstract

The invention discloses a power distribution network island operation mode intelligent identification and smooth switching control method, system and device and a medium, and belongs to the technical field of intelligent power distribution network operation and control, and the method comprises the steps: collecting real-time operation data of key nodes of a power distribution network, carrying out multiple quality verification, extracting characteristic parameters, building an island identification classifier, and carrying out the smooth switching of the key nodes of the power distribution network. Judging an island identification result, and starting an island stability control strategy to keep the island operation stable after the formation of the island is confirmed; detecting whether a main network recovers power supply during stable operation of the island, entering a grid-connected preparation stage, entering a grid-connected closing stage after synchronization is completed, sending a closing instruction to perform grid-connected switching, performing performance evaluation and offline optimization, obtaining optimal control parameters, and updating the optimal control parameters to a control system. According to the method, rapid and accurate identification of the island is realized, power imbalance in the island is rapidly responded, the stability of island operation is improved, smooth and impact-free grid connection is realized, impact on electrical equipment is avoided, and the grid connection success rate is improved.

Description

Technical Field

[0001] This invention relates to the field of operation and control technology of smart distribution networks, specifically to intelligent identification and smooth switching control methods, systems, equipment and media for islanded operation modes of distribution networks. Background Technology

[0002] As the penetration rate of distributed generation in distribution networks continues to increase, when a fault occurs in the main grid, local areas of the distribution network containing distributed generation may form islands and continue to operate. Existing islanding operation control technologies mainly have the following shortcomings: Islanding detection is slow and inaccurate: Traditional islanding detection methods mainly include passive detection and active detection. Passive detection identifies islanding by monitoring changes in electrical quantities such as voltage and frequency, but it has blind spots when the load and distributed power supply are in balance, resulting in a high failure rate.

[0003] Insufficient stability control capability in islanded operation: After islanding is formed, the voltage and frequency within the island are prone to large fluctuations due to the loss of main grid support. Existing control methods mainly focus on droop control for single distributed power sources, lacking coordinated optimization for multiple distributed power sources.

[0004] Islanded grid connection switching experiences significant impacts: When the main grid recovers from a fault, the islanded system needs to reconnect to the main grid. Traditional grid connection methods use simple voltage and frequency detection, directly closing the circuit when the voltage amplitude, phase, and frequency of the islanded system and the main grid are close. Due to the lack of precise synchronization control, there are often large phase and frequency differences at the moment of grid connection, generating significant inrush currents, with peak values ​​reaching 3-5 times the rated current. This impacts electrical equipment and affects power supply reliability. Summary of the Invention

[0005] In view of the above-mentioned problems, the present invention is proposed.

[0006] Therefore, the present invention aims to achieve rapid and accurate identification of islands, ensure stable voltage and frequency during island operation, and achieve smooth and shock-free grid connection between islands and the main grid.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a method for intelligent identification and smooth switching control of islanded operation mode in distribution networks, comprising, Real-time operational data of key nodes in the distribution network is collected, and multiple quality checks are performed on the collected data. The checked data is smoothed, filtered, and feature parameters are extracted and normalized. An islanding identification classifier is established to determine the islanding identification result and record the islanding formation time. Once islanding is confirmed, the islanding stability control strategy is activated, and the active power output command of the distributed power source is output and issued to maintain stable islanding operation. During stable islanding operation, it is checked whether the main grid has restored power supply. After the main grid is restored, it enters the grid connection preparation stage and performs grid connection synchronization, and the grid connection switching operation is triggered upon synchronization. After synchronization is completed, it enters the grid connection closing stage, issues a closing command to perform grid connection switching, and performs performance evaluation and offline optimization on islanding operation and grid connection switching to obtain the optimal control parameters and update them to the control system.

[0008] As a preferred embodiment of the intelligent identification and smooth switching control method for islanded operation mode of distribution network described in this invention, the multiple quality verification includes: collecting real-time operation data of key nodes of the distribution network, including electrical quantity information and switching quantity information, and setting the data collection frequency; The collected data undergoes multiple quality checks, including data integrity check, data rationality check, and data consistency check. Data that passes all three checks is stored in the data cache.

[0009] As a preferred embodiment of the intelligent identification and smooth switching control method for islanded operation mode of distribution network described in this invention, the extraction of feature parameters includes: smoothing the data using a sliding window filtering method; and extracting feature parameters reflecting the islanded state after filtering. The feature vector is constructed based on the feature parameters, and then the feature vector is normalized.

[0010] As a preferred embodiment of the intelligent identification and smooth switching control method for islanded operation mode of distribution network described in this invention, the establishment of the island identification classifier includes using the extracted feature vectors and employing a support vector machine algorithm to establish the island identification classifier. By training samples from different scenarios using historical operational data, the optimal parameters of the support vector machine can be determined. Determine the island identification result, identify the island identification flag, immediately trigger island stabilization control, and record the island formation time.

[0011] The beneficial effects of the preferred technical solution in the embodiments of the present invention are as follows: by fusing multi-source electrical quantity information, an island intelligent recognition based on feature vectors is established, overcoming the blind spot problem of traditional passive detection and the slow speed problem of active detection.

[0012] As a preferred embodiment of the intelligent identification and smooth switching control method for islanded operation mode of distribution network described in this invention, the islanded stability control strategy includes: adopting adaptive droop control as the basic control strategy and introducing virtual inertia to improve dynamic response characteristics; the basic control strategy includes the frequency-active power droop characteristic and voltage-reactive power droop characteristic of adaptive droop control. Virtual inertia control is introduced to improve the dynamic response characteristics during island formation. An inertia compensation term is added to the frequency control stage. In the formula, For inertia compensation power, For virtual inertia coefficient, The rate of change of frequency; The final active power output command for distributed power sources is: In the formula, This is an instruction to contribute effort. This is a frequency reference value. The droop coefficient is adaptively adjusted. The output frequency is set; the command is sent through the communication network to the grid-connected inverters of each distributed power source in the island, and the inverters adjust their output power after receiving the command. Continuously monitor the voltage and frequency within the island, and calculate the voltage and frequency deviations: In the formula, The percentage of voltage deviation. This is the voltage reference value. For frequency deviation; after determining that the islanded operation is stable, the stability control continues to run until the main network is detected to be restored or a grid connection preparation command is received.

[0013] The beneficial effects of the preferred technical solution in the embodiments of the present invention are as follows: a multi-distributed power source coordinated control model is established, and a strategy combining adaptive droop control and virtual synchronous machine control is used to quickly respond to power imbalances within the island, thereby improving the stability and reliability of island operation.

[0014] As a preferred embodiment of the intelligent identification and smooth switching control method for islanded operation mode of distribution networks described in this invention, the grid connection preparation stage includes: continuously monitoring the operating status of the main grid during stable islanded operation to detect whether the main grid has resumed power supply; and monitoring the voltage amplitude of the main grid through a voltage sensor upstream of the circuit breaker. When detected in 10 consecutive sampling periods When the main network is determined to have recovered, the main network recovery flag is set. Set to 1, where, This is the rated voltage of the power grid; After the main grid is restored, the grid connection preparation phase begins. It is necessary to predict the synchronization conditions between the islanded side and the main grid side, and calculate the voltage amplitude difference between the islanded side and the main grid side. Frequency difference and phase angle difference ; The grid connection synchronization conditions are defined as follows: When the synchronization condition is met, the synchronization flag is set. Set to 1 to allow grid connection; when the main grid recovers and the islanding is stable but the synchronization conditions are not met, the dual-side synchronization control strategy is activated. The goal of the synchronization control is to track the main grid parameters by adjusting the voltage, frequency and phase on the islanding side to meet the grid connection synchronization conditions. Frequency synchronization control uses a proportional-integral controller, and the frequency adjustment command is as follows: In the formula, This is a frequency adjustment command. This is the frequency scaling factor. This is the frequency integral coefficient; the frequency adjustment command is superimposed on the frequency control loop to correct the frequency reference value, and through frequency adjustment, the islanded frequency becomes the main network frequency; Voltage amplitude synchronization control also uses a proportional-integral controller: In the formula, This is a voltage regulation command. This is the voltage proportionality coefficient. This is the voltage integral coefficient, used by the voltage regulation command to correct the voltage reference value. A phase angle feedforward compensation strategy is adopted to directly adjust the phase on the islanded side based on the phase angle difference: In the formula, This is the corrected island-side phase angle. The phase angle compensation coefficient is used to adjust the phase angle by modifying the phase of the reference signal of the grid-connected inverter. The synchronization control commands are continuously updated to track changes in main network parameters in real time, calculate synchronization deviations, and determine whether synchronization is complete.

[0015] The beneficial effects of the preferred technical solution in the embodiments of the present invention are as follows: the design of a dual-side synchronous control strategy for the island and the main grid, through the coordinated cooperation of voltage tracking on the main grid side and frequency regulation on the island side, avoids impact on electrical equipment and improves power supply continuity.

[0016] As a preferred embodiment of the intelligent identification and smooth switching control method for islanded operation mode of distribution networks described in this invention, wherein: the grid connection closing stage includes, when When entering the grid connection and closing phase, a closing timing selection strategy is adopted to reduce the closing impact. The instantaneous voltage difference between the islanded side and the main grid side is calculated: In the formula, For instantaneous voltage difference, and These are the instantaneous voltages on the islanded side and the main grid side, respectively; the zero-crossing point of the instantaneous voltage difference is monitored to determine the optimal closing time; A closing command is issued to the grid-connected circuit breaker at the optimal closing time, taking into account the circuit breaker's operating delay, and is initiated one circuit breaker operating time in advance. Issue closing command: In the formula, The time when the closing command is issued. For the predicted voltage zero-crossing time, This refers to the inherent operating time of the circuit breaker; After the circuit breaker is closed, monitor the grid connection point current and calculate the inrush current multiple as an evaluation indicator of control effectiveness; after grid connection is completed, the islanding identification flag is set. Reset to 0, the system exits islanded operation mode and returns to normal grid-connected operation; A performance evaluation is conducted on the entire islanding operation and grid connection switching process, and key performance indicators are defined, including islanding detection time, islanding operation stability, synchronization time, and grid connection impact. Based on key performance indicators, a genetic algorithm is used to optimize key control parameters offline, including droop coefficient, virtual inertia coefficient, and synchronous controller parameters; the optimization objective function is: In the formula, For comprehensive performance indicators, For the weighting coefficients assigned to different parameters, For the island detection time, As a stability indicator, To synchronize time, The factor is the multiple of the inrush current; by minimizing the comprehensive performance index, the optimal control parameters are obtained and updated to the actual control system.

[0017] Another objective of this invention is to provide an intelligent identification and smooth switching control system for islanded operation modes in power distribution networks.

[0018] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a distribution network islanding operation mode intelligent identification and smooth switching control system, comprising: a data acquisition and preprocessing module, a feature extraction module, an islanding intelligent identification module, an islanding stability control module, a synchronization control module, and a grid connection switching module; The data acquisition and preprocessing module collects real-time operating data of key nodes in the power distribution network and performs multiple quality checks on the collected data. The feature extraction module smooths the verified data, filters it, extracts feature parameters, and performs normalization processing. The island intelligent identification module establishes an island identification classifier, determines the island identification result, and records the time when the island is formed. The island stability control module, upon confirming the formation of an island, initiates the island stability control strategy, outputs and issues active power output commands from the distributed power source, and maintains the stable operation of the island. During the stable operation of the island, the synchronization control module detects whether the main grid has restored power supply. After the main grid is restored, it enters the grid connection preparation stage and performs grid connection synchronization. The synchronization is completed and the grid connection switching operation is triggered. After synchronization is completed, the grid-connected switching module enters the grid-connected closing stage, issues a closing command to perform grid-connected switching, performs performance evaluation and offline optimization of islanded operation and grid-connected switching, and obtains the optimal control parameters to update the control system.

[0019] The present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, characterized in that the processor executes the computer program to implement the steps of the intelligent identification and smooth switching control method for islanded operation mode of the power distribution network.

[0020] The present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that, when the computer program is executed by a processor, it implements the steps of the intelligent identification and smooth switching control method for islanded operation mode of the power distribution network.

[0021] The beneficial effects of this invention: By integrating multi-dimensional features such as voltage, frequency, phase angle, and power, and employing a support vector machine intelligent recognition algorithm, this invention achieves rapid and accurate identification of isolated islands.

[0022] This invention employs a strategy combining adaptive droop control and virtual inertia control to quickly respond to power imbalances within an island, thereby improving the stability of islanded operation.

[0023] This invention achieves smooth, shock-free grid connection through dual-sided coordinated synchronous control on the island side and the main grid side, avoiding impact on electrical equipment and improving the grid connection success rate. Attached Figure Description

[0024] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0025] Figure 1 The above is a flowchart of the intelligent identification and smooth switching control method for islanded operation mode of distribution network provided in one embodiment of the present invention. Detailed Implementation

[0026] 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.

[0027] Example 1, referring to Figure 1 This is one embodiment of the present invention, which provides a method for intelligent identification and smooth switching control of islanded operation mode in a distribution network, including: S100: Collects real-time operating data of key nodes in the power distribution network and performs multiple quality checks on the collected data. S200. Smooth the verified data, filter it, extract the feature parameters, and normalize them. S300. Establish an island identification classifier, determine the island identification result, and record the time when the island is formed. S400: Once islanding is confirmed, the islanding stability control strategy is activated, and the active power output command of the distributed power source is output and issued to maintain the stable operation of the island. During the stable operation of the S500 island, it detects whether the main grid has restored power supply. After the main grid is restored, it enters the grid connection preparation stage and performs grid connection synchronization, and synchronously completes the grid connection switching operation. After S600 synchronization is completed, it enters the grid connection and closing stage, issues a closing command to perform grid connection switching, performs performance evaluation and offline optimization on islanded operation and grid connection switching, and obtains the optimal control parameters to update the control system. It should be noted that existing technologies have technical defects such as inaccurate island detection, unstable island operation, and large grid connection impact.

[0028] Therefore, to address the aforementioned problems, through steps S100-S600, this invention collects electrical quantity data such as voltage, current, frequency, and phase angle at various measuring points using a distribution network monitoring device, and performs quality verification and feature extraction; it establishes an islanding identification model using a support vector machine algorithm, and quickly determines whether an island has formed through pattern recognition of multi-dimensional feature vectors; when an island is detected, an islanding stability control strategy is immediately initiated, adjusting the active and reactive power output of distributed power sources through adaptive droop control, and introducing virtual inertia to improve dynamic response characteristics and maintain voltage and frequency stability within the island; when the main grid fault is recovered, the system enters the grid connection preparation stage, adjusting the voltage amplitude, frequency, and phase on the island side through a dual-side synchronization controller to ensure precise synchronization with the main grid side, and closing the circuit breaker to connect to the grid when the synchronization conditions are met, achieving smooth switching.

[0029] Example 2, refer to Figure 1 This is one embodiment of the present invention, which provides a method for intelligent identification and smooth switching control of islanded operation mode in a distribution network, including: In this embodiment of the invention, S100 involves collecting real-time operational data of key nodes in the power distribution network and performing multiple quality checks on the collected data, including the following steps S101-S103: S101. Real-time operating data of key nodes in the distribution network is collected through the communication interfaces of the distribution automation terminal, smart meter and distributed power grid-connected inverter. The real-time operating data includes electrical quantities and switching quantities. Electrical quantities include three-phase voltage amplitude, three-phase current amplitude, frequency, phase angle, active power, and reactive power; Switching quantities include information such as circuit breaker status and distributed power supply operating status.

[0030] The data acquisition frequency is set to 20 times per second, that is, a set of data is collected every 50 milliseconds to ensure that the rapid transient process when the island is formed can be captured.

[0031] To ensure data reliability, multiple quality checks are performed on the collected data, including data integrity check, data rationality check, and data consistency check. First, perform a data integrity check, defining the data integrity coefficient as follows: In the formula, The data integrity coefficient. This represents the actual number of data packets received. The number of data packets to be received; when If the data for the current time period is incomplete, data retransmission or data estimation from adjacent measurement points will be initiated.

[0032] S102. Then, perform data rationality verification. Taking voltage data as an example, the voltage rationality criterion is defined as follows: In the formula, Rated voltage, For measuring voltage; For data that exceeds the reasonable range, it is marked as an outlier and a data correction process is triggered, which is then corrected by linear interpolation of data from adjacent time points.

[0033] S103. Finally, perform data consistency verification. For different measuring points on the same busbar, calculate the voltage consistency deviation: In the formula, For voltage consistency deviation, and Voltage values ​​at different measuring points on the same busbar; when At that time, it was determined that there was a consistency problem with the data, and further verification of the measuring device was required; The data, after passing triple verification, is stored in the data cache area and serves as the basic input for subsequent feature extraction and island identification.

[0034] In this embodiment of the invention, step S200 involves smoothing the verified data, filtering it, extracting feature parameters, and performing normalization, including the following steps S201-S202: S201. The raw data collected contains rich information on power grid operation, but it has differences in dimensions and noise interference. Therefore, in the embodiments of the present invention, the data is smoothed, including the following steps A1: A1. The sliding window filtering method is used to smooth the data. The filter window length is set to 5 sampling points, i.e. 250 milliseconds. This can eliminate high-frequency noise and retain the dynamic characteristics when islands are formed. In an optional implementation, the smoothing process in S201 can employ a moving average filter to smooth the data.

[0035] In another alternative implementation, the smoothing process in S201 can also employ median filtering to smooth the data.

[0036] After filtering, it is necessary to extract the characteristic parameters that reflect the islanding state. Six key features are extracted to form a feature vector, which specifically include voltage change rate, frequency change rate, phase angle jump variable, active power imbalance, reactive power imbalance, and voltage harmonic distortion rate. Specifically, the rate of change of voltage is defined as: In the formula, The rate of change of voltage. The voltage at the current moment. The voltage at the previous moment. The sampling interval; When an isolated grid is formed, the voltage usually changes abruptly due to the loss of the main grid support, and the rate of voltage change increases significantly.

[0037] The rate of change of frequency is defined as: In the formula, The rate of change of frequency, The frequency at the current moment. The frequency of the previous moment; Once an island is formed, the frequency will change due to the imbalance of load power, and the rate of frequency change is an important identification feature.

[0038] Active power imbalance is defined as: In the formula, For active power imbalance, For the total output of distributed power sources within the isolated island, This represents the total load power within the isolated island. Before the island is formed, the power is balanced by the main grid, and the imbalance is close to zero; Once an island is formed, the imbalance will increase due to the loss of the main network.

[0039] In an embodiment of the present invention, S202, since the dimensions and numerical ranges of each feature are different, normalization processing is required, including the following steps B1-B3: B1. Using the maximum-minimum value normalization method: In the formula, Here, x represents the normalized eigenvalues, and x represents the original eigenvalues. and These are the historical minimum and maximum values ​​of this feature, respectively. B2. The normalized feature values ​​range from 0 to 1, eliminating the influence of dimensions and facilitating subsequent pattern recognition. B3. Extract and normalize the six-dimensional feature vector. As input for island identification.

[0040] In an optional implementation, the normalization process in S202 can employ a Z-score normalization method based on statistical distribution.

[0041] In another alternative implementation, the normalization process in S202 can also employ a range scaling method based on decimal scaling.

[0042] In this embodiment of the invention, S300 establishes an island identification classifier, determines the island identification result, and records the island formation time, including the following steps S301-S302: S301. Using the extracted feature vectors, a support vector machine (SVM) algorithm is used to establish an island identification classifier. The SVM can effectively handle nonlinear classification problems by finding the optimal classification hyperplane in the high-dimensional feature space, and is suitable for complex pattern recognition tasks such as island identification.

[0043] In an embodiment of the present invention, establishing an island identification classifier includes the following step C1: C1. Constructing a support vector machine model using radial basis kernel functions: In the formula, For kernel function, and For feature vectors, These are kernel function parameters; In an optional implementation, the island identification classifier established in S301 can be built using the random forest algorithm. The random forest model is trained offline using a sample set of distribution network operation feature vectors, generated historically or through simulation and labeled as either islanded or non-islanded. During training, multiple training subsets are generated using a bootstrap sampling method, each subset used to independently train a decision tree. When splitting a decision tree node, a feature subset is randomly selected from all six features for evaluation to choose the optimal split point. The real-time, normalized six-dimensional feature vector obtained in step S202 is input into the trained random forest classifier to determine whether a node is islanded or non-islanded. The results of all decision trees are statistically analyzed, and the final classification result is determined using a majority voting method. However, this implementation generates a relatively complex model structure and a large model file size, which poses a challenge when deployed and run on edge distribution terminals with extremely limited computing and storage resources.

[0044] In another optional implementation, the island identification classifier established in S301 can also employ the K-nearest neighbor algorithm. This involves collecting and storing a large number of historical or simulated distribution network operation feature vector samples with precisely labeled states, forming a feature sample database covering multiple operating conditions. This database serves as a reference for classification. In the real-time identification phase, the real-time normalized feature vector generated in step S202 is used to calculate the distance between it and all samples in the feature sample database. After calculation, the samples in the database are sorted in ascending order of distance, and the K closest historical samples are selected. The state labels corresponding to these K samples are then examined. The number of samples belonging to the island state is counted. If the number exceeds a preset threshold, the current real-time operating state is determined to be islanded, and this moment is recorded; otherwise, it is determined to be a non-islanded state. However, this implementation requires real-time distance calculation and comparison with a massive amount of historical samples during real-time identification. When the sample database is large, it consumes high online computing speed and memory resources, affecting the real-time performance of the identification.

[0045] The optimal parameters of the support vector machine are determined by training with a large amount of historical operating data, including samples of different scenarios such as normal operation, fault operation, and isolated operation. The training samples include 2000 sets of normal operation data and 800 sets of isolated operation data. Parameters are optimized using cross-validation. Finally determined ; S302. For newly acquired feature vectors The support vector machine outputs the discriminant function value: In the formula, Judgment results For Lagrange multipliers, Here, b represents the class label of the training samples, b is the bias term, and n is the number of support vectors. when When it is determined to be in an isolated state, This is considered a normal grid connection status. To improve the reliability of identification, the final determination is based on three consecutive consistent discrimination results. That is, only when the discrimination results of three consecutive sampling periods are all in an island state can the formation of an island be finally confirmed, and the island identification flag is then set. Set to 1 if the value is not specified, otherwise set to 0. This continuous discrimination mechanism can effectively avoid misjudgments caused by accidental disturbances and keep the misjudgment rate below 2%.

[0046] when Immediately trigger island stabilization control and record the moment the island forms. This provides a time reference for subsequent control strategy switching; the island identification results and the island formation time are transmitted as key information to island stability control.

[0047] In this embodiment of the invention, after confirming the formation of an island in step S400, an islanding stability control strategy is initiated, outputting and issuing active power output commands from the distributed power source to maintain stable islanding operation, including the following steps S401-S402: S401. Once island formation is confirmed, immediately initiate the island stabilization control strategy.

[0048] The core of islanded stability control is to quickly balance power and stabilize voltage frequency by adjusting the output of distributed power sources within the island. It adopts adaptive droop control as the basic control strategy and introduces virtual inertia to improve dynamic response characteristics. Specifically, the basic control strategy includes the frequency-active droop characteristic and the voltage-reactive droop characteristic of adaptive droop control. The frequency-active power droop characteristic of adaptive droop control is as follows: In the formula, For output frequency, Here, P is the frequency reference value, and P is the current active power output. This is a reference value for active power output; This is the active power droop coefficient. Adaptive adjustment based on power imbalance within the island: In the formula, The baseline droop factor is set at 0.0002 Hz / W. This is the adaptive adjustment coefficient, with a value of 0.5. Active power imbalance; When the power imbalance is large, increase the droop factor to improve the response speed of frequency regulation; When power approaches equilibrium, reducing the droop factor improves frequency stability.

[0049] The voltage-reactive droop characteristic is as follows: In the formula, V is the output voltage amplitude. This is the voltage reference value. Where Q is the reactive power droop coefficient, and Q is the current reactive power output. This is a reference value for reactive power output; The reactive power droop coefficient also adopts an adaptive adjustment mechanism, which is dynamically adjusted according to the voltage deviation.

[0050] S402. To improve the dynamic response characteristics at the moment of island formation, the system introduces virtual inertia control.

[0051] Add an inertia compensation term to the frequency control stage: In the formula, For inertia compensation power, For virtual inertia coefficient, For the rate of change of frequency, use the calculated value directly. The virtual inertia coefficient is determined based on the island capacity. For a typical 2MW island, the virtual inertia coefficient is set to 5s.

[0052] The final active power output command for distributed power sources is: In the formula, It integrates the functions of droop control and virtual inertia control for active power output commands; The active power output command is sent to the grid-connected inverters of each distributed power source within the island via the communication network. After receiving the command, the inverter adjusts its output power. The update cycle of the control command is 50 milliseconds, which is consistent with the data acquisition cycle, forming a real-time closed-loop control.

[0053] Continuously monitor the voltage and frequency within the island, and calculate the voltage and frequency deviations: In the formula, The percentage of voltage deviation. This is for frequency deviation; when and At Hz, the islanded operation is determined to be stable, and the stability flag is set. Set to 1; The stability control module continues to run until it detects that the main network has recovered or receives a grid connection preparation command.

[0054] In this embodiment of the invention, during the stable operation of the island in S500, it is detected whether the main grid has restored power supply. After the main grid is restored, it enters the grid connection preparation stage and performs grid connection synchronization. The synchronization completes the triggering of the grid connection switching operation, including the following steps S501-S502: S501. During stable islanded operation, continuously monitor the operating status of the main grid side and detect whether the main grid has restored power supply.

[0055] The voltage amplitude of the main grid is monitored by a voltage sensor upstream of the circuit breaker. When detected in 10 consecutive sampling periods When the main network is determined to have recovered, the main network recovery flag is set. Set to 1.

[0056] After the main grid is restored, the grid connection preparation phase begins. At this time, it is necessary to predict the synchronization conditions between the islanded side and the main grid side, and calculate the voltage amplitude difference, frequency difference, and phase angle difference between the islanded side and the main grid side. In the formula, For voltage amplitude difference, For frequency difference, For phase angle difference, , , These are the island-side voltage, frequency, and phase angle, respectively. , , These are the main grid side voltage, frequency, and phase angle, respectively. The phase angle is obtained in real time through a phase-locked loop algorithm.

[0057] The grid connection synchronization conditions are defined as follows: When the synchronization condition is met, the synchronization flag is set. Set to 1 to allow closing and grid connection.

[0058] However, initially, there is usually a large deviation between the isolated side and the main network side, which does not meet the synchronization conditions. It is necessary to activate the dual-side synchronization control strategy and actively adjust the parameters on the isolated side to synchronize it with the main network.

[0059] S502, When the main network recovers and the isolated network is stable but the synchronization condition is not met, i.e. and and The system initiates a dual-side synchronous control strategy; The goal of synchronization control is to adjust the voltage, frequency, and phase of the islanded side to enable it to quickly track the main grid parameters and meet the grid synchronization conditions.

[0060] Frequency synchronization control uses a proportional-integral controller, and the frequency adjustment command is as follows: In the formula, This is a frequency adjustment command; This is the frequency proportionality coefficient, with a value of 0.2; This is the frequency integral coefficient, with a value of 0.05; It is the frequency difference; Frequency adjustment commands are superimposed on the frequency control circuit to correct the frequency reference value. : In the formula, The corrected frequency reference value is used to replace the original frequency reference value for droop control; By adjusting the frequency, the island frequency is gradually brought closer to the main network frequency.

[0061] Voltage amplitude synchronization control also uses a proportional-integral controller: In the formula, This is a voltage regulation command; This is the voltage proportionality coefficient, with a value of 0.15. This is the voltage integral coefficient, with a value of 0.03; The voltage adjustment command corrects the voltage reference value.

[0062] Phase angle synchronization control is the most critical aspect. A phase angle feedforward compensation strategy is adopted to directly adjust the phase on the islanded side based on the phase angle difference. In the formula, This is the corrected island-side phase angle; This is the phase angle compensation coefficient, with a value of 0.8; Phase angle adjustment is achieved by modifying the phase of the reference signal of the grid-connected inverter.

[0063] Synchronization control commands are continuously updated every 50 milliseconds to track changes in main network parameters in real time. The synchronization deviation is continuously calculated. Synchronization is considered complete when the grid-connected synchronization conditions are met for 20 consecutive sampling periods. Set to 1 to trigger grid connection switching operation; With dual-sided coordinated control, synchronization can usually be completed within 1-2 seconds, much faster than the 5-10 seconds of traditional methods.

[0064] In this embodiment of the invention, after synchronization is completed in S600, the grid connection closing stage begins, a closing command is issued to perform grid connection switching, performance evaluation and offline optimization of islanded operation and grid connection switching are performed, and the optimal control parameters are obtained and updated to the control system, including the following steps S601-S602: S601, when At that time, the system enters the grid connection and shutdown phase; To further reduce closing impact, a precise closing timing selection strategy is adopted, and the instantaneous voltage difference between the islanded side and the main grid side is calculated: In the formula, For instantaneous voltage difference, and These are the instantaneous voltages on the islanded side and the main grid side, respectively; Monitor the zero-crossing point of the instantaneous voltage difference, when Approaching zero and When the voltage on both sides is equal and changes in the same direction, the closing impact is minimized. A closing command is issued to the grid-connected circuit breaker at the optimal closing time. The circuit breaker's operating time is approximately 30-50 milliseconds. Considering the circuit breaker's operating delay, the closing time is advanced by one circuit breaker operating time. Issue closing command: In the formula, The time when the closing command is issued. For the predicted voltage zero-crossing time, The inherent operating time of the circuit breaker was determined to be 40 milliseconds through offline testing. After the circuit breaker is closed, the system monitors the grid connection point current and calculates the inrush current multiple: In the formula, This is a multiple of the impact current. This represents the maximum current during the first three cycles after the circuit breaker closes. This is the rated current.

[0065] Record the impact current multiple as an evaluation index of control effectiveness; Through precise synchronous control and selection of closing time, the inrush current multiple can usually be controlled below 1.2, which is far lower than the 3-5 times of the traditional method.

[0066] After grid connection is completed, the island identification flag will be activated. Reset to 0, exit islanded operation mode, and return to normal grid-connected operation.

[0067] The control strategy of the distributed power source switches from islanded mode back to grid-connected mode and operates according to the grid dispatch instructions. At this point, a complete closed-loop control process of island formation, stable operation, and smooth grid connection is completed.

[0068] S602. Conduct performance evaluation of the entire islanding operation and grid connection switching process to provide a basis for parameter optimization. Define four key performance indicators, including islanding detection time, islanding operation stability, synchronization time, and grid connection impact. The island detection time is the time interval from the actual formation of the island to its identification and confirmation by the system. In the formula, For the island detection time, To confirm the time of the isolated island for the system, The detection time must be less than 200 milliseconds to represent the actual moment the island forms.

[0069] The stability of islanded operation is evaluated by the time integral of the voltage frequency deviation: In the formula, As a stability indicator, This represents the total operating time of the isolated island. and This is the voltage frequency deviation calculated in step four; the smaller the stability index, the more stable the islanded operation.

[0070] The synchronization time is the time from the start of synchronization control to the fulfillment of grid connection conditions. In the formula, To synchronize time, To meet the synchronization conditions at the time, The synchronization time must be less than 3 seconds to initiate synchronization control.

[0071] In an embodiment of the present invention, performance evaluation and offline optimization of islanded operation and grid-connected switching are performed, including the following steps D1-D3: D1. Based on these performance indicators, a genetic algorithm is used to optimize key control parameters offline, including droop coefficient, virtual inertia coefficient, and synchronous controller parameters; the optimization objective function is: In the formula, For comprehensive performance indicators; The weighting coefficients are 0.2, 0.3, 0.2, and 0.3, respectively, indicating the importance attached to stability and shock suppression.

[0072] D2. By minimizing the comprehensive performance index, the optimal control parameters are obtained and updated into the actual control system; D3. Parameter optimization is performed quarterly to ensure that the control strategy adapts to changes in the operating characteristics of the distribution network.

[0073] In an optional implementation, offline optimization in S602 can be performed by manually tuning parameters periodically based on expert experience rules. All isolated events occurring within the current period are automatically archived monthly or quarterly, and the four key performance indicators defined in this invention are calculated to generate a performance report. A parameter adjustment rule base based on expert experience is established. Operations experts review the performance report, compare historical data trends, and, based on the rule base and their personal experience, determine whether optimization is needed and the direction of optimization. After determining a new set of candidate parameters, the experts verify them on a simulation platform or in an offline testing environment. Once verification is successful, the new parameters are manually updated to the control parameter library of the production system, replacing the original parameters. However, this implementation relies heavily on expert experience, resulting in a subjective and inefficient adjustment process. Furthermore, it struggles to handle complex optimization problems involving multiple coupled parameters, leading to unstable optimization results.

[0074] Example 3 is an embodiment of the present invention, which provides an intelligent identification and smooth switching control method for islanded operation mode of distribution network. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through experiments.

[0075] The experiment of this invention is based on the simulation verification of an actual 10kV distribution network in a certain region. The distribution network includes a 12-kilometer main line and two branch lines, connected to three distributed photovoltaic power sources with a total capacity of 2.5MW, as well as several conventional loads with a total load capacity of approximately 2.0MW.

[0076] A detailed simulation model of the power distribution network was built on the MATLAB / Simulink platform, including line models, load models, distributed power generation and its control system models, circuit breaker models, etc.

[0077] The simulation time step was set to 50 microseconds to ensure accurate simulation of electromagnetic transient processes.

[0078] The simulation scenarios include: islanding caused by main grid failure, load fluctuations during islanding operation, photovoltaic output fluctuations, and grid-connected switching after main grid recovery. The performance of traditional passive islanding detection methods, traditional droop control methods, and the method of this invention is compared to verify the superiority of this method.

[0079] Validation was performed, as shown in Tables 1, 2, and 3. Table 1 Comparison of Island Detection Performance

[0080] In 100 islanding scenario tests, the average detection time of the method of this invention was only 120 milliseconds, which is much faster than the traditional method. The detection accuracy reached 98.2%, with only 2 false positives and 1 false negative. The false positives and false negatives mainly occurred under extreme load change conditions. Traditional passive detection has obvious blind spots under power balance conditions, with a false negative rate as high as 15%, and the detection time is as long as 1-2 seconds. Although traditional active detection has high accuracy, it requires the injection of disturbance signals, which leads to an increase in total harmonic distortion rate of 2-3%, affecting power quality, and the detection time is still about 1 second. The method of this invention achieves fast and accurate islanding detection through multi-dimensional feature fusion and intelligent recognition, without affecting power quality.

[0081] Table 2 Comparison of stability during islanded operation

[0082] In the islanded operation stability test, the method of this invention controls the maximum voltage deviation within 2.7% and the maximum frequency deviation within 0.15Hz, which is far superior to the traditional method. The traditional fixed droop control has a slow response, low accuracy, large voltage and frequency fluctuations, and a stabilization time of up to 4-5 seconds. The method of this invention, through adaptive droop coefficient and virtual inertia control, responds quickly to power imbalance, stabilizes the voltage and frequency within 1 second, and has a stability index of only 0.04, which is one-tenth of the traditional method, significantly improving the stability and power supply quality of islanded operation.

[0083] Table 3 Comparison of Grid Connection Switching Performance

[0084] In grid connection switching tests, the average synchronization time of the method of this invention is 1.8 seconds, and the voltage difference during grid connection is only 42V, the frequency difference is 0.04Hz, and the phase difference is 2.5 degrees, all of which meet the strict synchronization conditions. The inrush current multiple is only 1.15, which is close to the ideal impact-free grid connection. The traditional simple detection method has poor synchronization accuracy, and the inrush current multiple is as high as 4.2, which causes a large impact on the equipment. The grid connection success rate is only 76%, and in some scenarios, the grid connection fails due to excessive impact triggering protection actions. The method of this invention achieves true smooth grid connection through dual-sided coordinated synchronous control and precise closing time selection, with a grid connection success rate of 97%. The 3% of failures are mainly due to communication delay causing synchronous control failure.

[0085] Example 4 is an embodiment of the present invention, illustrating the schematic scheme of the intelligent identification and smooth switching control method for islanded operation modes in a distribution network. It should be noted that the technical solution of the intelligent identification and smooth switching control system for islanded operation modes in a distribution network belongs to the same concept as the technical solution of the aforementioned intelligent identification and smooth switching control method for islanded operation modes in a distribution network. Details not described in detail in the technical solution of the intelligent identification and smooth switching control system for islanded operation modes in this embodiment can be found in the description of the technical solution of the aforementioned intelligent identification and smooth switching control method for islanded operation modes in a distribution network.

[0086] This embodiment provides an intelligent identification and smooth switching control system for islanded operation mode of distribution network, including: a data acquisition and preprocessing module, a feature extraction module, an islanded intelligent identification module, an islanded stability control module, a synchronization control module, and a grid connection switching module; The data acquisition and preprocessing module collects real-time operating data of key nodes in the power distribution network and performs multiple quality checks on the collected data. The feature extraction module smooths the verified data, filters it, extracts feature parameters, and performs normalization processing. The island intelligent identification module establishes an island identification classifier, determines the island identification result, and records the time when the island is formed. The island stability control module, upon confirming the formation of an island, initiates the island stability control strategy, outputs and issues active power output commands from the distributed power source, and maintains the stable operation of the island. During the stable operation of the island, the synchronization control module detects whether the main grid has restored power supply. After the main grid is restored, it enters the grid connection preparation stage and performs grid connection synchronization. The synchronization is completed and the grid connection switching operation is triggered. After synchronization is completed, the grid-connected switching module enters the grid-connected closing stage, issues a closing command to perform grid-connected switching, performs performance evaluation and offline optimization of islanded operation and grid-connected switching, and obtains the optimal control parameters to update the control system.

[0087] This embodiment also provides an electronic device applicable to the intelligent identification and smooth switching control method for islanded operation mode in power distribution networks, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the intelligent identification and smooth switching control method for islanded operation mode in power distribution networks as proposed in the above embodiment.

[0088] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, it implements the intelligent identification and smooth switching control method for islanded operation mode of distribution network proposed in the above embodiment.

[0089] The storage medium proposed in this embodiment and the method for intelligent identification and smooth switching control of islanded operation mode in distribution networks proposed in the above embodiments belong to the same inventive concept. 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.

[0090] 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.

[0091] 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 method for intelligent identification and smooth switching control of islanded operation mode in distribution networks, characterized in that: include, Collect real-time operational data of key nodes in the power distribution network and perform multiple quality checks on the collected data; The validated data is smoothed, filtered, and then the feature parameters are extracted and normalized. Establish an island identification classifier, determine the island identification results, and record the time when an island is formed; Once island formation is confirmed, the island stability control strategy is activated, outputting and issuing active power output commands from the distributed power source to maintain stable island operation. During the stable operation of the isolated system, it checks whether the main grid has restored power supply. After the main grid is restored, it enters the grid connection preparation stage and performs grid connection synchronization. The grid connection switching operation is triggered synchronously. After synchronization is completed, the grid connection and switching phase begins. A switching command is issued to perform grid connection switching. Performance evaluation and offline optimization are performed on islanded operation and grid connection switching to obtain the optimal control parameters and update them to the control system.

2. The intelligent identification and smooth switching control method for islanded operation mode of distribution network as described in claim 1, characterized in that: The multiple quality verification includes collecting real-time operating data of key nodes in the distribution network, including electrical quantity information and switching quantity information, and setting the data collection frequency; The collected data undergoes multiple quality checks, including data integrity check, data rationality check, and data consistency check. Data that passes all three checks is stored in the data cache.

3. The intelligent identification and smooth switching control method for islanded operation mode of distribution network as described in claim 2, characterized in that: The extraction of feature parameters includes smoothing the data using a sliding window filtering method; and extracting feature parameters reflecting the islanding state after filtering. The feature vector is constructed based on the feature parameters, and then the feature vector is normalized.

4. The intelligent identification and smooth switching control method for islanded operation mode of distribution network as described in claim 3, characterized in that: The process of establishing an island identification classifier includes using the extracted feature vectors and employing a support vector machine algorithm to establish the island identification classifier. By training samples from different scenarios using historical operational data, the optimal parameters of the support vector machine can be determined. Determine the island identification result, identify the island identification flag, immediately trigger island stabilization control, and record the island formation time.

5. The intelligent identification and smooth switching control method for islanded operation mode of distribution network as described in claim 4, characterized in that: The islanding stability control strategy includes adopting adaptive droop control as the basic control strategy and introducing virtual inertia to improve dynamic response characteristics; the basic control strategy includes the frequency-active droop characteristics and voltage-reactive droop characteristics of adaptive droop control. Virtual inertia control is introduced to improve the dynamic response characteristics during island formation. An inertia compensation term is added to the frequency control stage. In the formula, For inertia compensation power, For virtual inertia coefficient, The rate of change of frequency; The final active power output command for distributed power sources is: In the formula, This is an instruction to contribute effort. This is a frequency reference value. The droop coefficient is adaptively adjusted. The output frequency is set; the command is sent through the communication network to the grid-connected inverters of each distributed power source in the island, and the inverters adjust their output power after receiving the command. Continuously monitor the voltage and frequency within the island, and calculate the voltage and frequency deviations: In the formula, The percentage of voltage deviation. This is the voltage reference value. For frequency deviation; after determining that the islanded operation is stable, the stability control continues to run until the main network is detected to be restored or a grid connection preparation command is received.

6. The intelligent identification and smooth switching control method for islanded operation mode of distribution network as described in claim 5, characterized in that: The grid connection preparation phase includes continuously monitoring the operating status of the main grid during stable islanding operation to detect whether the main grid has resumed power supply; and monitoring the main grid voltage amplitude through a voltage sensor upstream of the circuit breaker. When detected in 10 consecutive sampling periods When the main network is determined to have recovered, the main network recovery flag is set. Set to 1, where, This is the rated voltage of the power grid; After the main grid is restored, the grid connection preparation phase begins. It is necessary to predict the synchronization conditions between the islanded side and the main grid side, and calculate the voltage amplitude difference between the islanded side and the main grid side. Frequency difference and phase angle difference ; The grid connection synchronization conditions are defined as follows: When the synchronization condition is met, the synchronization flag is set to... Set to 1 to allow grid connection; when the main grid recovers and the islanding is stable but the synchronization conditions are not met, the dual-side synchronization control strategy is activated. The goal of the synchronization control is to track the main grid parameters by adjusting the voltage, frequency and phase on the islanding side to meet the grid connection synchronization conditions. Frequency synchronization control uses a proportional-integral controller, and the frequency adjustment command is as follows: In the formula, This is a frequency adjustment command. This is the frequency scaling factor. This is the frequency integral coefficient; the frequency adjustment command is superimposed on the frequency control loop to correct the frequency reference value, and through frequency adjustment, the islanded frequency becomes the main network frequency; Voltage amplitude synchronization control also uses a proportional-integral controller: In the formula, This is a voltage regulation command. This is the voltage proportionality coefficient. This is the voltage integral coefficient, used by the voltage regulation command to correct the voltage reference value. A phase angle feedforward compensation strategy is adopted to directly adjust the phase on the islanded side based on the phase angle difference: In the formula, This is the corrected island-side phase angle. The phase angle compensation coefficient is used to adjust the phase angle by modifying the phase of the reference signal of the grid-connected inverter. The synchronization control commands are continuously updated to track changes in main network parameters in real time, calculate synchronization deviations, and determine whether synchronization is complete.

7. The intelligent identification and smooth switching control method for islanded operation mode of distribution network as described in claim 6, characterized in that: The grid connection and shutdown phase includes, when When entering the grid connection and closing phase, a closing timing selection strategy is adopted to reduce the closing impact. The instantaneous voltage difference between the islanded side and the main grid side is calculated: In the formula, For instantaneous voltage difference, and These are the instantaneous voltages on the islanded side and the main grid side, respectively; the zero-crossing point of the instantaneous voltage difference is monitored to determine the optimal closing time; A closing command is issued to the grid-connected circuit breaker at the optimal closing time, taking into account the circuit breaker's operating delay, and is initiated one circuit breaker operating time in advance. Issue closing command: In the formula, The time when the closing command is issued. For the predicted voltage zero-crossing time, This refers to the inherent operating time of the circuit breaker. After the circuit breaker is closed, monitor the grid connection point current and calculate the inrush current multiple as an evaluation indicator of control effectiveness; after grid connection is completed, the islanding identification flag is set. Reset to 0, the system exits islanded operation mode and returns to normal grid-connected operation; A performance evaluation is conducted on the entire islanding operation and grid connection switching process, and key performance indicators are defined, including islanding detection time, islanding operation stability, synchronization time, and grid connection impact. Based on key performance indicators, a genetic algorithm is used to optimize key control parameters offline, including droop coefficient, virtual inertia coefficient, and synchronous controller parameters; the optimization objective function is: In the formula, For comprehensive performance indicators, For the weighting coefficients assigned to different parameters, For the island detection time, As a stability indicator, To synchronize time, The factor is the multiple of the inrush current; by minimizing the comprehensive performance index, the optimal control parameters are obtained and updated to the actual control system.

8. A distribution network islanding operation mode intelligent identification and smooth switching control system, using the distribution network islanding operation mode intelligent identification and smooth switching control method as described in any one of claims 1 to 7, characterized in that, include: Data acquisition and preprocessing module, feature extraction module, island intelligent identification module, island stability control module, synchronization control module, grid connection switching module; The data acquisition and preprocessing module collects real-time operating data of key nodes in the power distribution network and performs multiple quality checks on the collected data. The feature extraction module smooths the verified data, filters it, extracts feature parameters, and performs normalization processing. The island intelligent identification module establishes an island identification classifier, determines the island identification result, and records the time when the island is formed. The island stability control module, upon confirming the formation of an island, initiates the island stability control strategy, outputs and issues active power output commands from the distributed power source, and maintains the stable operation of the island. During the stable operation of the island, the synchronization control module detects whether the main grid has restored power supply. After the main grid is restored, it enters the grid connection preparation stage and performs grid connection synchronization. The synchronization is completed and the grid connection switching operation is triggered. After synchronization is completed, the grid-connected switching module enters the grid-connected closing stage, issues a closing command to perform grid-connected switching, performs performance evaluation and offline optimization of islanded operation and grid-connected switching, and obtains the optimal control parameters to update the control system.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the intelligent identification and smooth switching control method for islanded operation mode of distribution network as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the intelligent identification and smooth switching control method for islanded operation mode of distribution network as described in any one of claims 1 to 7.

Images