Transient stability control method for power distribution network containing high-proportion distributed energy

Abstract

The invention provides a transient stability control method for a power distribution network containing high-proportion distributed energy, and belongs to the technical field of electrical variable adjustment and control, and the method comprises the steps: obtaining real-time monitoring data, distributed energy output information and external environment data of the power distribution network, carrying out the fusion processing of the obtained data, and generating comprehensive monitoring data; performing transient stability prediction analysis on the comprehensive monitoring data, and outputting a prediction result; performing collaborative decision on the prediction result and the comprehensive monitoring data, and generating a dynamic adaptive adjustment strategy for cooperatively controlling the distributed energy equipment, the energy storage equipment and the controllable load; and executing the dynamic self-adaptive adjustment strategy, collecting feedback data after execution, and performing secondary adjustment by using the feedback data, so that the voltage parameter or the frequency parameter approaches a preset reference value. According to the invention, the transient stability of the voltage or frequency of a distributed energy grid-connected point and a sensitive load node can be realized by adopting a real-time monitoring and predictive analysis technology and combining a dynamic adaptive adjustment strategy.

Description

Technical Field

[0001] This invention belongs to the field of electrical variable regulation and control technology, and in particular relates to a transient stability control method for a distribution network with a high proportion of distributed energy resources. Background Technology

[0002] In modern power regulation systems, especially in networked architectures that include multiple power supply units such as photovoltaics and wind power, the stability of output electrical parameters faces severe challenges. Because these distributed power supply units are connected via power electronic interfaces, their dynamic response characteristics differ from those of traditional rotating electric machines. When the system encounters dynamic disturbances such as sudden load changes, power outages, or source-load imbalances, the power supply units lack sufficient dynamic inertia support, leading to sharp fluctuations in key electrical parameters and even triggering system-wide cascading instability.

[0003] Existing technologies typically employ local compensation strategies to address such problems: a correction mechanism based on local steady-state parameters is set at the power supply unit interface, passively adjusting the output after detecting excessive electrical parameters. However, due to reliance on local information perception and a lack of global dynamic coordination, this can easily lead to multi-unit adjustment conflicts under systemic disturbances, and even exacerbate oscillations. Another approach is to adjust the reference output of the power supply unit at fixed intervals based on historical data or short-term prediction models. However, its response delay is difficult to match the time-varying characteristics of millisecond-level dynamic disturbances; essentially, it remains a post-event correction and cannot suppress the evolution of disturbance trends in advance.

[0004] The aforementioned methods have the following drawbacks: they rely on static models or steady-state parameters, making them insufficiently adaptable to nonlinear, strongly coupled transient processes of electrical variables; they lack cross-unit optimization and coordination mechanisms, failing to generate a unified control strategy based on the system's dynamic state; and the adjustment parameters are fixed, making it impossible to optimize the control response accuracy in real time based on execution feedback. Therefore, a precise adjustment method for electrical variables that can deeply integrate system dynamic prediction and adaptive closed-loop control is needed to improve the stability of multi-source power supply systems under complex disturbances. Summary of the Invention

[0005] The purpose of this invention is to propose a transient stability control method for distribution networks with a high proportion of distributed energy resources. By adopting real-time monitoring and predictive analysis technology and combining dynamic adaptive adjustment strategies, the method can achieve transient stability of voltage or frequency at distributed energy grid connection points and sensitive load nodes.

[0006] The above objectives can be achieved through the following approach:

[0007] A transient stability control method for a distribution network with a high proportion of distributed energy resources includes acquiring real-time monitoring data of the distribution network, distributed energy output information, and external environmental data. The real-time monitoring data includes voltage parameters, current parameters, and frequency parameters. The acquired data is fused to generate comprehensive monitoring data. Transient stability prediction analysis is performed on the comprehensive monitoring data, and prediction results are output. The prediction results and comprehensive monitoring data are used for collaborative decision-making to generate a dynamic adaptive adjustment strategy for the coordinated control of distributed energy devices, energy storage devices, and controllable loads. The dynamic adaptive adjustment strategy is executed, and feedback data after execution is collected. The feedback data is used for secondary adjustment to make the voltage parameters or frequency parameters approach preset reference values.

[0008] Optionally, generating comprehensive monitoring data includes: collecting real-time monitoring data of the power distribution network and external environmental data, the external environmental data including solar irradiance, ambient temperature and wind speed; obtaining distributed energy output information, including active power, reactive power, available power margin and operating mode; and integrating the real-time monitoring data, external environmental data and distributed energy output information to generate comprehensive monitoring data.

[0009] Optionally, the transient stability prediction analysis of the comprehensive monitoring data and the output prediction results include: extracting features from the comprehensive monitoring data at multiple time scales, and performing pattern matching in conjunction with a pre-established system operation state pattern library to identify the state, thereby obtaining a system transient feature sequence; and performing transient stability prediction analysis on the system transient feature sequence through a pre-trained model to output the prediction results.

[0010] Optionally, the generation of a dynamic adaptive adjustment strategy for the coordinated control of distributed energy devices, energy storage devices, and controllable loads includes: extracting power fluctuation trends based on the prediction results and generating fluctuation characteristic parameters; generating inertial adjustment parameters based on the fluctuation characteristic parameters; generating partition control commands by combining the inertial adjustment parameters and a preset partition coordination mechanism; and integrating the inertial adjustment parameters and the partition control commands to generate a dynamic adaptive adjustment strategy for the coordinated control of distributed energy devices, energy storage devices, and controllable loads.

[0011] Optionally, generating inertial adjustment parameters includes: calculating the power change rate based on the fluctuation characteristic parameters; generating preliminary inertial parameters based on the power change rate; obtaining the physical constraints of the distributed energy equipment and the safe operation boundary of the distribution network to form an inertial threshold; and using the inertial threshold to perform constraint optimization on the preliminary inertial parameters to generate inertial adjustment parameters.

[0012] Optionally, the generation of zonal control instructions includes: dividing the distribution network into multiple dynamic zones based on the network topology information in the integrated monitoring data; assigning control tasks to the multiple dynamic zones according to the inertial adjustment parameters and the preset zonal coordination mechanism, and generating local adjustment instructions; each dynamic zone coordinating the local adjustment instructions through inter-regional communication based on its real-time status and the assigned control tasks, and generating zonal control instructions.

[0013] Optionally, executing the dynamic adaptive adjustment strategy includes: parsing the dynamic adaptive adjustment strategy and generating a control signal; sending the control signal to the distributed energy device, energy storage device, and controllable load for execution; monitoring the execution status of the control signal and generating execution status data; and updating the dynamic adaptive adjustment strategy in real time based on the execution status data.

[0014] Optionally, the step of using the feedback data for secondary adjustment to make the voltage parameter or frequency parameter approach a preset reference value includes: performing deviation analysis and adaptive correction calculation on the voltage parameter or frequency parameter based on the feedback data to obtain the parameter adjustment amount; and performing multi-objective optimization and coordinated control on the parameter adjustment amount to make the voltage parameter or frequency parameter approach the reference value. The optimization objectives include minimizing the steady-state deviation of the frequency and voltage, and minimizing the adjustment cost.

[0015] Optionally, the method further includes: monitoring the operating status of the power distribution network in real time, generating an event trigger signal if a transient event is detected; and triggering the real-time update of the dynamic adaptive adjustment strategy and the increase of its execution priority based on the event trigger signal.

[0016] Based on the same inventive concept, this invention also provides a transient stability control system for a distribution network with a high proportion of distributed energy resources. The system includes: a data acquisition module for acquiring real-time monitoring data of the distribution network, distributed energy output information, and external environmental data, wherein the real-time monitoring data includes voltage parameters, current parameters, and frequency parameters; a data fusion processing module for performing data fusion processing on the real-time monitoring data, distributed energy output information, and external environmental data to generate comprehensive monitoring data; a prediction and analysis module for performing transient stability prediction analysis on the comprehensive monitoring data and outputting prediction results; a dynamic strategy generation module for making collaborative decisions based on the prediction results and the comprehensive monitoring data to generate a dynamic adaptive adjustment strategy for collaboratively controlling distributed energy devices, energy storage devices, and controllable loads; and a strategy execution and feedback control module for executing the dynamic adaptive adjustment strategy, collecting feedback data after execution, and using the feedback data for secondary adjustment to make the voltage parameters or frequency parameters approach a preset reference value.

[0017] Compared with the prior art, the present invention has the following advantages:

[0018] This invention enhances the transient stability of power distribution networks under high penetration of distributed energy resources. By acquiring comprehensive monitoring data in real time, including multi-source information such as voltage and frequency parameters, it outputs accurate prediction results and combines them with a dynamic adaptive adjustment strategy to quickly and collaboratively control distributed energy devices, energy storage devices, and controllable loads, thereby suppressing power fluctuations and voltage or frequency deviations. The feedback data after execution further optimizes parameter adjustment, ensuring that the system approaches the preset reference value when facing transient disturbances, reducing the risk of instability and shortening recovery time.

[0019] This invention achieves intelligent and adaptive operation and control capabilities. After generating prediction results based on feature extraction and state identification at multiple time scales, the dynamic adaptive adjustment strategy adaptively generates local adjustment commands and inertial adjustment parameters according to fluctuation characteristic parameters and partition coordination mechanisms. This process takes into account equipment constraints and network topology, realizes dynamic linkage of partitions and optimization of control priorities, enabling the system to flexibly respond to environmental changes and external event triggers, and improves overall resilience and fault tolerance.

[0020] This invention improves the resource coordination efficiency and safe operation boundary of the power distribution network. By integrating zone control mechanisms and local adjustment commands to coordinate the output of distributed energy devices, energy storage devices, and controllable loads, it optimizes power allocation and the generation of inertial regulation parameters. During execution, it monitors the strategy execution status and updates control signals in real time. Combined with multi-objective optimization and coordinated control to eliminate deviations, it not only reduces energy waste and equipment overload, but also ensures efficient and stable operation within the safe operation boundary.

[0021] This invention improves the dynamic response speed and parameter control accuracy of power distribution networks. After executing the dynamic adaptive adjustment strategy, feedback data is immediately collected, and deviation analysis and adaptive correction calculations are performed to generate parameter adjustment amounts. Multi-objective optimization is used to rapidly bring voltage or frequency parameters close to reference values, while an event-triggered signal mechanism prioritizes strategy updates and improves response levels. This shortens transient event processing time, improves control accuracy, and reduces system fluctuation frequency and power quality losses.

[0022] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart illustrating a transient stability control method for a distribution network with a high proportion of distributed energy resources, according to an embodiment of the present invention.

[0025] Figure 2 This is a comprehensive monitoring data integration diagram according to an embodiment of the present invention.

[0026] Figure 3 This is a multi-objective optimization parameter control surface diagram according to an embodiment of the present invention.

[0027] Figure 4 This is a schematic diagram of the structure of a transient stability control system for a power distribution network with a high proportion of distributed energy resources, according to an embodiment of the present invention. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] Reference Figure 1 One embodiment of the present invention proposes a transient stability control method for a distribution network with a high proportion of distributed energy resources. By employing real-time monitoring and predictive analysis technology, combined with a dynamic adaptive adjustment strategy, the method can achieve transient stability of voltage or frequency at the grid connection point of distributed energy resources and sensitive load nodes.

[0030] The method described in this embodiment specifically includes:

[0031] Acquire real-time monitoring data of the power distribution network, distributed energy output information, voltage parameters, frequency parameters, and external environmental data;

[0032] The real-time monitoring data, distributed energy output information, voltage parameters, frequency parameters, and external environmental data are fused together to generate comprehensive monitoring data.

[0033] Perform predictive analysis on the comprehensive monitoring data and output the prediction results;

[0034] The prediction results and the comprehensive monitoring data are used to make collaborative decisions to generate a dynamic adaptive adjustment strategy for the coordinated control of distributed energy devices, energy storage devices and controllable loads.

[0035] The dynamic adaptive adjustment strategy is executed, and feedback data is collected after execution. The feedback data is used to adjust the voltage parameter or frequency parameter to approach a preset reference value.

[0036] Specifically, firstly, multi-source data fusion technology integrates real-time electrical quantities of the distribution network, the output status of distributed energy resources, and external environmental influencing factors into a comprehensive and synchronous integrated monitoring dataset, laying the foundation for accurate characterization of the system state. Next, predictive analysis is performed using this dataset to proactively assess potential transient instability trends in the system. Based on this prediction and combined with the current real-time system state, a dynamic adaptive adjustment strategy is generated through collaborative decision-making. The core of this strategy is the collaborative control of widely distributed distributed energy devices, energy storage devices, and controllable loads, enabling them to participate as a whole in system stability regulation. Finally, this method forms a complete closed-loop control. After strategy execution, secondary adjustments are performed by collecting feedback data, precisely guiding key system parameters such as voltage or frequency to preset stable reference values, completing the entire control process from prediction, decision-making, execution to correction.

[0037] Optionally, the generation of comprehensive monitoring data includes:

[0038] Collect real-time monitoring data, voltage parameters, frequency parameters, and external environmental data of the power distribution network;

[0039] Obtain distributed energy output information;

[0040] By integrating the real-time monitoring data, voltage parameters, frequency parameters, external environmental data, and the distributed energy output information, comprehensive monitoring data is generated.

[0041] Specifically, the first step is to construct a multi-source heterogeneous data acquisition system. This system deploys phasor measurement units (PMUs) at key nodes in the distribution network, such as substation busbars, feeder heads, and points of connection with large-capacity distributed energy sources. These PMUs are used to collect dynamic electrical quantities such as voltage and current phasors, node frequencies, and frequency change rates at high frequencies. This data is precisely synchronized using GPS to generate high-precision real-time monitoring data. Simultaneously, remote terminal units (RTUs) and intelligent electronic devices (IEDs) in the distribution automation system are used to collect traditional steady-state operating information such as power flow and switch status of various lines in the distribution network, supplementing the real-time monitoring data. To address external environmental factors affecting distributed energy output, meteorological monitoring stations are deployed in typical areas to collect data on solar irradiance, ambient temperature, and wind speed. Secondly, distributed energy output information needs to be acquired. This requires establishing communication connections with the energy management systems (EMS) or inverter controllers of each distributed energy station to obtain real-time status information such as the actual active power, reactive power output, available power margin, and operating mode of each distributed energy unit. This information, along with the distribution network monitoring data, is transmitted to the central control system. Finally, the collected and acquired multi-source data are integrated and processed to generate structured comprehensive monitoring data. This integrated monitoring data is as follows: Figure 2 As shown. The integration process includes the following steps: First, data time synchronization, aligning data from different sources and sampling rates, such as PMUs, RTUs, and distributed energy controllers, based on a unified GPS timestamp to ensure consistency across all data in the time dimension. Second, data cleaning and verification, removing outliers, repairing bad data, and verifying logical consistency in the synchronized data sequence. For example, interpolating missing measurement data using state estimation methods to ensure data integrity and accuracy. Third, data fusion and formatting, fusing the cleaned and verified real-time monitoring data of the distribution network, voltage parameters, frequency parameters, external environmental data, and distributed energy output information into a unified system state vector. This vector can be represented as:

[0042] ,

[0043] In the formula, for A vector of comprehensive monitoring data at any given moment; It is the electrical state sub-vector of the distribution network, which includes real-time monitoring data such as the voltage phasor of each node, the line current phasor, and the system frequency obtained through PMU and RTU; The distributed energy output information subvector includes the actual active and reactive power obtained from each distributed energy site; This is a subvector of external environmental data, including solar irradiance and wind speed obtained from weather stations; This represents the transpose of the matrix. Through the above steps, comprehensive monitoring data that fully, accurately, and in real-time reflects the dynamic characteristics of the distribution network is finally generated, providing a high-quality data foundation for subsequent transient stability prediction and control.

[0044] Optionally, the predictive analysis of the comprehensive monitoring data, and the output of the prediction results, include:

[0045] The comprehensive monitoring data is subjected to feature extraction and state identification at multiple time scales to obtain a system transient feature sequence;

[0046] Perform transient stability prediction analysis on the transient characteristic sequence of the system and output the prediction results.

[0047] Specifically, the data first needs to be deeply processed to extract key information reflecting the system's dynamic behavior. The first step in this process is feature extraction and state identification at multiple time scales. For the acquired comprehensive monitoring data, signal processing techniques are used to decompose it at different time scales. For example, continuous wavelet transform is applied to analyze the transient components of key electrical quantities such as voltage and frequency to capture short-term events such as high-frequency oscillations and sudden changes; simultaneously, statistical methods such as moving average and root mean square are used to analyze data within longer time windows to extract medium- and long-term features such as power fluctuation trends and average deviations. Through this step, a series of quantitative indicators describing the system's dynamic characteristics can be obtained, such as the rate of frequency change, which is calculated as follows:

[0048] ,

[0049] in, express Rate of change of frequency at any given moment; The frequency parameter is obtained from the comprehensive monitoring data at the current moment; The frequency parameter at the previous sampling time; The sampling time interval is defined as follows. Similarly, key features such as voltage change rate and active power ramp-up rate of distributed energy sources can also be calculated. By combining these features extracted at different time scales and performing pattern matching with a pre-established system operating state pattern library, state identification can be completed, forming a system transient feature sequence containing a summary of the system's current and recent dynamic behavior. The next step is to perform transient stability prediction analysis on this system transient feature sequence and output the final prediction result. The core of this step is to build an intelligent model capable of processing time-series data and having predictive capabilities, such as a Long Short-Term Memory (LSTM) network or a Gated Recurrent Unit (GRU). This model needs to be pre-trained offline using a large amount of simulation data and historical fault data to learn the complex nonlinear mapping relationship from the system transient feature sequence to the future stable state. During actual operation, the real-time generated system transient feature sequence is used as input and fed into this trained prediction model. By analyzing the evolution trend and intrinsic correlation of features in the sequence, the model predicts the system transient stability within a very short future time window. The output prediction result is a quantitative transient stability index (TSI), which comprehensively evaluates the system's ability to maintain synchronous operation under disturbances. For example, the TSI value range can be set from 0 to 1. The closer the value is to 1, the more stable the system is. If the value drops rapidly and approaches 0, it indicates that the system may become unstable.

[0050] Optionally, the dynamic adaptive adjustment strategy for collaboratively controlling distributed energy devices, energy storage devices, and controllable loads includes:

[0051] Based on the prediction results, the power fluctuation trend is extracted, and fluctuation characteristic parameters are generated;

[0052] Based on the aforementioned wave characteristic parameters, inertial adjustment parameters are generated;

[0053] By combining the inertial adjustment parameters and the preset partition coordination mechanism, partition control commands are generated;

[0054] By integrating the inertial adjustment parameters and the partition control commands, a dynamic adaptive adjustment strategy for the coordinated control of distributed energy devices, energy storage devices, and controllable loads is generated.

[0055] Specifically, the power fluctuation trend is first extracted based on the prediction results, generating fluctuation characteristic parameters. This step involves in-depth analysis of the time-series changes of the Transient Stability Index (TSI) output by the transient stability prediction analysis, as well as the prediction curves for key variables such as future system frequency and power deficit. By performing mathematical processing such as differentiation and peak finding on these prediction curves, the system dynamics that may occur in the short term are quantified. For example, the estimated maximum rate of frequency change is extracted from the predicted frequency curve. The predicted maximum power imbalance is extracted from the predicted power balance curve at the lowest frequency point. These quantified indicators collectively constitute the fluctuation characteristic parameters characterizing the future severity of system fluctuations. Next, based on these fluctuation characteristic parameters, inertial regulation parameters are generated. Inertial regulation parameters are key coefficients determining the strength of the synthetic inertia and damping support required by power electronic devices such as distributed inverter power supplies and energy storage systems to provide virtual synchronous machine functionality. This step establishes a mapping function from fluctuation characteristic parameters to inertial regulation parameters. The core idea of ​​this function is that the more severe the predicted system fluctuations, the stronger the required inertial support. For example, a key inertial regulation parameter is the virtual rotational inertia. It can be determined based on the predicted rate of change of frequency, and the relationship can be expressed as:

[0056] ,

[0057] in, These are the virtual moment of inertia parameters to be generated, which determine the speed and intensity of the device's response to changes in the system's frequency. It is one of the fluctuation characteristic parameters generated in the previous step, representing the predicted rate of change of frequency; It is a nonlinear gain function that ensures that when the predicted rate of frequency change is large, the virtual moment of inertia can be dynamically and nonlinearly increased, thereby providing stronger inertial support. Subsequently, combined with the inertial adjustment parameters and the preset zoning coordination mechanism, zoning control commands are generated. First, the distribution network is dynamically divided into several relatively independent control areas based on its network topology, line impedance, and controllable resource distribution. The preset zoning coordination mechanism is a set of optimized allocation rules designed to rationally decompose the overall control task based on the adjustment capacity, electrical distance, and impact on global stability of each area. This step transforms the inertial adjustment parameters, representing the required global support strength, generated in the previous step, into specific control objectives for each zone, i.e., zoning control commands, according to the zoning coordination mechanism. These commands clarify the share of inertial support responsibility or power adjustment range that each zone needs to undertake. Finally, the inertial adjustment parameters and the zoning control commands are integrated to generate the final dynamic adaptive adjustment strategy. The inertial adjustment parameters define the dynamic characteristics of the control response, i.e., how to respond; while the zoning control commands specify the allocation scheme of control actions, i.e., where to respond and how much to respond. Integrating these two elements creates a hierarchical and structured collaborative control blueprint. This dynamic adaptive adjustment strategy details how distributed energy devices, energy storage devices, and controllable loads across the network should proactively and differentiatedly respond in a collaborative manner to impending transient events, thus forming a complete and executable control scheme.

[0058] Optionally, the generated inertial adjustment parameters include:

[0059] The power change rate is calculated based on the fluctuation characteristic parameters.

[0060] Based on the power change rate, preliminary inertial parameters are generated;

[0061] Obtain the physical constraints of the distributed energy equipment and the safe operation boundary of the power distribution network to form an inertial threshold;

[0062] The initial inertial parameters are constrained and optimized using the inertial threshold to generate inertial adjustment parameters.

[0063] Specifically, the power change rate is first calculated based on the aforementioned fluctuation characteristic parameters. This step is crucial for quantifying the intensity of the power surge the system will face. More specifically, a predicted time series of the active power imbalance in the distribution network over a short period is extracted from the fluctuation characteristic parameters generated in the previous step. By differentiating the predicted sequence over time, the predicted power change rate of the system can be obtained. This rate characterizes the growth rate of power deficit or excess during transient events. Next, based on this power change rate, preliminary inertial parameters are generated. These preliminary inertial parameters are idealized control parameters without considering any physical or operational constraints. The logic behind their generation is that the larger the predicted power change rate, the greater the required inertial support strength should be to effectively suppress rapid frequency changes. For example, a preliminary virtual moment of inertia can be set. Its value is directly proportional to the absolute value of the predicted power change rate, and its calculation can be expressed as: ,

[0064] In this formula, These are the initial inertial parameters to be generated, i.e., the initial virtual moment of inertia; It is a pre-set gain coefficient, the value of which is determined through offline system simulation and stability analysis, and is used to calibrate the proportional relationship between the power change rate and the required inertial support; This represents the power change rate calculated based on the fluctuation characteristic parameters. Then, the physical constraints of the distributed energy devices and the safe operating boundaries of the distribution network are obtained to form the inertia threshold. This step is fundamental to ensuring the practical feasibility of the control strategy. For distributed energy devices, the physical constraints mainly include the upper and lower limits of the state of charge (SOC) of the energy storage devices and their maximum charging and discharging power, the rated capacity of the inverter, and the DC-side voltage stability range. For the distribution network, the safe operating boundaries include that the voltage deviation of critical nodes must not exceed the legal range, and the current carrying capacity of lines and transformers must not exceed their thermal stability limits. These constraints collectively define the maximum regulating power that a single device or local grid can safely withstand or generate. By substituting these power constraints inversely into the virtual inertia control equation, the maximum allowable value of the inertia regulation parameters can be calculated. and minimum value These two factors together constitute the inertia threshold. Finally, the inertia threshold is used to constrain and optimize the preliminary inertia parameters, generating the final inertia adjustment parameters. This process corrects the idealized preliminary parameters to adapt them to physical reality. The preliminary inertia parameters generated in the previous step are then... With inertia threshold and Compare. If If it is within the threshold range, then the final inertial adjustment parameter is equal to ;like Exceeded Then the inertial adjustment parameter is saturated and set to... ;like Below Then the inertia adjustment parameter is set to This process ensures that the final inertial regulation parameters not only respond to the system's stability requirements to the greatest extent possible, but also never exceed the safe tolerance of any related equipment or power grid component.

[0065] Optionally, the generation of partition control instructions includes:

[0066] Based on the network topology information in the comprehensive monitoring data, the power distribution network is divided into multiple dynamic areas;

[0067] Based on the inertial adjustment parameters and the preset partition coordination mechanism, control tasks are assigned to the multiple dynamic regions, and local adjustment instructions are generated.

[0068] The local adjustment instructions are coordinated through communication between the dynamic regions to generate partition control instructions.

[0069] Specifically, firstly, based on the network topology information in the comprehensive monitoring data, the distribution network is divided into multiple dynamic regions. This step does not employ fixed geographical or administrative divisions, but rather utilizes a community detection algorithm from graph theory to dynamically partition the distribution network in real time. The algorithm's input is a weighted adjacency graph of the power grid constructed based on the comprehensive monitoring data. Nodes in the graph represent buses or load points, and the edge weights are determined by the electrical distance between nodes. Electrical distance is a quantity characterizing the degree of electrical coupling between two points in the power grid, typically proportional to the line impedance between them, and can be calculated based on real-time monitored voltage and current data. This algorithm aggregates a group of electrically connected and mutually influential nodes into a dynamic region, thereby ensuring a more efficient and direct control response within the region. Then, according to the inertial adjustment parameters and a preset partitioning coordination mechanism, control tasks are assigned to the multiple dynamic regions, generating local adjustment commands. The preset partitioning coordination mechanism is an optimized allocation strategy whose goal is to use inertial adjustment parameters generated in the previous step, representing global requirements, such as the total virtual rotational inertia requirement. This involves fairly and efficiently allocating resources to various dynamic zones. The allocation is based on factors including the controllable resources within each dynamic zone, such as distributed energy devices and energy storage devices, their total regulation capacity, response speed, and the zone's impact on the overall network's transient stability or sensitivity. For example, a dynamic zone... The control task, namely the inertial regulation parameters it should undertake. According to its regulatory capacity The allocation is based on the proportion of the total network regulation capacity, calculated as follows:

[0070] ,

[0071] in It is allocated to the dynamic region. The local adjustment instructions, i.e., the virtual inertia that the region needs to provide; It is the total virtual inertia required globally, determined by the inertia adjustment parameters; It is a region The total maximum adjustable power that all controllable devices inside can provide is obtained from the device status information in the comprehensive monitoring data; It is the sum of the adjustment capabilities of all dynamic regions across the entire network. This serves as the index variable for the dynamic regions. Next, the local adjustment instructions are coordinated through communication between the dynamic regions to generate the final partition control instructions. Upon receiving the initial local adjustment instructions, each dynamic region's local controller combines this with more refined real-time status within its region, such as the energy storage charge state, to perform a feasibility check. If a region cannot fully meet its assigned tasks, it will broadcast its adjustment deficit to adjacent or designated coordination regions through an inter-regional communication network, such as fiber-optic-based industrial Ethernet. Other region controllers with sufficient capacity, upon receiving this information, will proactively assume part or all of the deficit based on a preset collaborative response protocol and update their own adjustment instructions accordingly. This process may iterate several times until the control tasks of the entire network are fully covered by the actual feasible adjustment capabilities of all regions, forming a globally coordinated and locally feasible instruction set—the final partition control instructions.

[0072] Optionally, executing the dynamic adaptive adjustment strategy includes:

[0073] The dynamic adaptive adjustment strategy is analyzed to generate control signals;

[0074] The control signal is sent to the distributed energy equipment, energy storage equipment, and controllable load for execution.

[0075] Monitor the execution status of the control signals and generate execution status data;

[0076] The dynamic adaptive adjustment strategy is updated in real time based on the execution status data.

[0077] Specifically, the dynamic adaptive adjustment strategy is first analyzed to generate control signals. This step transforms the macro-level strategy defined at the upper level into specific operational instructions that the lower-level devices can directly recognize and execute. The dynamic adaptive adjustment strategy includes inertial adjustment parameters and power scheduling targets for each distributed energy device, energy storage device, and controllable load. The execution unit, typically a zone controller or energy management system, needs to convert these parameters into the internal control parameters of the corresponding device controller. For example, for an inverter using virtual synchronous machine control, the virtual inertia setting value in its internal controller needs to be... Meanwhile, its active power reference value will also be based on the target power. Adjustments are made. This parsing process ultimately generates a series of standardized control signals. These control signals are then sent to the distributed energy devices, energy storage devices, and controllable loads for execution. This requires a reliable, low-latency communication network. The control signals are transmitted from the area controller or central controller to each device-level controller in the field, such as photovoltaic inverters, energy storage converters (PCS), or smart load switches, via fiber optics, 5G private networks, or other dedicated communication links. Upon receiving the signals, the device controllers immediately adjust their operating logic and power output to participate in the transient stability support of the power grid according to the instructions. Then, the execution of the control signals is monitored, generating execution status data. After the control signals are sent, the control system does not terminate its operation but continuously monitors the actual response of each controlled device in real time through a high-frequency data acquisition system. The monitored data includes the actual power output, terminal voltage, current, and internal states of the device, such as the SOC change rate of energy storage. These actual response data are compared with the sent control signals to analyze the deviation and latency between them. For example, the actual power response is calculated. With target power The difference :

[0078] ,

[0079] These deviations, delays, and the real-time operating status of the equipment collectively constitute the execution status data. Finally, the dynamic adaptive adjustment strategy is updated in real time based on the execution status data. This step constitutes a rapid inner-loop feedback in the control process. When the monitoring system detects a deviation between the actual response of a device and the command, such as due to communication interruption, equipment failure, or reaching physical limits, or when it detects that the actual transient evolution of the power grid does not match the initial prediction, the execution status data will trigger the real-time update mechanism of the adjustment strategy. For example, if an energy storage unit cannot provide the required power support due to a low SOC, the system will immediately recalculate the task allocation based on the reported execution status data, dynamically transferring the adjustment task of the energy storage unit to other controllable resources with sufficient margin. This real-time update ensures that the entire control strategy can always adapt to the constantly changing field conditions and power grid dynamics, forming a closed-loop self-correction process.

[0080] Optionally, adjusting the voltage parameter or frequency parameter using the feedback data to approach a preset reference value includes:

[0081] Based on the feedback data, deviation analysis and adaptive correction calculations are performed on the voltage or frequency parameters to obtain the parameter adjustment amount;

[0082] The parameter adjustment is optimized and coordinated by multiple objectives to make the voltage parameter or frequency parameter approach the reference value.

[0083] Specifically, firstly, based on the feedback data, deviation analysis and adaptive correction calculations are performed on the voltage or frequency parameters to obtain the parameter adjustment amount. The feedback data is the actual power grid operation data continuously collected by high-precision measurement equipment across the entire network after the dynamic adaptive adjustment strategy is implemented. This data mainly includes the actual voltage parameters of each key bus and the actual frequency parameters of the system. Deviation analysis involves comparing these feedback data with preset reference values ​​to calculate the steady-state deviation. The preset reference values ​​are usually the nominal values ​​of the power grid, such as frequency reference values. 50 Hz, voltage reference value The standard value is 1.0. Therefore, the frequency deviation can be obtained. and voltage deviation at each node Subsequently, adaptive correction calculations are performed based on these deviations to generate parameter adjustment amounts. This calculation typically employs an adaptive proportional-integral (PI) control algorithm. For example, to eliminate steady-state frequency deviations, a parameter adjustment amount for active power needs to be generated, which can be calculated as follows:

[0084]

[0085] in, yes The total active power adjustment calculated at any given time; yes The frequency deviation at time is equal to Subtract the actual frequency from the feedback data. This represents the time from the start of the secondary adjustment phase following the transient event to the current calculation time. any point in time between; and These are the proportional and integral gain coefficients, respectively. These coefficients are adaptive, and their values ​​are adjusted online based on the current grid conditions, such as the system's moment of inertia and load levels, to ensure the speed and stability of the regulation process. Similarly, the required reactive power regulation can be calculated based on the voltage deviation at each node. Next, multi-objective optimization and coordinated control is performed on the parameter adjustment amount to make the voltage parameter or frequency parameter approach the reference value. This step addresses how to adjust the calculated total adjustment amount. and The goal is to rationally allocate resources among numerous distributed energy devices, energy storage devices, and controllable loads across the entire network. This is a multi-objective optimization problem, whose objective function typically includes minimizing steady-state deviations in frequency and voltage, while simultaneously minimizing regulation costs, such as incremental network losses or the cost of calling up energy storage devices. The multi-objective optimization parameter control surface is as follows: Figure 3 As shown. The decision variables of the optimization problem are assigned to the first... 'Active power adjustment command of a controllable device' and reactive power regulation commands The optimization process needs to be performed under a series of constraints, including that the sum of all device adjustment commands should equal the total parameter adjustment amount, i.e. equal and equal Meanwhile, the adjustment of each device cannot exceed its own physical constraints, such as maximum power limits and energy storage state of charge range. By solving this multi-objective optimization problem, a set of optimal control command allocation schemes is obtained, and these commands are issued to each device for execution, thereby collaboratively and economically restoring the system's voltage and frequency parameters to near the reference values.

[0086] Optionally, the method further includes:

[0087] The system monitors the operating status of the power distribution network in real time, and generates an event trigger signal if a transient event is detected.

[0088] Based on the event trigger signal, the dynamic adaptive adjustment strategy is updated in real time and its execution priority is increased.

[0089] Specifically, the operation status of the distribution network is first monitored in real time. If a transient event is detected, an event trigger signal is generated. This process relies on a highly sensitive transient event monitoring module, which analyzes comprehensive monitoring data acquired from advanced measurement equipment in parallel and continuously. Event identification does not rely on a single threshold but rather on a decision logic based on multi-dimensional feature fusion. The monitoring module calculates key indicators in real time, such as the rate of frequency change, voltage dip depth, and power surge. When one or more combinations of these indicators exceed preset dynamic boundaries, a transient event is determined to have occurred. For example, when the absolute value of the rate of frequency change exceeds a certain threshold and the duration of the change exceeds several milliseconds, it can be determined as a severe power imbalance event. Once the identification conditions are met, the monitoring module immediately generates a digital event trigger signal. This signal is an internal interrupt signal with the highest priority, and its content may include the preliminary judgment of the event type, such as a line short-circuit fault or high-capacity load switching, and the timestamp of the event occurrence. Then, based on the event trigger signal, the dynamic adaptive adjustment strategy is triggered to update in real time and its execution priority is increased. When the main processor of the control system receives this high-priority event trigger signal, it immediately interrupts the currently executing routine or periodic calculation tasks. System resources, including computing power and communication bandwidth, are preferentially allocated to the transient stability control process. This trigger signal immediately initiates or forcibly refreshes the entire dynamic adaptive adjustment strategy generation process. This means that the system will immediately call upon the latest comprehensive monitoring data, re-perform multi-timescale feature extraction and transient stability prediction analysis, and generate a completely new dynamic adaptive adjustment strategy specific to this particular event. Compared with conventional periodic updates, this event-triggered update is instantaneous, significantly reducing the delay between the event occurrence and the formation of control decisions. Simultaneously, the generated control signals are marked with the highest execution priority when issued, ensuring they are transmitted with minimal delay in the communication network and unconditionally and immediately executed by the field device controllers, thus intervening before the transient process deteriorates.

[0090] Based on the same inventive concept, such as Figure 4 As shown, the present invention also provides a transient stability control system for a distribution network with a high proportion of distributed energy resources, the system comprising:

[0091] The data acquisition module is used to acquire real-time monitoring data of the power distribution network, distributed energy output information, voltage parameters, frequency parameters, and external environmental data.

[0092] The data fusion processing module is used to perform data fusion processing on the real-time monitoring data, distributed energy output information, voltage parameters, frequency parameters and external environmental data to generate comprehensive monitoring data;

[0093] The prediction and analysis module is used to perform predictive analysis on the comprehensive monitoring data and output the prediction results;

[0094] The dynamic strategy generation module is used to make collaborative decisions based on the prediction results and the comprehensive monitoring data to generate dynamic adaptive adjustment strategies for the collaborative control of distributed energy devices, energy storage devices and controllable loads.

[0095] The strategy execution and feedback control module is used to execute the dynamic adaptive adjustment strategy, collect feedback data after execution, and use the feedback data to adjust the voltage parameter or frequency parameter to approach a preset reference value.

[0096] To verify the feasibility of this invention in practice, it was applied to the power distribution network of a high-tech industrial park. This park's power distribution network is connected to a high proportion of photovoltaic power plants and wind turbines, and also supplies power to companies such as semiconductor manufacturers that have extremely high power quality requirements. Due to the intermittency and volatility of distributed energy resources, this power distribution network frequently suffers from transient disturbances, posing a serious threat to the stable production of companies in the park. The park hopes to use the method of this invention to improve the transient stability level and power supply reliability of the power distribution network under various disturbances.

[0097] In this embodiment, the transient stability control method described in this invention is deployed in the park's power distribution network control center. The system first constructs a multi-source heterogeneous data acquisition and synchronization system using phasor measurement units (PMUs) and intelligent electronic devices (IEDs) deployed at key nodes and lines, as well as solar intensity meters and wind turbines installed in photovoltaic and wind farms. All data, synchronized via GPS clock, is integrated into comprehensive monitoring data containing grid electrical status, distributed energy output, and external environmental information. Based on this data, the system uses a Long Short-Term Memory (LSTM) network model to predict transient stability and generate a transient stability index. Once an instability risk is predicted, the system generates a dynamic adaptive adjustment strategy to proactively suppress transient fluctuations by coordinating the control of distributed energy inverters, energy storage systems, and controllable loads within the park.

[0098] To verify the effectiveness of this invention, a typical transient event that occurred on a certain day of a certain month was selected for analysis. At a certain time on that day, a main power supply line in the park experienced a transient short-circuit fault due to external damage, causing severe fluctuations in system voltage and frequency.

[0099] Prior to the incident, the system continuously acquired comprehensive monitoring data of the distribution network. Voltage amplitude and phase angle were collected at a high sampling rate by PMUs deployed on each bus, and combined with line power flow data acquired by IEDs to form a real-time monitoring dataset for the distribution network. Simultaneously, the system communicated with each photovoltaic inverter via the IEC61850 protocol to obtain its real-time active power output, reactive power output, and reserve capacity, forming a distributed energy output information set. This data is identical to the external environmental datasets such as solar irradiance and wind speed provided by weather stations; after precise timestamp alignment, they were merged into unified comprehensive monitoring data.

[0100] When the event occurred, the system rapidly calculated transient features such as the rate of change of frequency by extracting features from comprehensive monitoring data across multiple time scales. Data showed that the system's rate of change of frequency surged dramatically within a short period. This feature sequence was input into a pre-trained LSTM prediction model. The model output indicated that the transient stability index was expected to rapidly decrease from a stable 0.95 to below the danger threshold of 0.4 within the next 500 milliseconds, suggesting a risk of power angle instability in the system.

[0101] Based on this prediction, the system immediately initiates the generation of a dynamic adaptive adjustment strategy. First, the system extracts the estimated maximum power imbalance from the predicted power balance curve and calculates the power change rate, generating preliminary inertial parameters. Subsequently, the system acquired physical constraints such as the state of charge (SOC) of the energy storage devices and the rated capacity of the inverters within the park, thus forming an inertial threshold. and Through the analysis of Constraint optimization was performed, and feasible inertial adjustment parameters were finally generated.

[0102] Next, based on the network topology, the system dynamically divides the park's power distribution network into three control zones. Then, according to the adjustment capabilities of controllable resources within each zone, and through a pre-defined zone coordination mechanism, the inertial adjustment parameters required for the global inertial adjustment task are decomposed into local adjustment commands for each zone. After receiving the commands and performing local verification, each zone controller quickly coordinates through inter-zone communication, ultimately generating globally consistent zone control commands.

[0103] These instructions are parsed into specific control signals and distributed via the 5G private network to various distributed energy inverters, energy storage converters, and controllable loads. For example, the photovoltaic inverters in area A are instructed to enter virtual synchronous machine mode, providing specified virtual rotational inertia support; the energy storage system in area B is instructed to rapidly output active power to compensate for power deficits. During the execution of control instructions, the system monitors the actual response of each device in real time and fine-tunes the adjustment strategy based on the execution status data, ensuring the achievement of the overall control objectives.

[0104] Once the transient process is successfully suppressed and the system frequency and voltage stabilize, the system enters the secondary regulation phase. Analyzing the feedback data, the system calculates that a steady-state frequency deviation of 0.05 Hz still exists, and the voltage at some key nodes deviates from the reference value by approximately 2%. The system then constructs a multi-objective optimization problem aimed at minimizing regulation costs and power quality deviations, and optimally allocates the calculated total active and reactive power regulation to each controllable resource, enabling the system frequency and voltage to accurately recover to the reference value within seconds of the event occurring.

[0105] It should be noted that the electrical connections between the various units described above do not necessarily represent direct or indirect connections. Any indirect connection method can be applied to the embodiments of the present invention as long as it achieves the purpose of the present invention. The above descriptions are merely exemplary embodiments of the present invention and should not be construed as limiting the scope of the present invention.

[0106] All equivalent changes and modifications made in accordance with the teachings of this invention are still within the scope of this invention. Those skilled in the art will readily conceive of other embodiments of this invention upon considering the specification and the disclosure of practical truth. This application is intended to cover any variations, uses, or adaptations of this invention that follow the general principles of this invention and include common knowledge or conventional techniques in the art not described herein.

Claims

1. A transient stability control method for a distribution network containing a high proportion of distributed energy resources, characterized in that, The method includes: The system acquires real-time monitoring data of the power distribution network, distributed energy output information, and external environmental data, including voltage parameters, current parameters, and frequency parameters. The real-time monitoring data, distributed energy output information, and external environmental data are fused together to generate comprehensive monitoring data. Perform transient stability prediction analysis on the comprehensive monitoring data and output the prediction results; The prediction results and the comprehensive monitoring data are used to make collaborative decisions to generate a dynamic adaptive adjustment strategy for the coordinated control of distributed energy devices, energy storage devices and controllable loads. The dynamic adaptive adjustment strategy is executed, and feedback data after execution is collected. The feedback data is used for secondary adjustment so that the voltage parameter or frequency parameter approaches the preset reference value.

2. The transient stability control method for a distribution network with a high proportion of distributed energy resources according to claim 1, characterized in that, The generated comprehensive monitoring data includes: Real-time monitoring data and external environmental data of the power distribution network are collected, including solar irradiance, ambient temperature and wind speed. Obtain distributed energy output information, including active power, reactive power, available power margin, and operating mode; The real-time monitoring data, external environmental data, and distributed energy output information are integrated to generate comprehensive monitoring data.

3. The transient stability control method for a distribution network with a high proportion of distributed energy resources according to claim 1, characterized in that, The transient stability prediction analysis performed on the comprehensive monitoring data outputs prediction results including: The comprehensive monitoring data is subjected to feature extraction at multiple time scales, and pattern matching is performed in combination with a pre-established system operation state pattern library to identify the state and obtain the system transient feature sequence. The transient stability prediction analysis of the system's transient feature sequence is performed using a pre-trained model, and the prediction results are output.

4. The transient stability control method for a distribution network with a high proportion of distributed energy resources according to claim 1, characterized in that, The dynamic adaptive adjustment strategy for coordinating the control of distributed energy devices, energy storage devices, and controllable loads includes: Based on the prediction results, the power fluctuation trend is extracted, and fluctuation characteristic parameters are generated; Based on the aforementioned wave characteristic parameters, inertial adjustment parameters are generated; By combining the inertial adjustment parameters and the preset partition coordination mechanism, partition control commands are generated; By integrating the inertial adjustment parameters and the partition control commands, a dynamic adaptive adjustment strategy for the coordinated control of distributed energy devices, energy storage devices, and controllable loads is generated.

5. The transient stability control method for a distribution network with a high proportion of distributed energy resources according to claim 4, characterized in that, The generated inertial adjustment parameters include: The power change rate is calculated based on the fluctuation characteristic parameters. Based on the power change rate, preliminary inertial parameters are generated; Obtain the physical constraints of the distributed energy equipment and the safe operation boundary of the power distribution network to form an inertial threshold; The initial inertial parameters are constrained and optimized using the inertial threshold to generate inertial adjustment parameters.

6. The transient stability control method for a distribution network with a high proportion of distributed energy resources according to claim 4, characterized in that, The generated partition control instructions include: Based on the network topology information in the comprehensive monitoring data, the power distribution network is divided into multiple dynamic areas; Based on the inertial adjustment parameters and the preset partition coordination mechanism, control tasks are assigned to the multiple dynamic regions, and local adjustment instructions are generated. Each dynamic region generates a partition control command by coordinating the local adjustment commands through inter-regional communication based on its real-time status and the assigned control tasks.

7. The transient stability control method for a distribution network with a high proportion of distributed energy resources according to claim 1, characterized in that, The execution of the dynamic adaptive adjustment strategy includes: The dynamic adaptive adjustment strategy is analyzed to generate control signals; The control signal is sent to the distributed energy equipment, energy storage equipment, and controllable load for execution. Monitor the execution status of the control signals and generate execution status data; The dynamic adaptive adjustment strategy is updated in real time based on the execution status data.

8. The transient stability control method for a distribution network with a high proportion of distributed energy resources according to claim 1, characterized in that, The step of using the feedback data for secondary adjustment to make the voltage parameter or frequency parameter approach a preset reference value includes: Based on the feedback data, deviation analysis and adaptive correction calculations are performed on the voltage or frequency parameters to obtain the parameter adjustment amount; The parameter adjustment is optimized and coordinated by multiple objectives to make the voltage parameter or frequency parameter approach the reference value. The optimization objectives include minimizing the steady-state deviation of frequency and voltage, and minimizing the adjustment cost.

9. The transient stability control method for a distribution network with a high proportion of distributed energy resources according to claim 1, characterized in that, The method further includes: The system monitors the operating status of the power distribution network in real time, and generates an event trigger signal if a transient event is detected. Based on the event trigger signal, the dynamic adaptive adjustment strategy is updated in real time and its execution priority is increased.

10. A transient stability control system for a distribution network with a high proportion of distributed energy resources, applied to the transient stability control method for a distribution network with a high proportion of distributed energy resources as described in any one of claims 1-9, characterized in that, The system includes: The data acquisition module is used to acquire real-time monitoring data of the power distribution network, distributed energy output information, and external environmental data. The real-time monitoring data includes voltage parameters, current parameters, and frequency parameters. The data fusion processing module is used to perform data fusion processing on the real-time monitoring data, distributed energy output information and external environmental data to generate comprehensive monitoring data. The prediction and analysis module is used to perform transient stability prediction and analysis on the comprehensive monitoring data and output the prediction results; The dynamic strategy generation module is used to make collaborative decisions based on the prediction results and the comprehensive monitoring data to generate dynamic adaptive adjustment strategies for the collaborative control of distributed energy devices, energy storage devices and controllable loads. The strategy execution and feedback control module is used to execute the dynamic adaptive adjustment strategy, collect feedback data after execution, and use the feedback data to perform secondary adjustment so that the voltage parameter or frequency parameter approaches the preset reference value.

Images