Distributed collaborative control method, system, device and medium for power distribution district

Abstract

The application relates to the technical field of power distribution area inverter collaborative control, and discloses a distributed collaborative control method, system, equipment and medium for a power distribution area, real-time risk values of a plurality of risk indexes corresponding to multi-dimensional operation data of the power distribution area are calculated, a risk adjustment mechanism based on operation scene characteristics is introduced to correct the real-time risk values, and a comprehensive risk index and a dominant risk type of the power distribution area are obtained; when the comprehensive risk index exceeds a preset high-risk threshold, each inverter node in the power distribution area is controlled to exchange state information, so that a master node and a plurality of slave nodes are determined through a distributed election algorithm; subsequently, the master node is controlled to calculate a total adjustment demand according to the dominant risk type, and the optimal resource allocation scheme is solved and issued to the corresponding slave nodes in combination with the maximum safe adjustment amount and the expected contribution degree received from each slave node, so that each slave node is controlled to execute; rapid response to real-time risk is realized, and the distributed collaborative control efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of distributed transformer substation inverter collaborative control technology, and in particular to a distributed collaborative control method, system, device and medium for distributed transformer substations. Background Technology

[0002] With the high proportion of distributed photovoltaic, energy storage, and flexible loads connected to distribution substations, the system's operating status is becoming increasingly complex, increasing real-time operational risks such as voltage exceeding limits and transformer overload. Traditional solutions mainly fall into two categories: one is local autonomous control based on fixed thresholds, which lacks global coordination and may introduce new global risks through local optimization; the other is centralized control relying on the distribution automation master station, which involves multiple communication layers and long latency, making it difficult to meet the millisecond-level response requirements for rapid risk suppression.

[0003] In recent years, distributed cooperative control has provided a new approach to solving the above problems. Existing technologies, such as distributed optimization based on consensus algorithms or cooperative strategies based on neighbor communication, mostly focus on steady-state objectives such as power sharing or voltage averaging, and cannot perceive risks. Existing methods are usually based on fixed physical or communication topologies for networking, which lacks flexibility and leads to low cooperative efficiency.

[0004] Therefore, how to solve the problems of existing distributed collaborative control schemes being unable to detect risks and having low collaborative efficiency has become a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] This invention provides a distributed collaborative control method, system, device, and medium for distribution radio areas, which solves the problems of existing distributed collaborative control schemes being unable to detect risks and having low collaborative efficiency.

[0006] To address the aforementioned technical problems, the first aspect of this invention provides a distributed cooperative control method for a distribution radio area, comprising:

[0007] Multidimensional operational data of the distribution substation is collected in real time, and the current observation value and short-term forecast value are integrated by weighting method to obtain the real-time risk value of multiple risk indicators corresponding to the multidimensional operational data;

[0008] A risk adjustment mechanism based on operational scenario characteristics is introduced to correct each of the real-time risk values, thereby obtaining the comprehensive risk index and dominant risk type of the distribution substation; the operational scenario characteristics are used to characterize the operational status and dominant energy flow mode of the distribution substation.

[0009] When the comprehensive risk index exceeds the preset high-risk threshold, the inverter nodes in the distribution area are controlled to exchange status information, and based on the evaluation system including controllable capacity, communication quality and topology centrality, a distributed election algorithm is used to determine the master node and several slave nodes.

[0010] The master node calculates the total adjustment requirement based on the dominant risk type, and combines the maximum safe adjustment amount and expected contribution received from each slave node to solve for the optimal resource allocation scheme and distribute it to the corresponding slave nodes to control the execution of each slave node.

[0011] A second aspect of the present invention provides a distributed cooperative control system for a distribution radio area, comprising:

[0012] The risk calculation module is used to collect multi-dimensional operation data of the distribution substation in real time, and to fuse the current observation value and short-term forecast value through a weighting method to obtain the real-time risk value of multiple risk indicators corresponding to the multi-dimensional operation data.

[0013] The risk correction module is used to introduce a risk adjustment mechanism based on the characteristics of the operating scenario to correct each of the real-time risk values, thereby obtaining the comprehensive risk index and dominant risk type of the distribution substation; the operating scenario characteristics are used to characterize the operating status and dominant energy flow mode of the distribution substation.

[0014] The node election module is used to control each inverter node in the distribution area to exchange status information when the comprehensive risk index exceeds a preset high-risk threshold, and to determine the master node and several slave nodes through a distributed election algorithm based on an evaluation system that includes controllable capacity, communication quality and topology centrality.

[0015] The collaborative control module is used to control the master node to calculate the total adjustment demand based on the dominant risk type, combine the maximum safe adjustment amount and expected contribution received from each of the slave nodes, solve for the optimal resource allocation scheme, and send it to the corresponding slave nodes to control each of the slave nodes to execute.

[0016] A third aspect of the present invention provides an electronic device including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the distributed cooperative control method for a distribution radio station as described above.

[0017] A fourth aspect of the present invention provides a computer-readable storage medium comprising a stored computer program, wherein when the device containing the computer-readable storage medium executes the computer program, it implements the distributed cooperative control method for a distribution radio station as described above.

[0018] Compared with the prior art, the beneficial effects of the embodiments of the present invention are as follows:

[0019] By employing a weighted approach that integrates real-time observation and short-term forecast data, the subjectivity or bias of single weighting methods is overcome, resulting in a more scientific allocation of risk indicator weights and a risk value calculation that better reflects the actual system state and future trends. An adjustment mechanism based on operational scenario characteristics is introduced, enabling risk assessment to adapt to the current operational mode of the distribution area, significantly improving the accuracy of capturing the true risk level under complex and changing operating environments. When high risk is detected, master and slave nodes are dynamically generated through distributed election based on real-time status (controllable capacity, communication quality, topology centrality), ensuring the optimality of the leader node and improving the generation and transmission of control commands. It achieves both efficiency and reliability, while avoiding the risk of single-point failures and enhancing the robustness of the system. When calculating resource allocation, it not only considers the total regulation demand and the maximum safe regulation amount of slave nodes (hard constraints), but also the expected contribution of each node. Thus, it obtains the globally optimal allocation scheme under the premise of ensuring the operational safety of each node, realizing a balance between overall risk mitigation at the distribution area level and individual operational safety at the equipment level. This improves the economy and safety of regulation actions, effectively enhances the intelligent, adaptive, and collaborative control capabilities of distribution areas in dealing with various operational risks under high-proportion distributed energy access, and optimizes the overall operating status while ensuring equipment safety. Attached Figure Description

[0020] To more clearly illustrate the technical solution of the present invention, the drawings used in 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.

[0021] Figure 1 This is a flowchart of a distributed cooperative control method for a distribution radio station area provided in a certain embodiment of the present invention;

[0022] Figure 2 This is a curve showing the change of comprehensive risk indicators of a distribution radio station area according to a certain embodiment of the present invention;

[0023] Figure 3 This is a diagram showing the optimized reactive power resource allocation results provided in a certain embodiment of the present invention;

[0024] Figure 4 This is a structural diagram of a distributed collaborative control system for a distribution radio station area provided in a certain embodiment of the present invention;

[0025] Figure 5 This is a structural diagram of an electronic device provided in a certain embodiment of the present invention;

[0026] Figure label:

[0027] Among them, 10 is the risk calculation module; 20 is the risk correction module; 30 is the node election module; 40 is the collaborative control module; 5000 is the electronic device; 5001 is the processor; 5002 is the bus; 5003 is the memory; and 5004 is the transceiver. Detailed Implementation

[0028] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings and examples. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0029] In the description of this application, the terms "first," "second," "third," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first," "second," "third," etc., may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items. Those skilled in the art will be able to understand the specific meaning of the above terms in this application according to the specific circumstances.

[0030] In the description of this application, it should be noted that, unless otherwise specified, all technical and scientific terms used in this invention have the same meaning as commonly understood by those skilled in the art. The terminology used in this specification is only for describing specific embodiments and is not intended to limit the invention. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0031] In one embodiment, such as Figure 1 As shown, the first aspect of the present invention provides a distributed cooperative control method for a distribution radio area, comprising:

[0032] S1. Real-time acquisition of multi-dimensional operation data of the distribution substation area, and fusion of current observation values ​​and short-term forecast values ​​through a weighted method to obtain real-time risk values ​​of multiple risk indicators corresponding to the multi-dimensional operation data; wherein, the multi-dimensional operation data includes operation time, operation power and operation load;

[0033] In one embodiment, step S1 includes:

[0034] Multidimensional operational data of the distribution radio area is collected in real time, multiple risk indicators are determined based on the multidimensional operational data, and the current observed value of each risk indicator is determined based on the multidimensional operational data.

[0035] The multidimensional operational data is processed by a time series forecasting algorithm to obtain short-term forecast values ​​for each risk indicator. The current observation values ​​and the short-term forecast values ​​are then weighted and fused to obtain a comprehensive evaluation value for each risk indicator.

[0036] The dynamic weights of each risk indicator are determined according to the weighting method, and the real-time risk values ​​of each risk indicator are determined based on the dynamic weights and the comprehensive evaluation values.

[0037] Specifically, each inverter node collects its local electrical quantities (such as node voltage, output active / reactive power data, operating current, operating time, rated voltage, operating frequency, and its own operating load) in real time through sensors or data collection devices. It also acquires common quantities of the distribution substation area (such as transformer load factor, total output, total load, apparent power, and rated power) via a shared bus or broadcast method. This collected data serves as the multi-dimensional operating data for the distribution substation area. Subsequently, multiple risk indicators are determined based on the collected data, including but not limited to: transformer load factor, node voltage deviation, and three-phase imbalance. The three-phase unbalance is calculated using real-time multidimensional operating data, frequency deviation, and source load power fluctuation rate. Transformer load rate is the ratio of the transformer's current apparent power to its rated power or rated capacity. Node voltage deviation is the difference between the current actual voltage and its rated voltage, or the ratio of voltage deviation to the allowable voltage deviation threshold. Frequency deviation is the difference between the current actual frequency and the rated frequency of the power grid. Three-phase unbalance is the difference between the maximum and minimum phase currents divided by the average three-phase current. Source load power fluctuation rate is quantified using short-term power fluctuation rate, calculated as the absolute value of the difference between the current power and the power one minute ago, divided by the power one minute ago. Based on these calculation principles, the values ​​of each risk indicator calculated using real-time and historical multidimensional operating data are their current observed values.

[0038] A time series forecasting algorithm (such as ARIMA, LSTM, or a lighter-weight moving average autoregressive model) is used to predict the voltage value sequence for the next 15-30 minutes (3-6 sampling periods) for each risk indicator's underlying data source (such as voltage or power). Specifically, using voltage time series data from the previous two hours or real-time voltage data, the voltage value sequence for the next 15 minutes is predicted. Then, based on the predicted voltage sequence, the node voltage deviation for each predicted point is calculated using the same formula logic as the observed values. The maximum value is taken as the final short-term predicted value of the node voltage deviation. Similarly, the short-term predicted values ​​for other risk indicators are calculated. The real-time observed values ​​and short-term predicted values ​​of each risk indicator are weighted according to a preset ratio (preferably 7:3) to obtain the comprehensive evaluation value of the corresponding risk indicator.

[0039] In one embodiment, determining the dynamic weights of each risk indicator according to the weighting method includes:

[0040] A judgment matrix is ​​constructed using the analytic hierarchy process (AHP), and the subjective weights of each risk indicator are determined based on the judgment matrix.

[0041] The information entropy of each risk indicator is determined based on the multidimensional operational data, and an objective weight reflecting the volatility characteristics of each risk indicator is determined by an improved entropy weight method based on the information entropy.

[0042] A preference coefficient is introduced to weight the subjective weights and objective weights to obtain the dynamic weights of the risk indicators.

[0043] Several experts in the field of power distribution networks (including operation and scheduling experts, equipment management experts, and new energy access experts) were invited to conduct pairwise importance comparisons of the risk indicators of this invention based on industry standards and the load characteristics of the industrial park. Saaty's 1-9 scale method was used for scoring to form a judgment matrix. The eigenvectors of each expert's judgment matrix were calculated to obtain preliminary weights. The preliminary weights of each risk indicator were then checked for consistency. If the check passed, the eigenvectors were normalized to obtain the subjective weights of each risk indicator. Otherwise, the experts needed to readjust their judgments.

[0044] Collect operational data for various risk indicators over a recent period (e.g., the past hour). Each indicator has multiple observations at various time points (e.g., one point per minute, for a total of 60 samples). Normalize the data for each indicator. Since risk indicators may have different dimensions and directions (some are more dangerous the larger, others the smaller), they need to be uniformly treated so that larger values ​​indicate greater danger. For example, for voltage deviation, take the absolute value and then normalize; for frequency deviation, do the same; for power imbalance, which is a percentage, normalize directly; and for line load rate, which is also a percentage, normalize according to the actual situation of the indicator. The information entropy of each indicator is calculated based on the normalized data. The traditional entropy weighting method uses the redundancy of information entropy (the difference between 1 and information entropy Ej) to calculate the weight. However, sometimes the information entropy is close to 1, which can lead to a very small weight. The improved entropy weighting method adopted in this invention can introduce an adjustment factor γ, γ=exp(-α×Ej); α is an adjustment parameter (which can be 1-3); the difference coefficient is the product of the redundancy of the original information entropy and the adjustment factor. Then, the difference coefficient is normalized to obtain the objective weight of the risk indicator.

[0045] A preference coefficient is introduced and used as the weight of the subjective weight of the risk indicator, while the objective weight of the risk indicator is the difference between 1 and the preference coefficient. By linearly weighting the subjective and objective weights of the risk indicators, the dynamic weight of each risk indicator can be obtained. This invention can also dynamically adjust the preference coefficient according to the operating status and management strategy of the distribution area. For example, during normal operation (stable load): 0.5 (balance between subjective and objective factors); during special operating conditions (such as equipment maintenance): 0.7 (emphasis on expert experience); during periods of high penetration of renewable energy: 0.3 (emphasis on data objectivity); during holidays or low load at night: 0.6 (appropriately conservative, emphasis on experience), or selected within the range [0.3, 0.6] according to actual needs. Finally, the comprehensive evaluation value of each risk indicator is multiplied by its corresponding dynamic weight, and then normalized to output the real-time risk value of the distribution area.

[0046] This invention, by simultaneously incorporating current observations and short-term forecasts, not only assesses the static risk of a transformer area at this moment but also provides insights into its recent dynamic trends, enhancing the comprehensiveness and foresight of risk perception. Traditional fixed-weight or single-weighting methods are difficult to adapt to the ever-changing operating states of a system. This invention combines the Analytic Hierarchy Process (AHP) with an improved entropy weighting method, effectively integrating the experience and knowledge of domain experts and the statistical characteristics of data, making the weight allocation of risk indicators more scientific. By flexibly adjusting the fusion ratio of subjective and objective weights through preference coefficients, the comprehensive weights can adapt to changes in the current operating state of the transformer area, providing accurate and structured decision-making basis for subsequent coordinated control.

[0047] S2. A risk adjustment mechanism based on operational scenario characteristics is introduced to correct each of the real-time risk values, thereby obtaining the comprehensive risk index and dominant risk type of the distribution substation; the operational scenario characteristics are used to characterize the operational status and dominant energy flow mode of the distribution substation.

[0048] In one embodiment, the introduction of a risk adjustment mechanism based on operational scenario characteristics to correct each of the real-time risk values ​​to obtain a comprehensive risk index for the distribution transformer area includes:

[0049] The operating scenario characteristics are determined based on the operating time, operating power, and operating load, and the risk indicators to be adjusted are determined based on the operating scenario characteristics.

[0050] Based on the running time, the running power, the running load, and the current observation, the time risk quantification value, the power characteristic quantification value, and the index approximation degree of the risk index to be adjusted are determined respectively.

[0051] The dynamic weights of the risk indicators to be adjusted are corrected by using the time risk quantification value, the power characteristic quantification value, and the indicator approximation degree to obtain the corrected weights;

[0052] The real-time risk value of the risk indicator to be adjusted is corrected using the corrected weight, and the corrected risk value is combined with other real-time risk values ​​other than the risk indicator to be adjusted to obtain the comprehensive risk index of the distribution area.

[0053] Specifically, this invention determines the characteristics of the operating scenario based on the operating time, operating power, and operating load in the multi-dimensional operating data. That is, if the operating time is between 10:00 and 14:00 (the natural peak period of photovoltaic output) and the transformer reverse power is greater than 0 (the photovoltaic output is greater than the local load demand, and the power is transmitted from the distribution area to the upper-level grid in reverse), it indicates that the distribution area is in an operating state dominated by the power source (photovoltaic) output and the energy flows in reverse. At this time, the core physical contradiction faced by the grid is the voltage rise caused by the photovoltaic surplus, and the overvoltage risk is the main threat. It is determined to be a photovoltaic reverse transmission scenario. At this time, the risk indicator to be adjusted is the node voltage deviation.

[0054] If the operating time is between 18:00 and 22:00 (the natural peak period for residential / industrial and commercial loads) and the transformer load rate is >80% (close to the rated capacity, with high equipment operating pressure), it indicates that the distribution area is in a state of demand-driven operation and positive high load flow of energy. At this time, the core physical contradiction faced by the power grid is transformer overload caused by excessive load, and the risk of equipment overload is the main threat. It is determined to be a characteristic of peak load scenario, and the risk indicator to be adjusted at this time is the transformer load rate.

[0055] If the operating time is between 00:00 and 06:00 (PV power output is zero, residential and industrial loads are at their lowest level throughout the day and are operating stably), the transformer load rate is less than 30% (low load operation), and the positive power fluctuation is less than 5% / hour (no drastic load changes, power curve is flat), then the distribution area is in a low-load, no-PV, and highly stable operating state. The energy flow shows unidirectional grid input and low-flow stable transmission. The problem of uneven three-phase load distribution is relatively amplified, which can easily lead to local overheating of transformers and increased line losses, becoming the dominant risk at present. It is determined to be a low-load scenario at night, and the risk indicator to be adjusted at this time is the three-phase imbalance.

[0056] If the operating time is between 08:00-09:00 (rapid increase in photovoltaic output + start-up of morning peak load) or 16:00-17:00 (rapid decrease in photovoltaic output + preheating of evening peak load), and the transformer power fluctuation rate in the operating power is >20% / 10 minutes (rapid imbalance of source and load power, drastic change in supply and demand), and the load rate in the operating load is between 30%-70% (medium load range, no extreme overload / low load), it indicates that the distribution area is in a dynamic transitional state of sudden change in both source and load and rapid imbalance in supply and demand. The energy flow shows a rapid bidirectional switching and drastic fluctuation in flow. The rapid fluctuation in power supply and demand leads to the grid frequency deviating from the rated value. If the response is not timely, it is easy to cause abnormal equipment operation. It is judged to be a characteristic of sudden change in source and load transition scenario. At this time, the risk indicator to be adjusted is frequency deviation.

[0057] If the operating time is between 10:00 and 16:00 (the photovoltaic output period, and cloudy weather leads to unstable sunlight), and the fluctuation rate of transformer reverse power (photovoltaic output) in the operating power is >15% / 5 minutes (frequent fluctuations in photovoltaic output), and the fluctuation rate of forward load power in the operating power is >10% / 5 minutes (superimposed random load changes), then the distribution area is in a state of dual instability due to frequent photovoltaic fluctuations and random load fluctuations. The energy flow is dominated by photovoltaic backfeeding and continuous flow oscillations. When photovoltaic output fluctuations and load fluctuations are coupled, the superposition of the two fluctuations can easily cause voltage fluctuations and frequency oscillations. The power fluctuation rate directly reflects the system stability and becomes the dominant risk. It is determined to be a characteristic of a continuous weather fluctuation scenario. In this case, the risk indicator to be adjusted is the source-load power fluctuation rate. Other situations are normal fluctuation scenarios of the power grid, with no risk indicator to be adjusted and no need to correct the dynamic weights.

[0058] For each operational scenario, the current duration of the scenario is determined based on real-time operating time. The typical total duration of the scenario within a day is calculated based on historical data. The risk level coefficient for each time period is then determined according to the rate of exceedance accidents of the core indicator for that period (Level 1 = 1.0, Level 2 = 1.3, Level 3 = 1.6; where the risk level of the scenario is divided according to thresholds: exceedance accident rate within the first threshold is Level 1; between the first and second thresholds is Level 2; and above the second threshold is Level 3, with the threshold determined based on historical data or expert experience). The time risk quantification value is set as: Current duration of scenario / Typical total duration of scenario × Time period risk level coefficient; the power characteristic quantification value is set as: Transformer core power parameter / [Values ​​not specified in the original text]. The threshold values ​​are as follows: For photovoltaic backfeed scenarios, the core parameter is the transformer power (threshold is rated power); for peak load scenarios, the core parameter is the load rate (preferably 80%); for low-load nighttime scenarios, the core parameter is the load rate (preferably 30%); for sudden source-load transition scenarios, the core parameter is the power fluctuation rate (preferably 20%); and for continuous fluctuation scenarios, the core parameter is the combined source-load fluctuation rate (preferably 25%). The index approximation degree is set as: current measured value of the core over-limit index / over-limit threshold. The over-limit threshold is set to: frequency offset ±0.2Hz, three-phase imbalance 15%, and source-load power fluctuation rate 20%. These thresholds can also be set based on historical data or human experience.

[0059] The correction coefficient is determined by calculating the time risk quantification value, the power characteristic quantification value, and the index approximation degree. This process is expressed by the following formula:

[0060] K=1.1+0.3×Tr+0.4×Pr+0.2×(Ia 1.2 -0.5)

[0061] In the formula, K is the correction coefficient; Tr is the time risk quantification value; Pr is the power characteristic quantification value; Ia is the index approximation degree, and the exponential term (Ia) 1.2 This allows the weight to be amplified more rapidly as the indicator approaches the threshold, reflecting the adaptive characteristic that the closer the risk, the higher the weight, thus breaking through the limitations of traditional linear correction.

[0062] It should be noted that the calculation process of the above correction coefficient is subject to the constraint that 1.2 ≤ K ≤ 1.9, meaning a minimum amplification of 20% to ensure the priority of core indicators and a maximum amplification of 90% to avoid ignoring other indicators. Multiplying the correction coefficient by the dynamic weight of the risk indicator to be adjusted yields the correction weight; for other risk indicators besides the one to be adjusted, K = 1.0, maintaining the basic weight and not adjusting. Finally, multiplying the newly obtained weight by its corresponding risk indicator yields the comprehensive risk indicator of the distribution substation; simultaneously, the risk indicator with the largest proportion (i.e., the product of the indicator and its weight is the largest) is selected from the comprehensive risk indicators, and this determines the dominant risk type (transformer load rate corresponds to transformer overload risk, node voltage deviation corresponds to overvoltage risk, frequency offset corresponds to power supply and demand imbalance risk, three-phase imbalance corresponds to three-phase imbalance risk, and source-load power fluctuation rate corresponds to voltage and frequency fluctuation risk), providing clear targets for subsequent coordinated control.

[0063] This invention determines the characteristics of the operating scenario through three dimensions: operating time, power, and load. This allows risk assessment to take into account different operating scenarios, thereby enabling differentiated risk assessment under different scenarios, improving the accuracy of the assessment, achieving deep coupling between risk evaluation and operating scenario, and enhancing the spatiotemporal adaptability of the assessment. Based on the scenario characteristics, it determines the risk indicators to be adjusted, achieving precise focus on key risks, so that the final comprehensive risk indicator can more accurately reflect the overall risk level under the current operating scenario.

[0064] S3. When the comprehensive risk index exceeds the preset high-risk threshold, control each inverter node in the distribution area to exchange status information, and determine the master node and several slave nodes through a distributed election algorithm based on an evaluation system that includes controllable capacity, communication quality and topology centrality. Specifically, the present invention presets a risk level threshold, for example, setting a comprehensive risk index > 0.7 as a high-risk threshold. When the comprehensive risk index calculated in step S2 exceeds this threshold, the self-organizing network process is automatically triggered.

[0065] In one embodiment, the control unit exchanges status information among the inverter nodes in the distribution transformer area, and determines the master node and several slave nodes through a distributed election algorithm based on an evaluation system that includes controllable capacity, communication quality, and topology centrality, including:

[0066] Each inverter node is controlled to broadcast its own status information to neighboring nodes via a wireless local area network, so that each inverter node can determine its topology centrality based on the status information of its neighboring nodes and interact with its own neighboring nodes; the status information includes controllable capacity, communication quality, and physical location information.

[0067] Each inverter node is controlled to calculate its own known inverter node communication score based on its own state information and topology centrality, as well as the state information and topology centrality of its neighboring nodes, and then compare it with the neighboring nodes after interacting with them.

[0068] After multiple rounds of information interaction and iteration, the inverter node with the highest communication score is designated as the master node, and the other inverter nodes besides the master node are designated as slave nodes.

[0069] Specifically, this invention treats each inverter as an inverter node, and each inverter node periodically (e.g., every 30 seconds) broadcasts its own status information packet within the local wireless LAN. This packet includes a unique node identifier; communication data: the number of single-hop neighbors, received signal strength, packet loss rate, and communication quality (a comprehensive score calculated by weighting the packet loss rate and latency of recent communications with all neighbors; the weights are determined based on actual needs or historical experience, preferably 4:6, and then normalized to between 0 and 1); physical-electrical data: the node's physical location coordinates (used to assist topology analysis, which can be logical locations such as feeder branch numbers, or the electrical distance between the node and the transformer in the distribution area); and equipment and risk data: equipment type (energy storage / photovoltaic / other), controllable capacity percentage (the maximum adjustable capacity currently available; positive values ​​indicate the ability to generate more capacity, negative values ​​indicate the ability to absorb more capacity), and the current dominant risk type (synchronized via a risk index). When the comprehensive risk index of the distribution area exceeds a threshold, all nodes receive a broadcast command to "start collaborative control," which then synchronously triggers the distributed election process.

[0070] Each inverter node continuously listens to and receives state information packets from its one-hop neighbors. Based on the collected information from neighboring inverters and its own information, it calculates the topology centrality and exchanges the result with its neighboring nodes. The physical meaning of topology centrality is to characterize a node's connectivity and relay reliability as a communication hub. Nodes with higher values ​​score higher in elections and, when acting as master nodes, can effectively reduce packet loss and latency in network-wide broadcasts. The topology centrality is calculated using the following formula:

[0071]

[0072] In the formula, C topo，i Let be the topological centrality of node i; N neighbor,i Let be the number of single-hop neighbors of node i; N total This represents the total number of network nodes. Let be the average signal strength of node i and its neighboring nodes.

[0073] Each node calculates a comprehensive communication score for itself and every known node based on its own state information and topology centrality, as well as the state information and topology centrality received from its neighbors. The contribution score for each node is obtained by weighting the controllable capacity percentage (50%), communication link quality (30%), and topology centrality (20%). Each node selects the node with the highest score from its known set of nodes (including itself and its neighbors) as its current "temporary master node" candidate. The node then packages the temporary master node ID and its score and sends it to all its neighbors. Each node receives the "temporary master node" votes from its neighbors and compares all received candidates (including its own selection), selecting the node with the highest score as the master node.

[0074] After multiple rounds of information exchange and iteration (such as using flooding or consensus algorithms, with the convergence condition being: when any node discovers that in two consecutive iterations, the master node IDs it receives from all its neighbors are exactly the same, and also consistent with its own selection, or the maximum preset number of iterations has been reached), the cluster will eventually converge and recognize the node with the highest score as the master node, with the remaining nodes as slave nodes, thus forming a star network with the master node as the coordination center, or forming a partial mesh network according to communication conditions; this topology is not fixed and can be dynamically reconstructed each time a high-risk event is triggered or the master node fails.

[0075] This invention enables inverter nodes to autonomously compete and negotiate for the optimal leader node based on real-time status information through distributed election. This ensures that the control architecture can dynamically reconfigure itself quickly and autonomously when equipment is put into operation or when a partial failure occurs, maintaining a highly efficient collaborative network led by the currently "most suitable" node. This significantly enhances the system's resilience to uncertainties and disturbances. By comprehensively considering multiple key attributes, the invention ensures the overall optimality of the leader node and its reliable communication hub status (high communication quality and high topology centrality). This ensures the efficient and stable broadcasting and feedback collection of control commands, allowing the elected leader node to theoretically maximize the benefits of collaborative control. Furthermore, it achieves consensus based on distributed consensus, avoiding single-point dependencies and bottlenecks. Global decision-making is achieved through local information interaction, resulting in low communication overhead and good scalability. This significantly reduces network communication overhead in the core election process. The addition or removal of a new node only affects the local information of its neighboring nodes, and the election algorithm can adaptively adjust, giving the system good scalability. This makes it very suitable for plug-and-play applications of distributed energy devices.

[0076] S4. Control the master node to calculate the total adjustment requirement based on the dominant risk type, combine the maximum safe adjustment amount and expected contribution received from each slave node, solve for the optimal resource allocation scheme and send it to the corresponding slave node to control each slave node to execute.

[0077] In one embodiment, the control master node calculates the total adjustment requirement based on the dominant risk type, and, in conjunction with the maximum safe adjustment amount received from each of the slave nodes and the expected contribution, solves for the optimal resource allocation scheme, including:

[0078] The master node calculates the total adjustment requirement based on the dominant risk type and broadcasts a collaboration request frame to all slave nodes; the collaboration request frame includes contribution calculation rules.

[0079] Each slave node determines its maximum safe adjustment amount and corresponding cost data based on its current operating data, and determines its expected contribution based on the contribution calculation rules, and sends it to the master node;

[0080] The master node controls the construction and solution of a resource allocation optimization model based on the maximum safe adjustment amount of each slave node received, its corresponding cost data, and expected contribution, to obtain the optimal resource allocation scheme.

[0081] Specifically, the master node deconstructs the collaborative control objective based on the dominant risk type (e.g., suppressing overvoltage requires absorbing reactive power). The master node's deconstruction of the collaborative control objective is based on calculations using a specific physical model. If the dominant risk is overvoltage, then the target adjustment amount (the total reactive power Q to be absorbed) is... total The calculation formula is:

[0082]

[0083] In the formula, K s The comprehensive voltage-reactive power sensitivity coefficient of the transformer area (preferred value range) ); U real This is the highest voltage currently measured. U limit This is the upper limit threshold for voltage. The safe overshoot coefficient is (preferably 1.05-1.2).

[0084] If the dominant risk is transformer overload, then the target regulation (the total active power P to be reduced) is... total The calculation formula is:

[0085]

[0086] In the formula, S real This represents the current apparent power. S rated Rated power capacity; The target load rate threshold is (preferably 0.95). The current power factor is obtained through monitoring the total power of the transformer area, and its value ranges from 0.85 to 0.95.

[0087] If the dominant risk is an imbalance between power supply and demand, then the target adjustment (total active power P) total (Positive during injection, negative during reduction) The calculation formula is:

[0088] P total =K f ×|Δf|×λ f ×S rated ×cosφ

[0089] In the formula, K f The frequency-active power sensitivity coefficient of the transformer substation represents the active power regulation required per unit frequency offset and per unit transformer capacity. It is obtained through offline simulation or actual measurement of the transformer substation, with a typical value range of 10-30 kW / (Hz·MVA); Δf is the measured value of the current frequency offset; λ f This is the frequency regulation safety overshoot coefficient, with a value range of 1.05-1.2.

[0090] If the dominant risk is the three-phase imbalance risk, then the target adjustment (Q) total Prioritize reactive power compensation to balance the load and avoid active power loss. The compensation amount is positively correlated with the imbalance difference and the transformer's rated capacity (injection is positive, absorption is negative). The calculation formula is as follows:

[0091] Q total =K u ×(ε-ε limit )×λ u ×S rated ×sinφ

[0092] In the formula, K u The three-phase unbalance compensation sensitivity coefficient represents the reactive power compensation required per unit unbalance difference and per unit transformer capacity. It is obtained through actual measurements in the transformer substation, with a typical value range of 8-15 kVar / (ε·MVA); ε is the measured value of the current three-phase current unbalance. limit The allowable threshold for three-phase unbalance is preferably 15%; λ u The unbalanced compensation safety overshoot coefficient has a value range of 1.05-1.2; sinφ is the current system average reactive power factor.

[0093] If the dominant risk is voltage frequency fluctuation risk, then the target adjustment amount (the required reserve power P for rapid response) is... total / Q total The calculation formulas for suppressing frequency fluctuations and voltage fluctuations respectively are as follows:

[0094] P total =K p×[(ΔP / Δt)-(ΔP / Δt) limit ]×λ vf ×S rated ×cosφ

[0095] Q total =K q ×[(ΔP / Δt)-(ΔP / Δt) limit ]×λ vf ×S rated ×sinφ

[0096] In the formula, K p The power fluctuation-active power reserve sensitivity coefficient characterizes the active power reserve required per unit power change rate and per unit capacity. It is obtained through actual measurements in the transformer area, with a typical value of 5-10 kW / ((kW / min)·MVA); K q The power fluctuation-reactive power reserve sensitivity coefficient characterizes the reactive power reserve required per unit power change rate and per unit capacity. It is obtained through actual measurements in the transformer area, with a typical value of 8-12 kW / ((kW / min)·MVA); ΔP / Δt is the current combined power fluctuation rate of the source and load; (ΔP / Δt) limit The allowable threshold for power fluctuation rate is set according to the source-load characteristics of the distribution area; the typical allowable threshold for the distribution network is 20-30 kW / min. vf The value is the overshoot coefficient for fluctuation suppression, ranging from 1.05 to 1.2.

[0097] Subsequently, the master node broadcasts a "cooperation request" frame to all slave nodes. The frame contains: risk type, node identifier, and contribution calculation rules (the expected contribution is the ratio of a node's adjustable capacity coefficient to its adjustment cost coefficient; the adjustable capacity coefficient is the ratio of the maximum safe reactive power absorption currently available to the node to the sum of the maximum safe reactive power absorption reported by all nodes; the adjustment cost coefficient is the ratio of the unit adjustment cost reported by the node to the average unit adjustment cost reported by all nodes. This rule encourages nodes with large adjustment capabilities and low adjustment costs to undertake more tasks, aiming to achieve economic optimization). Upon receiving the request, each slave node initiates its local optimization calculation. Calculation: Based on its current operating status (e.g., a photovoltaic inverter based on its current active power output and terminal voltage), each node calculates its maximum safe regulation capacity and estimates the cost required to complete the regulation (e.g., energy storage loss, photovoltaic curtailment cost, calculated using existing inverter cost calculation methods, which will not be elaborated here) and response time. This data is then reported to the master node. The master node calculates the sum of the maximum safe reactive power absorption required by the rules and the average unit regulation cost reported by all nodes, and distributes this to each slave node. This allows each slave node to calculate its expected contribution to this collaboration based on the rules published by the master node and reply to the master node.

[0098] After collecting all the information, the master node constructs and solves a resource allocation optimization model. The primary objective of this model is to minimize the total system regulation cost, while strong constraints must satisfy the total regulation requirements. The master node uses linear programming or a fast heuristic algorithm to solve the model, obtain the optimal resource allocation scheme, and issue it as a precise power regulation command to each slave node so that each slave node can execute it, thereby realizing the coordinated control of distributed resources in the distribution area.

[0099] This invention, when calculating resource allocation, not only considers the total amount of adjustment required for the entire distribution area, but also strictly respects the hard constraint of the maximum safe adjustment amount of each slave node. This ensures that no control command will force a node to exceed its safe operating range, fundamentally eliminating the risk of secondary equipment failures caused by collaborative control, and achieving a unity of global objectives and local safety. By introducing expected contribution based on cost data and explicit contribution calculation rules, this method elevates resource allocation from a simple "capacity-based allocation" problem to an optimization problem that comprehensively considers the economics of adjustment and the enthusiasm of node participation. Under the premise of meeting the total demand, it minimizes the total adjustment cost of the entire distribution area or maximizes the comprehensive benefits. A centralized-distributed hybrid optimization solution framework is established, balancing efficiency and feasibility, and has high implementability under actual communication and computing conditions.

[0100] In one embodiment, constructing and solving the resource allocation optimization model to obtain the optimal resource allocation scheme includes:

[0101] The resource allocation optimization model is constructed with the goal of minimizing the cost of fulfilling the total adjustment requirements.

[0102] The resource allocation optimization model is solved with the constraints that the sum of the target adjustment amounts of each slave node is not less than the total adjustment requirement and that the target adjustment amount of each slave node is not greater than its own maximum safe adjustment amount. The power adjustment commands of each slave node are then used as the optimal resource allocation scheme.

[0103] Specifically, the master node constructs a resource allocation optimization model based on the information received from the slave nodes, with the goal of minimizing the cost of fulfilling the total adjustment requirements. J The model is expressed by the following formula:

[0104]

[0105] In the formula, X i The adjustment instruction value assigned to node i; C cost,i The adjustment cost coefficient for node i (energy storage cost < photovoltaic cost); The target deviation penalty factor (take a large value, such as 1000, to ensure the overall target is achieved); To meet overall regulatory needs.

[0106] The constraints are that the sum of the target adjustment amounts of each slave node is not less than the total adjustment demand and the target adjustment amount of each slave node is not greater than its own maximum safe adjustment amount. ,in Let i be the maximum safe adjustment amount. This refers to the dynamic safety margin, with a preferred value range of [0.1, 0.2]. It explicitly indicates that the system will forcibly retain 10%-20% of its adjustment capability during each allocation to cope with sudden risk escalation. The resource allocation optimization model is solved using linear programming or a fast heuristic algorithm to obtain the power adjustment commands of each slave node as the optimal resource allocation scheme.

[0107] This invention aims to minimize total cost, ensuring that the resource combination with the lowest economic cost is used while meeting the adjustment requirements, thereby reducing the overall adjustment cost. The constraints ensure that the adjustment amount of each slave node does not exceed its maximum safe adjustment amount, preventing equipment overload or operation in dangerous conditions. It has high solution efficiency and is suitable for rapid solution on edge computing devices or master nodes, meeting real-time requirements.

[0108] Upon receiving the command, each inverter node rapidly executes the corresponding active / reactive power adjustment in its local control loop, ensuring that the entire coordinated action, from risk perception to command execution, takes less than 200 milliseconds. After execution, the system continues to monitor changes in the comprehensive risk index. At the end of a coordinated control cycle (e.g., after 3-5 power frequency cycles), the actual risk reduction is calculated and compared with the expected target of this coordination to evaluate control effectiveness. The entire coordination process data (including each node's expected contribution, actual output, actual communication performance, and final risk reduction) is used as a training sample. A reinforcement learning framework (such as the Actor-Critic algorithm) is used, with the risk reduction value serving as a reward signal. Through policy gradient updates, two optimizations are primarily achieved: 1) Optimizing the contribution prediction model: adjusting the weight parameters in the formula used to calculate the expected contribution locally from the node, making the predicted contribution more reflective of the actual risk suppression effect; 2) Optimizing election weights: dynamically adjusting the weight coefficients in the comprehensive communication score based on the master node's organizational efficiency in this coordination, resulting in a better future elected master node. Through continuous learning, the system can gradually adapt to changes in the transformer area structure and equipment aging, achieving continuous self-improvement.

[0109] In one embodiment, risk-coordinated control verification was implemented for a typical distribution area with a high proportion of distributed photovoltaic (PV) grid connection. This area includes 2.5MW of distributed PV, a 0.6MW / 1.2MWh energy storage system, and various flexible loads. Under typical operating conditions, it faces multiple risks such as midday PV backflow overvoltage, evening peak load overload, and three-phase imbalance. This invention was applied to deploy 286 intelligent inverters (single unit capacity 5-20kW) with end-to-end communication capabilities to construct a risk-driven self-organizing network coordinated control system. The verification focused on the dynamic response capabilities of reactive power cooperative absorption (Q<0) under overvoltage risk and active power cooperative reduction (P<0) under overload risk, as well as the adaptive effect of the networking strategy under topology change scenarios, comprehensively evaluating the system's control performance under multiple risk scenarios.

[0110] By deploying 286 smart inverters and multi-source data acquisition terminals, the current measured values ​​of key indicators such as transformer load rate (threshold 90%), node voltage (threshold 1.07 pu), and three-phase imbalance (threshold 15%) are monitored in real time as static observation values. An analytic hierarchy process (AHP) is used to construct a judgment matrix to determine the subjective weights of each indicator. Simultaneously, objective weights are calculated using the entropy weight method based on recent monitoring data. Finally, a combined weight vector W=[0.35,0.4,0.25] is obtained through linear weighted fusion. An ARIMA time series forecasting model is introduced, using historical 5-minute data as input. After differential stabilization and parameter estimation, the dynamic predicted values ​​of each risk indicator for the next 5 minutes are output. The static observation values ​​and dynamic predicted values ​​are weighted and fused at a 7:3 ratio. After multiplying by the corresponding dynamic weights, the comprehensive risk index Rrisk is calculated through normalization. The change curve of the comprehensive risk index of the distribution substation is shown in the figure. Figure 2 As shown.

[0111] Risk level thresholds are set: Rrisk < 0.4 (normal), 0.4 ≤ Rrisk < 0.7 (warning), and Rrisk ≥ 0.7 (high risk). When Rrisk ≥ 0.7, each inverter broadcasts its own status information (controllable capacity, communication quality, and location information) via the IPv6 over LoRaWAN protocol. A distributed election algorithm based on expected contribution is adopted, using controllable capacity (40% weight), communication quality (30% weight), and topology centrality (30% weight) as evaluation indicators, to elect the energy storage converter with the highest comprehensive score as the temporary master node, forming a star-shaped collaborative network.

[0112] The master node first calculates the total reactive power Q to be absorbed based on the real-time monitored voltage over-limit data (ΔU=0.03pu) and the preset voltage-reactive power sensitivity coefficient K=26.7kVar / pu (per unit value). total=800kVar. Subsequently, the master node broadcasts this adjustment target and contribution calculation rules (including a controllable capacity weight of 0.4, an adjustment cost weight of 0.3, and a response speed weight of 0.3) to all slave nodes via the ad hoc network. Upon receiving the coordination request, each photovoltaic inverter initiates local optimization calculations based on its current operating status: first, it calculates the current active power output P based on the real-time irradiance level and DC voltage value. out Then calculate the remaining reactive power capacity using the following formula:

[0113]

[0114] Then, a local optimization function is constructed with the goal of minimizing the adjustment cost. Q i For the reactive power to be provided, t response To estimate the response time, we solve for its maximum available reactive power Q. avail,i and the corresponding adjustment costs; expected contribution CD i The capacity percentage (Q) is then calculated using a weighted average. avail,i / ΣQ avail ), cost efficiency (1-C) i / ΣC i The results are derived from the communication quality (RSSI); the master node collects the (Q) of all nodes. avail,i CD i C i After obtaining the data, the bisection method is used to quickly solve the optimization problem with the objective of minimizing the total system adjustment cost: MinΣ[C i ·Q i The constraint condition is ΣQ. i ≥800kVar and Q of each node i ≤Q avail,i After obtaining the optimal allocation scheme, the master node retains a 15% safety margin (i.e., the actual instruction Q) for each node. cmd,i =0.85·Q alloc,i The results of reactive power resource allocation optimization are as follows: Figure 3 As shown, the energy storage inverter, due to its fast response speed and low regulation cost, undertakes the basic support task of 300kVar; the remaining 500kVar is shared by the photovoltaic inverters according to their contribution ratio, with the larger 20kW inverters allocated 150kVar and the 10kW inverters allocated 80kVar, forming a hierarchical and coordinated control system. This reserve margin can be activated immediately when the risk index is detected to be rising continuously, ensuring that the system has the ability to respond quickly to risk escalation.

[0115] The perception layer refreshes the risk index every 1 second; the networking layer completes topology reconstruction within 50ms after the risk exceeds the threshold; the negotiation layer completes resource allocation decisions within 100ms; the execution layer controls the inverter to complete reactive power output adjustment within 200ms; and the learning layer evaluates control effectiveness based on the actual risk reduction rate (in this example, from 0.82 to 0.31) and updates the contribution prediction model parameters.

[0116] A real-time simulation test platform based on OPAL-RT was built, including a power grid model of the distribution area, a hardware-in-the-loop device for the inverter cluster, and a communication network simulator. Test results show that the latency from self-networking triggering to the issuance of cooperative commands is less than 200ms; the voltage qualification rate is improved to 99.2% in the overvoltage risk suppression scenario; the transformer load rate is reduced by 12.5% ​​in the overload risk scenario; and the system reconverges to a stable state within 300ms after topology change, verifying the effectiveness and adaptability of the method.

[0117] This application addresses the shortcomings of existing distributed collaborative control schemes, such as their inability to detect risks and low collaborative efficiency. It proposes a distributed collaborative control method for distribution transformer areas, employing a weighted approach that integrates real-time observations and short-term forecasts. This overcomes the subjectivity or bias of single weighted methods, resulting in a more scientific allocation of risk indicators and a more accurate reflection of the system's actual state and future trends. An adjustment mechanism based on operational scenario characteristics is introduced, enabling risk assessment to adapt to the current operational mode of the distribution transformer area, significantly improving the accuracy of capturing the true risk level under complex and variable operating conditions. When high risk is detected, a distributed election is dynamically generated based on real-time status (controllable capacity, communication quality, topology centrality). The master-slave node system ensures the optimality of the leader node, improves the efficiency and reliability of control command generation and transmission, avoids single-point failure risks, and enhances the robustness of the system. When calculating resource allocation, it considers not only the total regulation demand and the maximum safe regulation amount of the slave nodes (hard constraints), but also the expected contribution of each node. The solution obtained is a globally optimal allocation scheme under the premise of ensuring the operational safety of each node. This achieves a balance between overall risk mitigation at the distribution area level and individual operational safety at the equipment level, improves the economy and safety of regulation actions, and effectively enhances the intelligent, adaptive, and collaborative control capabilities of distribution areas to cope with various operational risks under high-proportion distributed energy access. While ensuring equipment safety, it optimizes the overall operating status.

[0118] It should be noted that although the steps in the flowchart above are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order requirement for the execution of these steps, and they can be executed in other orders.

[0119] In another embodiment, such as Figure 4 As shown, a second aspect of the present invention provides a distributed cooperative control system for a distribution radio area, comprising:

[0120] The risk calculation module 10 is used to collect multi-dimensional operation data of the distribution area in real time, and to obtain the real-time risk value of multiple risk indicators corresponding to the multi-dimensional operation data by integrating the current observation value and the short-term forecast value through a weighting method.

[0121] Risk correction module 20 is used to introduce a risk adjustment mechanism based on operating scenario characteristics to correct each of the real-time risk values, thereby obtaining the comprehensive risk index and dominant risk type of the distribution substation; the operating scenario characteristics are used to characterize the operating status and dominant energy flow mode of the distribution substation.

[0122] The node election module 30 is used to control each inverter node in the distribution area to exchange status information when the comprehensive risk index exceeds a preset high-risk threshold, and to determine the master node and several slave nodes through a distributed election algorithm based on an evaluation system that includes controllable capacity, communication quality and topology centrality.

[0123] The collaborative control module 40 is used to control the master node to calculate the total adjustment demand according to the dominant risk type, combine the maximum safe adjustment amount and expected contribution received from each of the slave nodes, solve for the optimal resource allocation scheme, and send it to the corresponding slave nodes to control each of the slave nodes to execute.

[0124] It should be noted that the various modules in the aforementioned distributed cooperative control system for distribution transformer areas can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module. For specific limitations regarding the distributed cooperative control system for distribution transformer areas, please refer to the limitations regarding the distributed cooperative control method for distribution transformer areas described above; both have the same function and role, and will not be repeated here.

[0125] A third aspect of the present invention provides an electronic device comprising:

[0126] Processor, memory, and bus;

[0127] The bus is used to connect the processor and the memory;

[0128] The memory is used to store operation instructions;

[0129] The processor is configured to execute operations corresponding to a distributed cooperative control method for a distribution radio area as shown in the first aspect of this application by invoking the operation instructions.

[0130] In one alternative embodiment, an electronic device is provided, such as Figure 5 As shown, Figure 5 The illustrated electronic device 5000 includes a processor 5001 and a memory 5003. The processor 5001 and the memory 5003 are connected, for example, via a bus 5002. Optionally, the electronic device 5000 may also include a transceiver 5004. It should be noted that in practical applications, the transceiver 5004 is not limited to one type, and the structure of this electronic device 5000 does not constitute a limitation on the embodiments of this application.

[0131] Processor 5001 may be a CPU, a general-purpose processor, a DSP, an ASIC, an FPGA, or other programmable logic device, transistor logic device, hardware component, or any combination thereof. It may implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 5001 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0132] Bus 5002 may include a path for transmitting information between the aforementioned components. Bus 5002 may be a PCI bus or an EISA bus, etc. Bus 5002 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 5 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0133] The memory 5003 may be a ROM or other type of static storage device capable of storing static information and instructions, RAM or other type of dynamic storage device capable of storing information and instructions, or it may be an EEPROM, CD-ROM or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but is not limited thereto.

[0134] The memory 5003 is used to store application code that executes the scheme of this application, and its execution is controlled by the processor 5001. The processor 5001 is used to execute the application code stored in the memory 5003 to implement the content shown in any of the foregoing method embodiments.

[0135] Among them, electronic devices include, but are not limited to: mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), and in-vehicle terminals (such as in-vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers.

[0136] The fourth aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements a distributed cooperative control method for a distribution radio station as shown in the first aspect of the present application.

[0137] Another embodiment of this application provides a computer-readable storage medium storing a computer program that, when run on a computer, enables the computer to execute the corresponding content in the aforementioned method embodiments.

[0138] Furthermore, embodiments of the present invention also provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0139] In summary, this invention relates to the field of distributed transformer substation inverter collaborative control technology, and discloses a distributed collaborative control method, system, device, and medium for distributed transformer substations. It calculates the real-time risk values ​​of multiple risk indicators corresponding to the multi-dimensional operating data of the distributed transformer substation, and introduces a risk adjustment mechanism based on operating scenario characteristics to correct these values, obtaining a comprehensive risk indicator and the dominant risk type of the distributed transformer substation. When the comprehensive risk indicator exceeds a preset high-risk threshold, the inverter nodes in the distributed transformer substation are controlled to exchange status information to determine the master node and several slave nodes through a distributed election algorithm. Subsequently, the master node is controlled to calculate the total adjustment demand based on the dominant risk type, and, combined with the maximum safe adjustment amount and expected contribution received from each slave node, solves for the optimal resource allocation scheme and distributes it to the corresponding slave nodes to control their execution. This achieves rapid response to real-time risks and improves the efficiency of distributed collaborative control.

[0140] The various embodiments in this specification are described in a progressive manner. For directly identical or similar parts of the embodiments, refer to each other. Each embodiment focuses on its differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. It should be noted that the technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification.

[0141] The embodiments described above are merely preferred embodiments of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various improvements and substitutions without departing from the technical principles of this invention, and these improvements and substitutions should also be considered within the scope of protection of this application. Therefore, the scope of protection of this patent application should be determined by the scope of the claims.

Claims

1. A distributed cooperative control method for a distribution radio area, characterized in that, include: Multidimensional operational data of the distribution substation is collected in real time, and the current observation value and short-term forecast value are integrated by weighting method to obtain the real-time risk value of multiple risk indicators corresponding to the multidimensional operational data; A risk adjustment mechanism based on operational scenario characteristics is introduced to correct each of the real-time risk values, thereby obtaining the comprehensive risk index and dominant risk type of the distribution area; The operational scenario features are used to characterize the operational status and dominant energy flow mode of the distribution substation. When the comprehensive risk index exceeds the preset high-risk threshold, the inverter nodes in the distribution area are controlled to exchange status information, and based on the evaluation system including controllable capacity, communication quality and topology centrality, a distributed election algorithm is used to determine the master node and several slave nodes. The master node calculates the total adjustment requirement based on the dominant risk type, and combines the maximum safe adjustment amount and expected contribution received from each slave node to solve for the optimal resource allocation scheme and distribute it to the corresponding slave nodes to control the execution of each slave node.

2. The distributed cooperative control method for a distribution radio area according to claim 1, characterized in that, The real-time acquisition of multi-dimensional operational data from the distribution substation is used to fuse current observations and short-term forecasts using a weighted method to obtain real-time risk values ​​for multiple risk indicators corresponding to the multi-dimensional operational data, including: Multidimensional operational data of the distribution radio area is collected in real time, multiple risk indicators are determined based on the multidimensional operational data, and the current observed value of each risk indicator is determined based on the multidimensional operational data. The multidimensional operational data is processed by a time series forecasting algorithm to obtain short-term predicted values ​​for each risk indicator. The current observed values ​​and the short-term predicted values ​​are then weighted and fused to obtain a comprehensive evaluation value for each risk indicator. The dynamic weights of each risk indicator are determined according to the weighting method, and the real-time risk values ​​of each risk indicator are determined based on the dynamic weights and the comprehensive evaluation values.

3. The distributed cooperative control method for a distribution radio area according to claim 2, characterized in that, The step of determining the dynamic weights of each risk indicator according to the weighting method includes: A judgment matrix is ​​constructed using the analytic hierarchy process (AHP), and the subjective weights of each risk indicator are determined based on the judgment matrix. The information entropy of each risk indicator is determined based on the multidimensional operational data, and an objective weight reflecting the volatility characteristics of each risk indicator is determined by an improved entropy weight method based on the information entropy. A preference coefficient is introduced to weight the subjective weights and objective weights to obtain the dynamic weights of the risk indicators.

4. The distributed cooperative control method for a distribution radio area according to claim 2, characterized in that, The multidimensional operational data includes operating time, operating power, and operating load; among which... The risk adjustment mechanism based on operational scenario characteristics is introduced to correct each of the real-time risk values, resulting in a comprehensive risk index for the distribution area, including: The operating scenario characteristics are determined based on the operating time, operating power, and operating load, and the risk indicators to be adjusted are determined based on the operating scenario characteristics. Based on the running time, the running power, the running load, and the current observation, the time risk quantification value, the power characteristic quantification value, and the index approximation degree of the risk index to be adjusted are determined respectively. The dynamic weights of the risk indicators to be adjusted are corrected by using the time risk quantification value, the power characteristic quantification value, and the indicator approximation degree to obtain the corrected weights; The real-time risk value of the risk indicator to be adjusted is corrected using the corrected weight, and the corrected risk value is combined with other real-time risk values ​​other than the risk indicator to be adjusted to obtain the comprehensive risk index of the distribution area.

5. A distributed cooperative control method for a distribution radio area according to claim 1, characterized in that, The system controls the exchange of status information between each inverter node in the distribution transformer area, and, based on an evaluation system that includes controllable capacity, communication quality, and topology centrality, determines the master node and several slave nodes through a distributed election algorithm, including: Each inverter node is controlled to broadcast its own status information to neighboring nodes via a wireless local area network, so that each inverter node can determine its topology centrality based on the status information of its neighboring nodes and interact with its own neighboring nodes; the status information includes controllable capacity, communication quality, and physical location information. Each inverter node is controlled to calculate its own known inverter node communication score based on its own state information and topology centrality, as well as the state information and topology centrality of its neighboring nodes, and then compare it with the neighboring nodes after interacting with them. After multiple rounds of information interaction and iteration, the inverter node with the highest communication score is designated as the master node, and the other inverter nodes besides the master node are designated as slave nodes.

6. The distributed cooperative control method for a distribution radio area according to claim 1, characterized in that, The master node calculates the total adjustment requirement based on the dominant risk type, and, combined with the maximum safe adjustment amount received from each slave node and the expected contribution, solves for the optimal resource allocation scheme, including: The master node calculates the total adjustment requirement based on the dominant risk type and broadcasts a collaboration request frame to all slave nodes; the collaboration request frame includes contribution calculation rules. Each slave node determines its maximum safe adjustment amount and corresponding cost data based on its current operating data, and determines its expected contribution based on the contribution calculation rules, and sends it to the master node; The master node controls the construction and solution of a resource allocation optimization model based on the maximum safe adjustment amount of each slave node received, its corresponding cost data, and expected contribution, to obtain the optimal resource allocation scheme.

7. A distributed cooperative control method for a distribution radio area according to claim 6, characterized in that, The process of constructing and solving the resource allocation optimization model to obtain the optimal resource allocation scheme includes: The resource allocation optimization model is constructed with the goal of minimizing the cost of fulfilling the total adjustment requirements. The resource allocation optimization model is solved with the constraints that the sum of the target adjustment amounts of each slave node is not less than the total adjustment requirement and that the target adjustment amount of each slave node is not greater than its own maximum safe adjustment amount. The power adjustment commands of each slave node are then used as the optimal resource allocation scheme.

8. A distributed cooperative control system for a distribution radio area, characterized in that, include: The risk calculation module is used to collect multi-dimensional operation data of the distribution substation in real time, and to fuse the current observation value and short-term forecast value through a weighting method to obtain the real-time risk value of multiple risk indicators corresponding to the multi-dimensional operation data. The risk correction module is used to introduce a risk adjustment mechanism based on the characteristics of the operating scenario to correct each of the real-time risk values, so as to obtain the comprehensive risk index and dominant risk type of the distribution area. The operational scenario features are used to characterize the operational status and dominant energy flow mode of the distribution substation. The node election module is used to control each inverter node in the distribution area to exchange status information when the comprehensive risk index exceeds a preset high-risk threshold, and to determine the master node and several slave nodes through a distributed election algorithm based on an evaluation system that includes controllable capacity, communication quality and topology centrality. The collaborative control module is used to control the master node to calculate the total adjustment requirement based on the dominant risk type, combine the maximum safe adjustment amount and expected contribution received from each of the slave nodes, solve for the optimal resource allocation scheme, and send it to the corresponding slave nodes to control each of the slave nodes to execute.

9. An electronic device, characterized in that, It includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the distributed cooperative control method for a distribution radio area as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein when the device containing the computer-readable storage medium executes the computer program, it implements the distributed cooperative control method for a distribution radio area as described in any one of claims 1 to 7.

Images