A Resilient Operation Control System for Distribution Networks Based on Distributed Power Generation

Abstract

This application relates to the field of distribution network control technology and discloses a distribution network resilient operation control system based on distributed generation. The system includes a distributed generation real-time status monitoring module, a distribution network resilient status assessment module, and a resilient operation control decision module. The distributed generation real-time status monitoring module collects the operating status data of all connected distributed generation sources in the distribution network, processes it, and generates a distributed generation operating status dataset. The distribution network resilient status assessment module calculates the overall resilient status of the distribution network based on this dataset and classifies it into levels, generating distribution network resilient status levels and associated influence parameter sets. The resilient operation control decision module generates operation control command sets for different distributed generation sources based on the levels and parameter sets, and performs dynamic operation regulation of the distribution network. This system can comprehensively grasp the operating status of distributed generation sources, accurately assess the resilient level of the distribution network, and achieve flexible and efficient regulation of the distribution network.

Description

Technical Field

[0001] This application relates to the field of power distribution network control technology, and in particular to a power distribution network resilient operation control system based on distributed generation. Background Technology

[0002] With the rapid development of renewable energy technologies, the penetration rate of distributed power sources in distribution networks continues to increase. While this change optimizes the energy structure and reduces carbon emissions, it also brings new challenges to the stable operation of distribution networks. Distributed power sources, such as photovoltaic power generation and wind power generation, are significantly affected by natural conditions, exhibiting strong intermittency and volatility. When a large number of distributed power sources are connected to the distribution network, it can easily lead to problems such as voltage fluctuations and frequency deviations, and even cause instability in local power grids.

[0003] Traditional power distribution network operation and control systems are mostly based on centralized management models, relying on preset operating schemes and fixed control strategies, making it difficult to respond in real time to the dynamic changes of distributed power sources. In the event of emergencies such as extreme weather or equipment failure, the power distribution network's anti-interference and rapid recovery capabilities are insufficient, often resulting in large-scale power outages that affect users' normal electricity consumption and socio-economic activities.

[0004] Existing distribution networks lack comprehensive condition monitoring of distributed generation sources, and their data processing methods are relatively simplistic, making it difficult to generate effective datasets to support resilient condition assessments. Furthermore, the lack of effective resilient condition assessment methods makes it impossible to accurately determine the resilience level of the distribution network under different operating conditions, thus affecting the scientific rigor and timeliness of control decisions. These problems hinder the efficient and flexible operation and control of distribution networks when dealing with the complex operating conditions brought about by the integration of distributed generation sources, thus restricting the further promotion and application of distributed energy resources. Summary of the Invention

[0005] To address the aforementioned technical issues, this application proposes a flexible operation control system for distribution networks based on distributed power sources.

[0006] The technical solution adopted in this application is: a flexible operation control system for distribution networks based on distributed power sources, comprising:

[0007] The distributed power generation real-time status monitoring module is used to collect the operating status data of all connected distributed power sources in the distribution network, and to standardize the operating status data to generate a distributed power generation operating status dataset.

[0008] The distribution network resilience assessment module is used to calculate and classify the overall resilience of the distribution network based on the distributed generation operation status dataset, and generate the distribution network resilience status level and associated impact parameter set.

[0009] The flexible operation control decision module is used to generate operation control command sets for different distributed power sources based on the flexible state level of the distribution network and the associated set of influence parameters, and to dynamically regulate the operation of the distribution network according to the operation control command sets.

[0010] Furthermore, the distributed power real-time status monitoring module identifies the types of abnormal states in the acquired distributed power operating status data.

[0011] The identified abnormal state types are encoded with state features to obtain the corresponding state feature identifier codes. The state feature identifier codes are then compared with the known sample state types recorded in the state feature identifier code mapping table.

[0012] When there is no sample state type that matches the abnormal state type in the state feature identifier mapping table, the state feature identifier is initialized to a specific value.

[0013] When a sample state type that matches the abnormal state type exists in the state feature identifier mapping table, the state feature identifier corresponding to the matching sample state type is associated with the abnormal state type.

[0014] Furthermore, the distributed power real-time status monitoring module locates the region to which the associated distributed power belongs based on the status feature identifier code that is not a specific value, and updates the abnormal status count corresponding to the region.

[0015] Extract the region identifier, topology location information, status feature identifier code, and time of occurrence of the abnormal state distributed power source, and sort and combine them to generate a region operation status sequence;

[0016] Based on the regional identifier in the regional operation status sequence, the regional operation status map in the distribution network resilience assessment module is updated. The regional operation status map is a graphical data model used for distribution network resilience analysis. It includes the physical connection topology of the distribution network and overlays a real-time operation status layer. The map nodes are divided into two categories: bottom-level nodes map to actual electrical nodes, which represent real-time collected data; upper-level nodes represent logical regions, which represent data aggregated from the bottom-level node data within the region. Node attributes include the number of normal and abnormal distributed power sources currently connected to the node, load level, and voltage status. The edges in the map are used to represent connection relationships; among them, the edge representing the line has attributes including load rate, and the edge representing the switch has attributes including on / off status.

[0017] Furthermore, the distribution network resilience assessment module triggers an assessment command when the regional operation status map is updated;

[0018] Mark the updated region as the target assessment region;

[0019] The total number of distributed power sources connected to the target assessment area and the number of distributed power sources in abnormal condition are statistically analyzed.

[0020] Data is integrated using regional resilience assessment calculation rules to output the resilience status identifier corresponding to the target assessment area.

[0021] Furthermore, the distribution network resilience assessment module includes a resilience level mapping module, which determines the resilience level of the target assessment area based on the value of the resilience status identifier.

[0022] When the elasticity status is marked with a specific value, the elasticity status of the target assessment area is determined to be normal.

[0023] When the elasticity status identifier is a non-specific value, a deep assessment instruction is triggered and secondary data integration is performed through the elasticity impact degree calculation rules to output the elasticity impact level identifier corresponding to the target assessment area. The elasticity impact level identifier includes the first level and the second level.

[0024] Furthermore, the elasticity state level mapping module marks the target assessment area corresponding to the elasticity impact level identifier with a value of first level as a low elasticity impact area;

[0025] The target assessment area corresponding to the elasticity impact level identifier with a value of level two is marked as a high elasticity impact area;

[0026] The flexible operation control decision module generates corresponding operation control instruction sets based on the markings of low-elasticity or high-elasticity influence areas.

[0027] Furthermore, the distribution network resilience assessment module obtains the status feature identification codes of all non-specific values;

[0028] Statistically analyze the frequency of occurrence of different state characteristic identifier codes and their total numerical value;

[0029] The type influence value corresponding to different abnormal state types is calculated using the state type influence analysis rules.

[0030] Furthermore, the flexible operation control decision module maintains the original operation control strategy for abnormal state types whose type influence value is a specific state value, and generates operation control strategy optimization instructions for abnormal state types whose type influence value is not a specific state value, wherein the specific state value is a type influence threshold preset within the system.

[0031] Furthermore, the flexible operation control decision module includes a control command verification unit. The control command verification unit performs a power distribution network operation status simulation based on the operation control command set and verifies the effectiveness of the operation control command set based on the simulation results.

[0032] When the verification fails, an instruction adjustment signal is generated and fed back to the elastic operation control decision module for iterative operation control instruction set iteration.

[0033] Furthermore, the system also includes:

[0034] The access control module is used to configure the operation permission levels of different users for the distributed power real-time status monitoring module, the distribution network resilient status assessment module, and the resilient operation control decision module.

[0035] The system operation monitoring module is used to monitor the operation status data stream of the distributed power real-time status monitoring module, the distribution network resilience status assessment module, and the resilience operation control decision module in real time.

[0036] The advantages of this application over the prior art are as follows:

[0037] By collecting and standardizing the operational status data of all distributed generation sources connected to the distribution network through a distributed generation real-time status monitoring module, the generated distributed generation operational status dataset comprehensively and accurately reflects the real-time operation of distributed generation sources, providing reliable basic information for subsequent distribution network resilience assessment. This comprehensive data collection and processing method avoids assessment biases caused by missing data or inconsistent formats, making the judgment of distribution network resilience more objective.

[0038] The distribution network resilience assessment module calculates and classifies the overall resilience status based on the distributed generation operating status dataset. The generated resilience status level and associated set of influencing parameters clearly present the current resilience level of the distribution network and the key factors affecting resilience. This process breaks through the traditional ambiguity in the understanding of resilience status in distribution network operation, allowing operation and management personnel to intuitively understand the resilience performance of the distribution network under different operating conditions, identify potential factors that may affect the stable operation of the distribution network, and thus carry out subsequent control work in a more targeted manner.

[0039] The flexible operation control decision module generates a set of control commands based on the distribution network's flexible state level and associated influencing parameter set, and dynamically regulates the distribution network accordingly, achieving refined management of distributed power sources. This dynamic regulation method can adjust the control strategy in real time according to changes in the distribution network's flexible state, adapting to the intermittent and fluctuating characteristics of distributed power sources, enabling the distribution network to maintain a relatively stable operating state under various complex conditions. Simultaneously, the differentiated control commands for different distributed power sources avoid the resource waste or insufficient regulation problems caused by a one-size-fits-all approach, improving the economy and reliability of distribution network operation. Attached Figure Description

[0040] The following description, in conjunction with the accompanying drawings, further illustrates this application:

[0041] Figure 1 A timing diagram of a distribution network resilient operation control system based on distributed power sources provided in an embodiment of this application;

[0042] Figure 2 A flowchart for identifying and encoding abnormal state types;

[0043] Figure 3 A flowchart for generating regional operational status sequences and updating the map;

[0044] Figure 4 A flowchart for determining the level and depth of elasticity assessment;

[0045] Figure 5 This is a flowchart for the impact analysis of abnormal state types. Detailed Implementation

[0046] like Figures 1 to 5 As shown, this application provides a distribution network resilient operation control system based on distributed generation, which mainly includes: a distributed generation real-time status monitoring module, a distribution network resilient status assessment module, and a resilient operation control decision module.

[0047] The distributed generation real-time status monitoring module is responsible for collecting operational status data of all connected distributed generation sources in the distribution network, including voltage, current, power output, equipment temperature, and grid connection point status information. This module standardizes the collected raw data, eliminating dimensional differences, unifying the data format, and generating a distributed generation operational status dataset. The distribution network resilience assessment module receives this dataset and calculates the overall resilience of the distribution network. The calculation process is based on a predefined resilience index algorithm, integrating the available capacity, response rate, and network topology of distributed generation sources to output a quantified resilience value. This quantified value is divided into different levels according to threshold ranges, and a set of associated impact parameters is generated, including the affected area identifier, the degree of resilience degradation, and the estimated recovery time. The resilience operation control decision module generates an operation control command set based on the resilience status level and the impact parameter set. The operation control command set includes commands for power adjustment, start / stop control, and operation mode switching for different types of distributed generation sources. This module issues commands through the distribution network energy management system interface to achieve real-time control of distributed generation sources and dynamically adjust the distribution network operation status.

[0048] Example 1: See Figure 2 The distributed power real-time status monitoring module identifies abnormal status types in the operating status data of the distributed power source by acquiring the operating status data, and encodes the identified abnormal status types with status features to obtain the corresponding status feature identification codes.

[0049] When there is no sample state type that matches the abnormal state type in the state feature identifier mapping table, the state feature identifier is initialized to a specific value.

[0050] When a sample state type that matches the abnormal state type exists in the state feature identifier mapping table, the state feature identifier corresponding to the matching sample state type is associated with the abnormal state type.

[0051] In the actual operation environment of a power distribution network, the implementation of a distributed generation real-time status monitoring module is a continuous and dynamic process. This module collects distributed generation operating status data at high frequency through sensor clusters, smart meters, and communication units deployed at various distributed generation grid connection points and key nodes. This data includes, but is not limited to, DC-side voltage, AC output current, active and reactive power, frequency offset, and inverter internal radiator temperature of photovoltaic inverters, as well as wind turbine speed, pitch angle, generator winding temperature, and converter status signals. All data is transmitted to the module's built-in data concentrator via power line carrier communication, a private wireless network, or a fiber optic network for preliminary data cleaning and timestamp alignment.

[0052] Because the data formats, sampling periods, and units of the devices come from diverse sources, the module's built-in processor performs a series of conversion operations. For example, it converts all voltage values ​​to per-unit values, using the system's rated voltage as a reference; current values ​​to amperes; power values ​​to megawatts; temperature values ​​to degrees Celsius; and time information to Coordinated Universal Time (UTC) format. This process generates a well-structured and time-consistent distributed power supply operating status dataset, providing standardized input for status identification.

[0053] The identification process is not a simple one-to-one threshold comparison, but rather employs a multi-parameter correlation rule engine. This engine incorporates rich judgment logic; for example, not all cases exceeding 105% of the rated voltage are immediately classified as "overvoltage anomalies." The system simultaneously checks whether the current at that node has increased abnormally, whether the reactive power compensation equipment in that area is operating, and whether the voltage at adjacent nodes has risen simultaneously. This correlation analysis effectively distinguishes between global voltage fluctuations and local equipment failures. Identified abnormal state types are assigned clear classification labels, such as "persistent overvoltage," "instantaneous frequency exceeding limits," and "communication interruption accompanied by zero power." Each identified type corresponds to a different physical meaning and potential impact. These textual descriptions of abnormal state types are then converted into standard codes that can be efficiently processed and compared by machines as state feature encodings. The encoding rules adopt a hierarchical structure; for example, the code "OV-01-SEV2" might represent the first seed type of overvoltage anomaly with a severity level of two. This encoding not only compresses data volume but also establishes a standardized description system for abnormal state types. The encoded state feature identifiers are easy to store, transmit, and quickly retrieve.

[0054] The status feature identifier mapping table is a dynamically updated knowledge base stored in the system's non-volatile memory. This table records all historically identified and processed abnormal status type samples, along with their unique identifiers. When a new abnormal status is identified and encoded, the system automatically searches the mapping table for a match. This matching process is not a simple string comparison but rather compares the core characteristic parameters of the abnormality, such as voltage deviation percentage, duration, and topological features of the location. If no sample record in the mapping table highly matches the current abnormal status feature identifier, it means that this may be a new abnormal status type that has never been encountered before or has significantly different characteristics. The system will mark the generated status feature identifier as the initial state, usually using a reserved specific value (such as "0" or "NULL"). This marking indicates that this abnormal status type requires special attention and further analysis from subsequent modules; its encoding will be temporarily stored, awaiting verification with more data.

[0055] Conversely, if a highly matching sample record is found in the mapping table, the system will perform an association operation. It associates the current anomaly instance with an existing sample state type in the mapping table and directly uses the verified state feature identifier code corresponding to that sample state type. For example, if an overvoltage anomaly is identified and its characteristic parameters highly match those of sample record number 15 in the mapping table, the system will mark the current anomaly instance as the known "OV-15" type and use its corresponding complete identifier code. This process ensures the consistency of the system's response to similar anomalies, avoids assigning different codes to the same essential problem, and ensures that subsequent statistical analysis and control decisions are based on stable and reliable data classification.

[0056] Example 2: See Figure 3 During the operation of the distribution network, the distributed generation real-time status monitoring module continuously generates a large amount of data streams with status characteristic identification codes. Once the module identifies anomalies and generates identification codes, the key to subsequent processing lies in associating these discrete anomaly events with the actual physical structure of the distribution network and transforming them into regionalized status information usable by advanced application modules. This process begins with the filtering of identification codes. The system data processing unit filters out all status characteristic identification codes that are not specific numerical values; these codes represent confirmed anomaly instances that require further processing.

[0057] Each valid status identifier is associated with a specific distributed power supply device. The system retrieves detailed access information for this device by querying the distributed power supply registration database. This database records the unique code of each distributed power supply, the name of the substation it is connected to, the feeder number, the management unit of the distribution area it belongs to, and the higher-level power supply area division. Based on this, the system can accurately locate abnormal devices to their respective electrical areas. This area is a logical concept, which may correspond to the range supplied by a distribution feeder or a power supply grid divided by tie switches. After location is completed, the system updates the abnormal status count value corresponding to that area. This count value is a basic indicator of the area's health status; an increase in its value indicates that operational disturbances are accumulating within the area.

[0058] Simultaneously, the system extracts in-depth information related to the anomaly from multiple data sources. This includes the regional identifier of the distributed power device, a unique geographical or logical partition code; its topological location information, detailing the device's specific location within the distribution network structure, such as which feeder and node it connects to, and its connections to adjacent switches, loads, and other power sources; a complete status characteristic identifier code, precisely describing the type and nature of the anomaly; and a timestamp recording the exact moment the anomaly occurred. This data is considered a complete information package.

[0059] To perform efficient time series analysis and regional comparison, the system does not process these information packets individually. Instead, it sorts and combines them according to predefined rules. The primary criterion for sorting is the region identifier, grouping all anomalous events occurring within the same region together. Secondly, within each region, the events are arranged chronologically according to their occurrence. This sorting and combination generates a structured sequence of regional operational states. Each sequence corresponds to a region, and each record in the sequence details a specific anomalous event that occurred at a specific time and on a specific device. This sequence dynamically reflects the evolution of the region's operational state, serving as a continuously updated data queue.

[0060] The generation of the regional operation status sequence triggers a deeper data fusion process. Based on the regional identifier in the sequence, the system updates the detailed information contained in the sequence to the regional operation status map in the distribution network resilience assessment module. This map is a graphical data model specifically designed for distribution network resilience analysis. It not only includes the physical connection topology of the distribution network but also overlays a real-time operation status layer. Map nodes are divided into two categories: bottom-level nodes map to actual electrical nodes (such as access points), representing real-time collected data; upper-level nodes represent logical regions, representing data aggregated from bottom-level node data within their jurisdiction. Node attributes include the number of currently connected normal and abnormal distributed power supply devices, load level, voltage status, etc. Edges in the map represent connection relationships; among them, edges representing lines have attributes including load rate, and edges representing switches have attributes including on / off status. Specifically, after the regional operation status sequence is generated, the system accurately locates the corresponding upper-level logical region node in the regional operation status map based on the clearly defined regional identifier in the sequence. The system traverses each abnormal state record within the sequence, and based on the detailed topological location information in the record, associates it with specific electrical nodes (such as the access points of distributed power sources) in the underlying structure of the graph. It then directly updates the real-time operating status attributes of these underlying nodes, including setting abnormal state flags and refreshing their real-time collected data such as voltage and power. Based on this, the system performs data aggregation, accumulating the number of distributed power sources in abnormal states among all underlying nodes under the jurisdiction of the upper-level logical region node. Simultaneously, it calculates the total number of normal distributed power sources in the region, the current total load level, and the region's average voltage state based on the integrated data from the underlying nodes, thereby updating the comprehensive attribute set of the upper-level region node. Meanwhile, the system processes the edges representing connection relationships in the graph in parallel: for edges representing distribution lines, their load rate attributes are updated based on real-time monitored power flow data; for edges representing switching equipment, their on / off status attributes are refreshed based on switch position signals received from the distribution automation system. This series of operations enables the deep integration of discrete regional operating state sequence information into a structured regional operating state map, completing the fusion of the real-time operating state layer and the static physical topology layer, and providing an immediate, accurate data model foundation with clear topological correlation for the distribution network resilient state assessment module.

[0061] When a new sequence of regional operating states arrives, the graph update algorithm locates the corresponding regional node and modifies its attributes. For example, it increases the count of abnormal devices in the region, adds new event records to the node's "abnormal event list," and may adjust the node's "status flag" based on the type of abnormality (e.g., changing from "normal" to "warning"). This process imbues the originally abstract distribution network topology with real-time, dynamic operating state semantics, transforming it into a truly meaningful "state" graph.

[0062] The distribution network resilience assessment module monitors changes in the regional operational status map. Any update to the attributes of a map node automatically triggers an assessment command within this module. The system marks the updated area as the target assessment area for this round of assessment. After the assessment command is initiated, the assessment algorithm first performs data statistics: it obtains the total number of distributed power generation devices connected to the target assessment area from the updated map, a relatively static value derived from power grid planning data; simultaneously, it obtains the number of distributed power generation devices currently in an abnormal state within the area, a dynamic value directly derived from the recently updated map node attributes.

[0063] The system invokes regional resilience assessment calculation rules to integrate and analyze the aforementioned data. This rule is not a simple proportional calculation, but rather an assessment function that comprehensively considers multiple factors. Its inputs include not only the ratio of the number of anomalous devices to the total number of devices, but may also incorporate the type weight of the anomalous devices (e.g., the impact of a failure in a megawatt-level photovoltaic power station differs from that of a failure in a kilowatt-level rooftop photovoltaic system), the average duration of the anomaly, and the importance coefficient of the region within the distribution network structure. Through calculation using these rules, the system outputs a quantitative result: the resilience status indicator corresponding to the target assessment region. This resilience status indicator is a comprehensive index; its value directly reflects the region's ability to withstand disturbances and maintain normal power supply at the current moment, providing the most direct numerical basis for subsequent classification and control decisions.

[0064] Example 3: See Figure 4 The distribution network resilience assessment module has a built-in resilience level mapping module, which is responsible for converting the calculated resilience status identifier values ​​into resilience status levels with clear operational guidance. This module maintains a resilience level classification table, which defines the levels corresponding to resilience status identifiers in different value ranges. For example, when the resilience status identifier value falls within the range [0, 0.2), the resilience status of the target assessment area is determined to be normal; when the value falls within the range [0.2, 0.5), it is determined to be a warning level; and when the value falls within the range [0.5, 1.0], it is determined to be a fault level. The core of the judgment logic lies in checking whether the resilience status identifier value is equal to a specific value, which is defined as "0" representing an absolutely normal state during system initialization. When the resilience status identifier equals this specific value, it means that there are no abnormal states within the target assessment area, or the impact of all abnormal states is negligible after calculation. Therefore, the module determines that the resilience status of the area is normal and ends the current assessment cycle for that area.

[0065] When the value of the resilience status indicator is not equal to the specific value, it indicates that there is a clear decline in resilience in the target assessment area, and the module immediately triggers a deep assessment command. The deep assessment process calls the resilience impact degree calculation rule, which performs secondary integration and analysis on the data in the regional operation status map. The factors considered in the secondary integration are more refined and comprehensive than those in the preliminary assessment, and its input data includes, but is not limited to: the distribution of current abnormal status types (i.e., the statistical results of different status characteristic identifier codes), the duration of various abnormalities, the proportion of the total installed capacity of abnormal equipment to the total capacity of regional distributed power sources, the current load level of the region, and the connection status of the region with the main grid (whether it has the conditions for islanded operation), etc.

[0066] The rule for calculating the degree of resilience impact uses a comprehensive weighted algorithm to process these multi-dimensional data, ultimately outputting a more refined quantitative indicator, namely the resilience impact level identifier. . The calculation method is as follows:

[0067] ;

[0068] in: The calculated elasticity impact level indicator is a continuous value between 0 and 1. This represents the total rated capacity of distributed power sources in the region that are in an abnormal state. This represents the total rated capacity of all distributed power sources within the region. Indicates the number of current abnormal status types within the region; Indicates the first The average duration of each abnormal state type from its occurrence to the current moment; Indicates the first The pre-defined impact weighting coefficients for each type of abnormal state are determined based on historical data analysis and expert knowledge, and are used to distinguish the potential threat level of different types of abnormalities to system stability. This indicates the maximum allowed duration threshold for anomalies in the system; The self-consistent operation capability coefficient of the region is an index calculated based on the region's internal power and load balance capability, energy storage configuration, and the completeness of islanding control strategy. The value ranges from 0 to 1, with 1 representing complete self-consistency capability. , , These are the capacity proportion factor, the time influence factor, and the self-consistency factor, which are three weighted coefficients that satisfy... The specific values ​​are determined through system debugging and simulation, and are used to adjust the contribution ratio of each factor in the overall evaluation.

[0069] The elastic state level mapping module receives the calculated The value is then used to determine its level based on another set of level mapping tables. This level mapping table will... The numerical range is divided into two main levels of influence. For example, when When the value falls within the range of [0.2, 0.6), its corresponding elasticity impact level is determined to be Level 1; when When the value falls within the range of [0.6, 1.0], it is classified as Level 2. The module then marks the target assessment area corresponding to the Level 1 elasticity impact indicator as a low elasticity impact area. This marking indicates that although there is a decrease in elasticity in this area, the degree is relatively mild, and the overall system operational stability is not fundamentally threatened. It may manifest as slight local voltage fluctuations or a small power deficit, but it still possesses a strong ability to self-regulate and maintain operation. Conversely, the module marks the target assessment area corresponding to the Level 2 elasticity impact indicator as a high elasticity impact area. This marking indicates that this area is experiencing severe elasticity decay, and the system operational stability faces a significant risk. It may manifest as severe voltage exceedances, frequency instability, or a large power deficit, requiring external intervention to prevent the situation from worsening.

[0070] The flexible operation control decision module monitors the marking status of each area in real time. Once it obtains the marking information of a low-elasticity impact area or a high-elasticity impact area, it immediately initiates the corresponding control command generation process. For targets marked as low-elasticity impact areas, the flexible operation control decision module retrieves the pre-set mild adjustment strategies in its strategy library. The generated operation control command set may include: adjusting the reactive power output of normally operating distributed power sources in the area to support voltage, slightly increasing or decreasing their active power to balance local power, switching reactive power compensation devices in the area, or requesting slight power support from adjacent areas. These commands aim to suppress the trend of decreased elasticity through fine-tuning, bringing the area's operating status back to the normal range.

[0071] For targets marked as high-resilience impact areas, the resilient operation control decision module will take more decisive and powerful control measures. The generated operation control command set will be more complex and comprehensive, potentially including: emergency start / stop of backup distributed generator units or energy storage systems in the area; execution of load shedding operations for non-critical loads; alteration of network topology (such as closing tie switches to transfer loads); and even initiation of islanded operation mode when conditions permit, disconnecting the area from the main grid and allowing it to operate independently using internal power sources. The command set generation process will fully consider the coupling relationships and execution timing between various control actions to avoid generating new instabilities. All generated command sets are ultimately converted into specific commands that can be recognized by lower-level execution units and distributed to corresponding circuit breakers, generator controllers, energy storage converters, and other equipment via the communication network for execution, thereby achieving dynamic regulation of the distribution network operation and responding to resilience degradation events.

[0072] Example 4: See Figure 5 The distribution network resilience assessment module continuously receives data streams from the distributed generation real-time status monitoring module, which contain a large number of abnormal state instances marked by status feature identifiers. These identifiers are standardized codes that accurately describe the type and nature of the anomalies. The module's processing logic is not limited to responding to individual events, but rather aims to identify the distribution patterns of anomaly types and their potential impact patterns at a macro level, thereby providing directional guidance for optimizing control strategies.

[0073] This module first filters all incoming status feature identifiers, eliminating codes with specific numerical values ​​(these codes represent unknown or initial anomaly types that require further observation). It focuses on processing identifiers already identified and categorized into known types by the system (known type identifiers exist in numerical code form, used to explicitly identify various anomaly types recognized by the system). Subsequently, the system initiates a statistical aggregation process. This process uses each unique status feature identifier (i.e., a specific identifier) ​​as a key, performing two statistical calculations: first, counting the frequency of the specific identifier within a set statistical time window; and second, calculating the sum of the values ​​of these occurring specific identifiers. It is important to emphasize that the "sum of values" here does not refer to the character value of the identifier itself, but rather to an internal weight value associated with each identifier in the system. This weight value is assigned when the identifier is defined and is used to initially quantify the severity of the anomaly type. For example, the code "FREQ-001" might be associated with a weight value of 5, while the code "VOLT-003" might be associated with a weight value of 8.

[0074] After statistical analysis, the system obtains a basic dataset for each known abnormal state type, including its occurrence frequency and the cumulative sum of its weights. Then, it invokes the state type impact analysis rule to deeply integrate and analyze this data. This rule is a multi-factor decision-making logic; its analysis process not only relies on simple frequency and weight accumulation but also introduces dynamic parameters such as time decay factors and spatial distribution factors. The time decay factor assigns higher weight to recently occurring abnormal events, believing they better reflect the current true state of the system; the spatial distribution factor assesses whether this type of anomaly is concentrated in a vulnerable area or widely dispersed throughout the network, with the former typically indicating higher local risk. By weighting these dimensions, the rule outputs a final quantified type impact value for each abnormal state type. This value is a comprehensive indicator reflecting the overall impact of this type of anomaly on the overall resilience of the distribution network within a specific time period.

[0075] The system internally sets a type impact threshold, a pre-calculated and predefined baseline value. This threshold categorizes abnormal state types into two processing categories. The resilient operation control decision module receives a real-time data stream of type impact values ​​from the distribution network resilient state assessment module. For abnormal state types with type impact values ​​equal to or lower than a specific state value, the resilient operation control decision module determines that the current state is within an acceptable range or has been effectively suppressed by existing control strategies. The specific state value refers to the aforementioned type impact threshold. For these types, the resilient operation control decision module maintains its original operation control strategy and does not initiate any strategy change instructions. Existing control logic, which may be strategies previously optimized for this type of abnormality, is considered sufficient and effective and continues to be executed.

[0076] Conversely, for abnormal state types whose type impact value is not a specific state value (i.e., higher than the type impact threshold), the elastic operation control decision module determines that the impact of this type of abnormality is expanding or that the existing control strategy is ineffective. The elastic operation control module then generates an operation control strategy optimization instruction. This instruction is a trigger signal; it does not directly contain specific strategy content but instead specifies the target abnormality type that needs optimization and its current estimated impact level. This instruction is sent to the strategy optimization engine.

[0077] The strategy optimization engine initiates a strategy reconstruction process based on instructions. This process first queries the historical database, retrieving all historical event records, executed control commands, and their effect evaluation reports related to the anomaly type. Then, the engine may use rule-based reasoning or case matching methods to generate one or more alternative optimization strategy schemes. These schemes may include: adjusting the response threshold for this type of anomaly, modifying control action parameters (such as increasing power regulation or shortening response delay), introducing new cooperative control objects (such as linking energy storage systems or reactive power compensation devices), or, in extreme cases, triggering higher-level system protection logic. After simulation and effectiveness verification, the generated optimization strategy schemes are updated in the strategy library, replacing or supplementing the original strategies, thereby completing the iterative optimization of the operational control strategy.

[0078] Table 1: Impact Analysis of Abnormal State Types.

[0079]

[0080] Referring to Table 1, the table shows the system's analysis results for five different abnormal state types. For example, the overvoltage type anomaly with the identifier "VOLT-003" occurred 27 times, with a cumulative weight of 216. Its high time decay coefficient indicates frequent recent occurrences, and its extremely high spatial concentration index indicates the anomaly is concentrated in a specific area. The calculated type impact value far exceeds the threshold, therefore the system's recommended approach is "optimization strategy." Conversely, although the "FREQ-001" type anomaly has a higher frequency, its spatial distribution is relatively dispersed, and its overall impact value is below the threshold; therefore, a "maintenance strategy" is recommended. This data-driven analysis process makes the adjustment of control strategies clear and targeted.

[0081] Example 5: The flexible operation control decision module integrates a control command verification unit, which plays a crucial role in verifying the safety and effectiveness of commands before they are actually issued and executed. The unit's workflow begins with receiving a set of operation control commands generated by the decision logic. These commands typically contain a series of specific control commands, such as adjusting the active / reactive power output setpoints of a specific distributed power source, opening or closing certain circuit breakers, switching the charging and discharging modes of the energy storage system, or initiating an islanded operation sequence. The command set also includes the expected time window and sequence for executing these commands.

[0082] The control command verification unit can initiate the distribution network operation status simulation process. This simulation is not a simple logic check, but rather a simulation of the distribution network operation status based on a highly realistic real-time digital twin model of the distribution network. This real-time digital twin model is a digital model that can synchronize the actual physical structure, equipment parameters, and operating status of the distribution network in real time. This model continuously receives and synchronizes data such as the actual distribution network topology, equipment parameters, and current operating points (node ​​voltages, line power flow, load levels, and distributed generation output status) to construct a virtual mirror consistent with the actual distribution network operation status. This provides a simulation environment highly consistent with the actual distribution network for subsequent simulation of control command execution effects. The simulation engine then injects the set of operation control commands to be verified into this model, simulating the dynamic response process of these commands after execution in the real distribution network environment. The simulation engine is the core functional unit in the real-time digital twin model of the distribution network used to simulate the execution process of control commands. Its role is to input the set of operation control commands to be verified (such as commands to adjust the output of distributed power sources and switch status) into the digital twin model, and simulate the changes in key parameters such as node voltage, line current, and system frequency when these commands are executed in the real distribution network, based on the physical operation laws of the distribution network (such as power flow calculation rules and equipment start-up and shutdown logic). At the same time, it follows the dynamic characteristics of the equipment (such as generator start-up delay and energy storage charging and discharging rate) to ensure that the simulation results can truly reflect the actual operating state of the distribution network after the command is executed, and provide a reliable basis for verifying the effectiveness of the control commands.

[0083] The simulation process covers the entire cycle of command execution, from the issuance of the first command to the completion of all commands and the system reaching a new steady or quasi-steady state. During the simulation, the engine strictly adheres to physical laws and the dynamic characteristic models of the equipment. For example, when simulating a command to "increase the active power output of the photovoltaic inverter by 10%", the engine calculates the impact of this operation on the local node voltage and then extrapolates the cascading effects on adjacent nodes and feeders through power flow calculations. When simulating a command to "close the tie switch", the engine calculates the resulting network topology changes, power flow redistribution, and potential circulating currents. When simulating a command to "start the diesel generator and connect it", the engine considers the generator's start-up delay, ramp rate, and its effect on system frequency regulation. The simulation process records the changes in all key parameters of the simulated system, including but not limited to node voltages, line currents and power, system frequency, load rates of critical equipment (such as transformers), and whether protection devices have activated.

[0084] Based on the simulation results, the control command verification unit performs effectiveness verification. The verification criteria are a predefined and configurable set of safety and performance constraints. Core constraints include: whether the voltage of all nodes remains within the allowable upper and lower limits after simulation; whether the load on all lines and equipment does not exceed their safe capacity; whether the system frequency fluctuates stably within a small range near its rated value; whether the initial resilience degradation that triggered this control decision has been eliminated (e.g., whether power deficits have been compensated, whether voltage overruns have been corrected); and whether new, more serious stability problems (such as voltage collapse, equipment overload, protection malfunctions, etc.) have been triggered. The control command verification unit compares and analyzes the simulation data against these constraints one by one.

[0085] When the simulation results meet all preset safety and performance constraints, the control command verification unit determines that the operation control command set has passed verification. Verified command sets are marked as valid and can be sent to field actuators (such as generator controllers, circuit breaker controllers, energy storage management systems, etc.) via the communication interface for actual execution. When the simulation results fail to meet one or more constraints, the control command verification unit determines that the verification has failed. At this time, the unit generates a detailed command adjustment signal. This signal not only contains a failure flag but also specific diagnostic information, indicating which constraints were not met and key problem points observed during the simulation (e.g., "node N12 voltage drops to 0.85 pu, below the lower limit," "feeder L7 load rate reaches 120%," "frequency drops to a minimum of 49.2 Hz"). The diagnostic information also relates to the specific command or combination of commands that caused the problem.

[0086] The instruction adjustment signal is fed back to the flexible operation control decision module in real time. Upon receiving the adjustment signal, the flexible operation control decision module initiates an iterative optimization process for the operation control instruction set. This iterative process may employ various strategies: modifying the instruction parameters that cause the problem (e.g., reducing the power increase from 10% to 5%); adjusting the execution order or timing of instructions (e.g., delaying the start-up time of a generator to avoid cumulative impact); replacing some instructions (e.g., replacing increasing the output of a photovoltaic power station with starting energy storage discharge); or even completely abandoning the current strategy and reselecting or generating an alternative control instruction set from the strategy library. The newly generated or modified instruction set is then sent back to the control instruction verification unit for a new round of simulation and verification. This iterative cycle of "generation-verification-adjustment" continues until an instruction set that passes all verification constraints is generated, or the preset maximum number of iterations is reached (at which point a higher-level alarm or manual intervention may be triggered).

[0087] The system architecture also includes a permission management module, which manages the access and operation permissions of different user roles for each functional module of the entire system. The permission hierarchy is clearly divided into three levels: read-only, operation, and management. Read-only users can only view the status information, monitoring data, and evaluation results of each system module; they have no right to modify any configurations or issue commands. Operation-level users have permission to perform routine operations, such as confirming alarm information, manually triggering status evaluations, viewing and confirming (but not modifying) the automatically generated set of operation control commands, and executing them after authorization. Management-level users have the highest permissions, allowing them to modify system configuration parameters, manage user accounts and permissions, and, in certain special circumstances, manually generate and issue control commands.

[0088] Access control is implemented through a strict user authentication mechanism, such as username and password combined with digital certificates. After a user logs in, their operation requests are checked by an access control list, which defines in detail the operation permissions (read, write, execute) of each role for each module, function menu, and even data item. All critical user operations (especially configuration modifications and command issuance) are recorded in detail in the operation log, including the operator, operation time, operation content, and operation result, meeting auditing requirements.

[0089] The system architecture also includes a system operation monitoring module, which acts as the system's "self-testing center," continuously monitoring the operational status of three core modules: the distributed power real-time status monitoring module, the distribution network resilience assessment module, and the resilience operation control decision module. It captures and analyzes the data flow between modules, monitoring key indicators including: whether the data acquisition and transmission rates are normal, whether there are abnormal delays in data processing, whether communication between modules is smooth, whether internal calculation tasks are completed on time, and whether error or alarm logs are generated. This module typically provides a comprehensive visual monitoring interface, intuitively displaying the real-time health status, performance indicators, and statistical information of key data flows for each module in the form of charts or dashboards. When any module is detected to be operating abnormally, this module generates corresponding alarm information, notifying maintenance personnel to intervene promptly and ensure the stable and reliable operation of the entire control system.

[0090] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A distributed power supply based power distribution network elastic operation control system, characterized in that: The application relates to a distributed power supply real-time state monitoring module for collecting operation state data of all distributed power supplies connected to a power distribution network and performing standardization processing on the operation state data to generate a distributed power supply operation state data set. The distributed power supply real-time state monitoring module identifies the abnormal state type in the operation state data through the obtained operation state data of the distributed power supply. The identified abnormal state type is subjected to state feature coding to obtain a corresponding state feature identification code, and the state feature identification code is compared with a known sample state type recorded in a state feature identification code mapping table. When there is no sample state type matching the abnormal state type in the state feature identification code mapping table, the state feature identification code is initialized as a specific numerical value. When there is a sample state type matching the abnormal state type in the state feature identification code mapping table, the state feature identification code corresponding to the matching sample state type is associated to the abnormal state type. The distributed power supply real-time state monitoring module locates the region to which the associated distributed power supply belongs according to the state feature identification code of the non-specific numerical value, and updates the abnormal state count corresponding to the region. The region identification, topological position information, state feature identification code and abnormal occurrence time of the abnormal state distributed power supply are extracted, sorted and combined to generate a region operation state sequence. According to the region identification in the region operation state sequence, the region operation state atlas in the power distribution network elasticity state evaluation module is updated. The region operation state atlas is a graphical data model for power distribution network elasticity analysis, contains the physical connection topology of the power distribution network, and is superimposed with a real-time operation state layer. The atlas nodes are divided into two categories: the bottom nodes map the actual electrical nodes, which represent the real-time collected data; the upper nodes represent the logical regions, which represent the data aggregated by the bottom nodes in the region. The node attributes include the number of normal and abnormal distributed power supplies currently connected to the node, the load level and the voltage state. The edges in the atlas are used to represent the connection relationship. The edges representing the lines have attributes including the load rate, and the edges representing the switches have attributes including the on-off state. The power distribution network elasticity state evaluation module is used for calculating and grading the overall elasticity state of the power distribution network according to the distributed power supply operation state data set, generating the power distribution network elasticity state grade and the associated influence parameter set. The elasticity operation control decision module is used for generating the operation control instruction set for different distributed power supplies based on the power distribution network elasticity state grade and the associated influence parameter set, and dynamically regulating and controlling the power distribution network according to the operation control instruction set. The power distribution network elasticity state evaluation module triggers the evaluation instruction when the region operation state atlas is updated. 2.The power distribution network resilience control system based on distributed power supply of claim 1, wherein: The updated region is marked as a target evaluation region. The total number of distributed power supplies connected to the target evaluation region and the number of distributed power supplies in abnormal states are counted. The data is integrated through the region elasticity evaluation calculation rule, and the elasticity state identification corresponding to the target evaluation region is output. ​ 3. The power distribution network resilience control system based on distributed power supply according to claim 2, characterized in that: The power distribution network resilience state evaluation module comprises a resilience state level mapping module, which determines the resilience state level corresponding to the target evaluation area according to the value of the resilience state identifier; When the resilience state identifier is a specific value, it is determined that the resilience state of the target evaluation area is normal; When the resilience state identifier is a non-specific value, a deep evaluation instruction is triggered and secondary data integration is performed through resilience impact degree calculation rules, and the resilience impact level identifier corresponding to the target evaluation area is output, which includes a first level and a second level.

4. The power distribution network resilience control system based on distributed power supply according to claim 3, characterized in that: The resilience state level mapping module marks the target evaluation area corresponding to the resilience impact level identifier with a value of the first level as a low resilience impact area; The target evaluation area corresponding to the resilience impact level identifier with a value of the second level is marked as a high resilience impact area; The resilience operation control decision module generates corresponding operation control instruction set according to the marking of the low resilience impact area or the high resilience impact area.

5. The power distribution network resilience control system based on distributed power supply according to claim 1, characterized in that: The power distribution network resilience state evaluation module obtains all state characteristic identifier codes with non-specific values; The occurrence frequency and the value sum of different state characteristic identifier codes are counted; The type impact value corresponding to different abnormal state types is calculated through the state type impact analysis rule.

6. The power distribution network resilience control system based on distributed power supply according to claim 5, characterized in that: The resilience operation control decision module maintains the original operation control strategy for the abnormal state type with a specific state value, and generates an operation control strategy optimization instruction for the abnormal state type with a non-specific state value, wherein the specific state value is a type impact threshold value preset in the system.

7. The power distribution network resilience control system based on distributed power supply according to claim 5, characterized in that: The resilience operation control decision module comprises a control instruction verification unit, which performs power distribution network operation state deduction according to the operation control instruction set, and verifies the effectiveness of the operation control instruction set according to the deduction result; When the verification fails, an instruction adjustment signal is generated and fed back to the resilience operation control decision module for operation control instruction set iteration.

8. The distributed power supply based power distribution network resilience operation control system according to any one of claims 1-7, characterized in that: The system further comprises: The permission management module is configured to configure the operation permission level of different users for the distributed power source real-time state monitoring module, the power distribution network resilience state evaluation module and the resilience operation control decision module; The system operation monitoring module is configured to monitor the running state data flow of the distributed power source real-time state monitoring module, the power distribution network resilience state evaluation module and the resilience operation control decision module.

Images