A regional autonomous hierarchical control method for active distribution network based on master station system

Abstract

The present application discloses a method for regional autonomous hierarchical control of an active distribution network based on a master station system, which relates to the field of power system dispatching, including: using a modeling method based on the Common Information Model (CIM) standard, and establishing a distributed power supply equipment model with the equipment model and equipment data of the distributed power supply that interacts with the external system; dividing multiple hierarchical control units according to the distribution network topology and the distributed power supply equipment model, constructing a logical relationship model of the hierarchical control unit, mapping the logical relationship model with the distributed power supply equipment model, and generating a hierarchical control unit model; through monitoring equipment, collecting the operating status data of each control unit in real time, inputting it into a pre-trained neural network evaluation model, and evaluating the operating status of each control unit; and collaboratively controlling each control unit based on the operating status evaluation results. In response to the problem of low coordinated control accuracy after the distributed power supply is connected to the distribution network in the prior art, the present application improves the control accuracy of the distribution network system.

Description

Technical Field

[0001] The present application relates to the field of power system dispatching, and in particular to a method for regional autonomous hierarchical control of active distribution networks based on a master station system. Background Art

[0002] Traditional distribution networks utilize a centralized control model, relying primarily on a master station system for overall dispatch and centralized control of the distribution network. This control model is suitable for passive distribution networks dominated by unidirectional power flows, but it struggles to cope with the complexity and uncertainty of distribution networks with high penetration rates of distributed power generation. Distributed power sources are intermittent and fluctuating, and their output is affected by natural conditions such as wind speed and sunlight, making them difficult to accurately predict and control. Furthermore, distributed power sources and loads are dispersed across various nodes in the distribution network, and their access points and power flow directions change at any time, resulting in a diverse and dynamic operating state for the distribution network. Traditional centralized control models struggle to acquire and process large amounts of distributed operating data in real time, making it impossible to precisely regulate the power flow and voltage of the distribution network, resulting in low control accuracy and efficiency.

[0003] To address the low accuracy of coordinated control of distribution networks after the integration of distributed power sources, academia and industry have conducted extensive research and practice. Some scholars have proposed a distributed collaborative control method based on a multi-user system, which achieves decentralized autonomous control of the distribution network through negotiation and interaction between autonomous users at each node of the distribution network. However, this method lacks consideration of the global optimization of the distribution network, making it difficult to ensure the coordination and optimality of the control behavior of each node. Other scholars have proposed a hierarchical control approach, which achieves hierarchical autonomy and collaborative control by setting up control units in various areas of the distribution network. However, most existing hierarchical control methods adopt a fixed partitioning approach and lack the ability to make adaptive adjustments based on the actual operating status of the distribution network. At the same time, existing methods lack specific autonomous control strategies within the control unit, making it difficult to adapt to complex source-load combinations and demand changes.

[0004] In related technologies, for example, Chinese patent document CN115483701A provides a regional autonomous hierarchical control method for active distribution networks based on a master station system. This method first models distributed power sources (DGs), enabling the DG device model and data to be updated from the source system to the DMS system. Based on the DG device model information and topology data within the system, three DG coordinated scheduling modes are defined, and the regional equipment and topology data required for monitoring and analysis for each DG coordinated scheduling mode are divided within the system. The operating conditions of distributed source and load devices within the autonomous region are monitored and analyzed in real time. When regional autonomous operation is detected to be disrupted, an adjustment strategy for the autonomous region is calculated based on various resources collected within the system, mobilizing various resources (active and reactive backup resources) to restore the regional operating conditions to the autonomous mode. However, the DG device modeling in this solution primarily relies on the source system's dynamic updates, resulting in the established device model failing to accurately reflect the actual operating status and regulation capabilities of the DGs, reducing the pertinence and effectiveness of coordinated control. Summary of the Invention

[0005] In response to the problem of low coordinated control accuracy after the access of distributed power sources to the distribution network in the prior art, the present application provides an active distribution network regional autonomous hierarchical control method based on the master station system, which improves the control accuracy of the distribution network system through autonomous system optimization and control of distributed power sources and loads in each area of ​​the distribution network.

[0006] Technical solution: The purpose of this application is achieved through the following technical solution.

[0007] This specification provides a method for regional autonomous hierarchical control of an active distribution network based on a master station system, including:

[0008] S1. Adopting a modeling method based on the Common Information Model (CIM) standard, the distributed power supply equipment model and equipment data are exchanged with the external system to establish a distributed power supply equipment model. S2. The master station system divides multiple hierarchical control units according to the distribution network topology and the established distributed power supply equipment model, constructs a logical relationship model of the hierarchical control unit, maps and integrates the logical relationship model with the distributed power supply equipment model, and generates a hierarchical control unit model that includes monitoring equipment, control equipment and internal logical structure. S3. The master station system collects the operating status data of each control unit in real time through the monitoring equipment in the hierarchical control unit model, inputs the data into the pre-trained neural network evaluation model, and evaluates the operating status of each control unit. S4. The master station system collaboratively controls each control unit based on the evaluation results of the operating status of each control unit.

[0009] Among them, the Common Information Model (CIM) standard: CIM is a set of information exchange standards developed by IEC, the international standardization organization for the power sector. It defines a common semantic model for various objects in the power system, including devices, network topologies, and measurement data, enabling data interoperability between different vendors and application systems. Using CIM standard modeling facilitates seamless integration of distribution automation systems with other systems. External systems refer to other systems that exchange data with the active distribution network master station system, such as distributed power generation (DG) monitoring systems, energy management systems, distribution automation master stations, and distribution network geographic information systems. The master station system obtains DG device models and real-time operating data from external systems through CIM standard interfaces. DG device models and device data: The device model describes the physical components and attributes of the DG, such as the rated capacity and control mode of photovoltaic inverters, wind turbines, and energy storage converters. Device data refers to real-time operating status information of DG devices, such as measured values ​​such as active and reactive power output and grid connection point voltage. The master station system obtains the DG device models and data from external systems and converts them into a unified format that complies with the CIM standard. A hierarchical control unit (HCU) is a regional unit with independent control capabilities, defined based on the distribution network topology and the location of distributed generation (DG) connections. It is generally categorized into three levels: microgrid, smart substation, and regional DG. Each HCU contains multiple physical devices, such as loads, DGs, and switchgear, and possesses a certain degree of autonomous control capabilities. The HCU's logical relationship model describes the logical control relationships between loads, DGs, switchgear, and other device objects within each HCU. This logical relationship model is constructed based on graph theory, with nodes representing device objects and edges representing control relationships. Logical relationships determine the coordinated control strategies for each device within the HCU. The HCU model is a comprehensive model that integrates the logical relationship model and the DG device model. This model includes mappings of the physical devices within the HCU, such as monitoring devices like smart meters and fault indicators, and control devices like circuit breakers and capacitors, as well as logical control relationships between these devices. The HCU model possesses both physical and logical attributes and serves as the core model for implementing HCU autonomous control. Operational status data: This refers to real-time operational measurement information from the hierarchical control unit, including load power, power output, line flow, node voltage, and other data. The master station system periodically collects operational status data from each unit through monitoring devices (such as smart meters) in the hierarchical control unit model as input for operational status assessment.

[0010] Furthermore, S1 establishes a distributed power supply equipment model, including: the master station system and external related systems define the public equipment model and proprietary equipment model of the distributed power supply based on the CIM standard; wherein the public equipment model includes static attribute information such as the type, capacity and geographical location coordinates of the distributed power supply, and the proprietary equipment model includes dynamic characteristic information such as the real-time dispatchable capacity, fault removal time and control mode of the distributed power supply; the master station system and external related systems establish an interactive interface between the distributed power supply equipment model and equipment data based on the CIM standard; wherein the equipment data includes the nameplate parameters, operating parameters, ledger information, etc. of the distributed power supply; the interactive interface It adopts an XML-based data format to support cross-system transmission of distributed power supply equipment models and equipment data; when distributed power supply equipment in external related systems is added, deleted or parameters are changed, the changed distributed power supply equipment model and equipment data are pushed to the master station system through the established interactive interface; after the master station system receives the distributed power supply equipment model and equipment data pushed by the external related system, it extracts the key parameters, updates the internally maintained distributed power supply equipment model and equipment database, and uses the updated distributed power supply equipment model and equipment data to support subsequent hierarchical control unit division, autonomous area operation status monitoring and evaluation, and adjustment control.

[0011] Among them, the master station system automatically generates a logical topology model of each hierarchical control unit based on the physical topology structure of the distribution network and the distributed power supply equipment model established in step S1; the logical topology model includes objects of monitoring equipment and control equipment and logical associations between devices; the master station system extracts real-time measurement point information of monitoring equipment and control equipment in each hierarchical control unit from the real-time database, and integrates the measurement point information with the logical topology model generated in step S21 to form a hierarchical control unit model containing real-time measurement points; the master station system traverses the hierarchical control unit model formed in step S22, and creates corresponding data acquisition tasks according to the type and sampling period of each monitoring equipment; creates corresponding control instruction issuance tasks according to the type and control mode of each control equipment; incorporates the created data acquisition tasks and control instruction issuance tasks into the task scheduling module of the master station system; the task scheduling module of the master station system periodically triggers the data acquisition tasks and control instruction issuance tasks, collects monitoring data of each hierarchical control unit through the distribution automation communication network, and issues control instructions to achieve closed-loop monitoring.

[0012] The common device model refers to a distributed power generation (DG) device model with common attributes that conforms to the CIM standard. This model includes basic DG characteristic information, such as: type (e.g., photovoltaic, wind power, or energy storage); capacity (e.g., the rated output of the DG, expressed in kW or MW); and geographic coordinates (e.g., the spatial location of the DG within the geographic information system, typically expressed in latitude and longitude coordinates). The master station system and external systems exchange information based on this unified common device model, ensuring data consistency and understandability.

[0013] Proprietary device model: This refers to a device model customized by the master system or external system based on the public device model and the specific attributes and control requirements of the distributed power source. The proprietary device model reflects the personalized information of the distributed power source, such as: Real-time dispatchable capacity: This indicates the active power output capacity of the distributed power source that can currently participate in distribution network dispatch, taking into account the power source's real-time output capability and grid connection restrictions, and is measured in kW or MW; Fault clearing time: This indicates the time it takes for the distributed power source to clear the fault through protective action after a fault occurs, and is measured in milliseconds or seconds; Control mode: This indicates the output control method of the distributed power source, such as constant power, reactive power priority, and voltage reactive power priority. Through the proprietary device model, the master system can obtain detailed information about the distributed power source, enabling more flexible and optimized regulation.

[0014] Master station system: This refers to the monitoring and control system at the master station level of distribution automation. It collects data and issues control commands to the various intelligent power devices in the distribution network through a communication network. The core functions of the master station system include: grid topology model management: maintaining the topology and asset status information of the distribution network; real-time data acquisition: periodically collecting telemetry and telesignals from each intelligent power device; monitoring and alarm processing: monitoring over-limit and fault alarms for collected real-time data; remote control execution: issuing remote control commands to optimize the operating status of the distribution network; and human-computer interactive display: presenting global information about the distribution network to operators via a graphical interface. In this solution, the master station system is responsible for exchanging device models and data of distributed power sources with external systems and applying them to the hierarchical autonomous control of the distribution network.

[0015] Equipment nameplate parameters: These are the inherent parameters determined at the time the equipment leaves the factory, reflecting the physical characteristics of the equipment. For distributed power sources, nameplate parameters typically include rated capacity, rated voltage, frequency, and power factor. Equipment operating parameters: These are the parameters that dynamically change during real-time operation and reflect the current operating conditions of the equipment. For distributed power sources, operating parameters typically include active output, reactive output, grid connection point voltage, and current.

[0016] Furthermore, S2 generates a hierarchical control unit model including monitoring equipment, control equipment and internal logical structure, including: S21, the master station system extracts the geographical location coordinates, load type and distributed power capacity of the equipment node as key features based on the established distributed power equipment model; based on the extracted key features, the DBSCAN clustering algorithm is used to group the equipment nodes, and the division results of the three types of hierarchical control units, namely microgrid, smart substation and regional distributed power supply, are output according to the clustering labels; S22, the master station system uses the object-oriented modeling method to construct the logical relationship model within each hierarchical control unit based on the hierarchical control unit division results; wherein, in the microgrid logical relationship model, the key load, general load and controllable load are defined. load, distributed power supply and energy storage device objects, and establish control associations between each object based on the control mode attributes; in the intelligent substation logical relationship model, define the switchgear, distribution transformer terminal and user smart meter objects, and establish a hierarchical structure between each object based on the topological relationship; in the regional distributed power supply logical relationship model, define the power generation control voltage and grid-connected control unit objects; S23, the master station system adopts the object association method to merge the logical relationship model objects with the device entity mapping of the distributed power supply device model, and generate a hierarchical control unit model containing the device composition topology and internal logic; extract the monitoring device and control device objects in the hierarchical control unit model, and establish data subscription and control instruction issuance channels respectively for collaborative control between control units.

[0017] Microgrids refer to small, localized power grids within distribution networks, consisting of distributed power sources (DGs), loads, energy storage devices, and control equipment. Microgrids are self-controlled, protected, and managed, and can operate both grid-connected and as isolated islands. In this solution, microgrids serve as the fundamental control unit for distribution automation. They optimize control of each device based on the real-time status of internal DGs and loads, achieving supply-demand balance and power quality management. Smart substations refer to local areas within distribution networks supplied by several distribution transformers. By installing smart distribution transformer terminals, smart meters, and switchgear within these substations, substation-level visibility and control are achieved. In this solution, smart substations serve as intermediate control units, collecting load data from each substation through distribution transformer terminals and smart meters, and controlling switchgear to optimize network losses and power reliability. Regional distributed power sources (RDGs) refer to large-scale distributed power sources (DGs) connected to the distribution network and distributed across various substations, such as centralized wind farms and photovoltaic power plants. Regional distributed power sources have larger capacity and higher dispatch priority. In this solution, regional distributed generation (RDGs) serve as the top-level control unit. They participate in distribution network voltage regulation by controlling the reactive power output of each DDG in each substation, and smooth out load fluctuations by rationally dispatching active power output. Object-oriented modeling is a commonly used software system analysis and design method. It uses objects as the basic unit, dividing the system into various objects and defining their attributes, behaviors, and relationships. In this solution, the master station system uses an object-oriented approach to abstract the logical relationship models for three types of control units: microgrids, smart substations, and regional distributed generation (RDGs). Each control unit contains multiple device objects and control logic. Critical loads, general loads, and controllable loads refer to different types of loads within the microgrid. Critical loads refer to important users with high requirements for power supply reliability and power quality, such as hospitals and transportation hubs; general loads refer to ordinary electrical equipment with relatively low requirements for power quality; and controllable loads refer to loads whose power consumption time and power can be adjusted through demand response measures, such as industrial equipment and electric vehicle charging stations. Differentiated control strategies are implemented for different load types within the microgrid's logical relationship model. Distributed power and energy storage device objects: Refers to distributed power generation equipment (such as photovoltaics and wind power) and energy storage equipment (such as batteries and supercapacitors) within the microgrid. Distributed power sources are the primary power source for microgrids, while energy storage devices can smooth fluctuations in renewable energy output and improve system flexibility. In the microgrid logical relationship model, distributed power and energy storage device objects coordinate and optimize operation with other load objects based on their own characteristics. Switchgear, distribution transformer terminals, and user smart meters: Refers to the intelligent primary and secondary equipment within the smart substation. Among them, switchgear (such as ring main units and circuit breakers) can realize topology reconstruction and fault isolation within the substation; distribution transformer terminals monitor the substation's gateway power and power quality; and user smart meters collect load data from each user. In the smart substation logical relationship model, the coordination of these device objects can optimize the substation's network losses and power supply reliability.Power generation control and grid connection control unit objects: refers to the control equipment of regional distributed power sources. Among them, the power generation control unit optimizes the scheduling of active and reactive power output of distributed power sources in each substation; the grid connection control unit is responsible for grid connection control functions such as reactive compensation and low voltage ride-through. In the regional distributed power supply logical relationship model, the power generation control and grid connection control units coordinate with each distributed power source to participate in the voltage regulation of the distribution network. Object association method: refers to the method of mapping and fusing different object models during the software modeling process. In this solution, the master station system adopts the object association method to correspond each device object in the logical relationship model to the distributed power supply device model at the physical layer one by one, forming a hierarchical control unit model that unifies the physical properties and logical relationships of the equipment.

[0018] Furthermore, it also includes: S24, the master station system embeds an autonomous control strategy module in the microgrid control layer, the autonomous control strategy module establishes data subscription with the monitoring equipment of the microgrid, and periodically receives real-time measurement data of the source and load nodes in the microgrid; the autonomous control strategy module uses the power flow calculation method based on the received real-time measurement data of the source and load to evaluate the supply and demand balance status within the microgrid; when the supply and demand within the microgrid is unbalanced, the autonomous control strategy module generates power adjustment curves for the key loads, general loads, and controllable loads defined in the microgrid, as well as active output dispatch instructions for the distributed power sources and energy storage devices in the microgrid; the generated power adjustment curves and active output dispatch instructions are respectively sent to the corresponding control devices to eliminate the supply and demand imbalance within the microgrid. The generated power adjustment curve and active power output dispatching instructions are respectively sent to the corresponding load control unit, distributed power supply control unit and energy storage control unit in the microgrid through the distribution automation communication network; each control unit receives and executes the corresponding power adjustment curve or active power output dispatching instruction, adjusts the load power consumption or source power generation power controlled by each control unit, and eliminates the supply and demand imbalance within the microgrid; each control unit feeds back the instruction execution results to the source-load coordination control submodule, the source-load coordination control submodule summarizes and evaluates the feedback results, and outputs the summarized evaluation results to the autonomous control strategy module, which updates the supply and demand balance status within the microgrid based on the summarized evaluation results.

[0019] Preferably, the distribution automation master station system continuously tracks the execution effect of the microgrid autonomous control strategy module. When the source-load coordination control result fed back by the autonomous control strategy module still fails to completely eliminate the supply and demand imbalance within the microgrid, the hierarchical control unit coordination control process at the master station system level is triggered: first, the master station system identifies several smart substations and regional distributed power sources directly adjacent to the current microgrid, extracts the real-time operation measurement data of each smart substation and regional distributed power source, and evaluates its operation margin; then, based on the evaluation results, the master station system formulates a power support plan between the hierarchical control units, generates on-off control instructions for the connection switches between the microgrid and the adjacent smart substations and regional distributed power sources, and sends the switch control instructions to the connection points of the microgrid, smart substations, and regional distributed power sources through the distribution automation communication network, controls the adjacent smart substations and regional distributed power sources to transmit power to the microgrid or the microgrid to feed power to it, and realizes source-load coordination and supply and demand balance in a larger range.

[0020] The autonomous control strategy module refers to the intelligent algorithm module embedded in the microgrid's control layer. By subscribing to real-time data from various monitoring devices within the microgrid, the autonomous control strategy module assesses the microgrid's supply-demand balance. Based on pre-set optimization objectives and constraints, it autonomously generates load power adjustment curves and distributed power generation output dispatch instructions, enabling self-control and management of the microgrid. This module enables the microgrid to operate autonomously, reducing the control frequency and communication pressure on the master station system. Source and load nodes refer to the distributed power generation (source) and load nodes within the microgrid. Source nodes include photovoltaic, wind power, and energy storage devices within the microgrid, which provide power to the microgrid. Load nodes include critical loads, general loads, and controllable loads within the microgrid, which consume power from the microgrid. The microgrid's supply-demand balance depends on the real-time output and power demand of each source and load node. Real-time measurement data refers to real-time operational data collected by various monitoring devices within the microgrid, including the active and reactive output of source nodes, the active and reactive power of load nodes, bus voltage, and line flow. The autonomous control strategy module obtains the current operating status of the microgrid by subscribing to these real-time measurement data, which serves as input for evaluating supply and demand balance and optimizing scheduling.

[0021] Power flow calculation methods refer to mathematical methods for analyzing power flow in power systems. By establishing an equivalent circuit model of the microgrid and solving for state variables such as node voltage and line power, the microgrid's static operating performance can be evaluated, including metrics such as active / reactive power balance and voltage compliance. The autonomous control strategy module uses the power flow calculation results to determine whether the microgrid is currently in a state of supply and demand balance. Power adjustment curves refer to power adjustment plans for various load types within the microgrid (critical loads, general loads, and controllable loads). When a supply and demand imbalance occurs in the microgrid, the autonomous control strategy module generates corresponding power adjustment curves based on the importance and controllability of each load type, representing the load power variation trajectory over each time period. Controllable loads have the largest power adjustment range, while critical loads have the smallest or even zero power adjustment range. Through coordinated power adjustment of various load types, supply and demand rebalancing is achieved in the microgrid. Energy storage devices refer to energy storage devices such as batteries, flywheels, and supercapacitors within the microgrid. Energy storage devices can charge when the microgrid is oversupplied and discharge when demand is insufficient, serving as a crucial regulatory resource for maintaining supply and demand balance in the microgrid. The autonomous control strategy module optimizes the active power output scheduling plan based on the capacity and charge-discharge status of the energy storage device, coordinating with other distributed power sources to smooth power fluctuations within the microgrid. Active power output scheduling instructions: These are active power adjustment instructions issued by the autonomous control strategy module to each distributed power source and energy storage device within the microgrid. Similar to the power adjustment curve of the load, the active power output scheduling instruction specifies the output change plan of each power source over a period of time in the future. The autonomous control strategy module comprehensively considers factors such as the supply and demand gap of the microgrid, the regulation capability of the power source, and the economic efficiency, and optimizes the generation of active power output scheduling instructions to coordinate the output of each power source, ultimately eliminating the supply and demand imbalance in the microgrid.

[0022] Furthermore, the autonomous control strategy module uses a power flow calculation method to evaluate the supply and demand balance state within the microgrid based on the received real-time source and load measurement data, including: the autonomous control strategy module periodically receives the real-time active power and reactive power measurement values ​​of the microgrid's key loads, general loads, and controllable loads, the real-time active output and reactive output measurement values ​​of the distributed power supply and energy storage device, and the flow measurement values ​​of the connection point between the microgrid and the main grid through the data subscription interface with the monitoring equipment in the microgrid, thereby forming a real-time flow measurement data set for each source point, load point, and connection point within the microgrid; the autonomous control strategy module inputs the acquired real-time flow measurement data set into a pre-established microgrid three-phase flow model, uses the measured voltage amplitude and phase angle of each source point, load point, and connection point as the initial value of the state variable, and uses the Newton-Raphson method to perform flow iterative calculation.

[0023] Active power and reactive power refer to two forms of power consumed by loads in a power system. Active power (measured in W or kW) is the power consumed by a load when performing useful work, such as the mechanical power of a motor or the luminous power of a light bulb. Reactive power (measured in var or kvar) is the power consumed by a load when generating an electromagnetic field, such as the excitation power of a transformer or motor. The autonomous control strategy module assesses the total load demand of the microgrid by acquiring the active and reactive power of each load within the microgrid. Active output and reactive output refer to two forms of output from generators or other power devices in a power system. Active output (measured in W or kW) is the active power output of a power device to the grid, which determines the grid's power supply capacity. Reactive output (measured in var or kvar) is the reactive power output or absorption of a power device to the grid, which participates in grid voltage regulation. The autonomous control strategy module assesses the total power supply capacity of the microgrid by acquiring the active and reactive output of each distributed power source and energy storage device within the microgrid.

[0024] Main grid: refers to the higher-level power grid to which the microgrid is connected, typically a distribution network or transmission network. A microgrid can purchase electricity from or sell electricity to the main grid. The autonomous control strategy module evaluates the power exchange status between the microgrid and the main grid by obtaining power flow measurements at the connection point between the microgrid and the main grid. A microgrid three-phase power flow model: refers to a method for mathematically modeling the three-phase power flow of a microgrid using node equations or branch equations. The three-phase power flow model accounts for the three-phase unbalanced characteristics of various components in the power system (such as lines and transformers), enabling accurate analysis of the microgrid's operating status under non-ideal conditions such as asymmetric faults. The autonomous control strategy module uses the microgrid three-phase power flow model to assess the microgrid's real-time supply and demand balance. Phase angle: refers to the phase angle of the AC voltage or current in the power system, reflecting the phase relationship between various electrical quantities. In power flow calculations, the phase angle of each bus voltage is an important state variable, directly affecting the distribution of power flow. The autonomous control strategy module uses the measured voltage phase angle of each measurement point in the microgrid as the initial value of the three-phase power flow model and solves the optimal power flow state of the microgrid through iterative calculation.

[0025] The Newton-Raphson method is a commonly used numerical iterative method for solving nonlinear equations. This method linearizes the nonlinear equations based on the Taylor series expansion principle and continuously modifies the values ​​of the variables through iteration until the residuals of the equations meet the convergence criteria. In power flow calculations, the Newton-Raphson method uses the voltage amplitude and phase angle of each bus as state variables to establish a power imbalance equation. By iteratively solving for the voltage amplitude and phase angle, the voltage is calculated to determine the line power flow and network losses. The autonomous control strategy module uses the Newton-Raphson method to solve the three-phase power flow model of the microgrid. Power imbalance refers to the difference between the actual and calculated power injected by each bus in the power flow model. For balanced nodes (PQ nodes), power imbalance includes both active and reactive power imbalance; for balanced nodes (PV nodes), power imbalance includes only active power imbalance. Power imbalance reflects the deviation between the current power flow calculation results and the actual operating status and serves as a criterion for iterative convergence of the power flow calculation. The autonomous control strategy module determines whether the three-phase power flow calculation of the microgrid has converged by comparing the power imbalance of each bus with the convergence threshold.

[0026] Preferably, the source-load power changes of the microgrid are usually continuous, and the power flow distributions at the previous and next moments are similar. The node voltage amplitude and phase angle at the previous moment can be used as the initial values ​​for the power flow calculation at the current moment to reduce the number of iterations. At the same time, the node power at the current moment can be predicted based on the changing trend of the source-load power, and the initial value setting can be further optimized. The traditional Newton-Raphson method uses a deterministic iterative search strategy, which is difficult to handle the nonlinear and multi-peak characteristics of the microgrid power flow model. Intelligent optimization algorithms, such as particle swarm optimization and differential evolution, can be introduced to transform the power flow calculation into an optimization problem for solution. Intelligent optimization algorithms can jump out of the local optimal solution and find the global optimal power flow distribution through group search and random mutation.

[0027] After each power flow iteration, the autonomous control strategy module calculates the power imbalance of each node in the three-phase power flow model of the microgrid; compares the calculated power imbalance with the convergence threshold; when the power imbalance of each node is less than the convergence threshold, the iterative calculation is stopped, and the voltage amplitude and phase angle of each source point, load point and connection point of the current iteration step are output as the optimal power flow calculation result of the microgrid; the autonomous control strategy module compares the optimal power flow calculation result of the microgrid with the preset microgrid supply and demand balance constraints; if the supply and demand balance constraints are met at the same time, the state evaluation result of the microgrid supply and demand balance is output, otherwise the state evaluation result of the supply and demand imbalance is output.

[0028] Furthermore, the Newton-Raphson method simplifies the Jacobian matrix of the power flow model based on the approximate decoupling of the power angle and the approximate decoupling of the voltage amplitude. Specifically, the autonomous control strategy module receives the real-time power flow measurement data set of the microgrid, extracts the three-phase active and reactive power, voltage amplitude and phase angle measurement values ​​of each source point, load point and connection point, and forms the input vector of the power flow iterative calculation. The autonomous control strategy module constructs the node power imbalance equation based on the three-phase power flow model of the microgrid, adopts Taylor series expansion and ignores the second-order and above terms, and linearizes it into a modified equation group. The power angle approximate decoupling hypothesis is introduced to decouple the modified equation group into a power-voltage amplitude sub-equation group and a power-voltage phase angle sub-equation group. The autonomous control strategy module solves the two decoupled sub-equation groups separately, calculates the correction amount through LU decomposition, and updates the state variables. The autonomous control strategy module iteratively calculates the corrected voltage amplitude and phase angle of each node until the remainder of the corrected equation group is less than the set convergence accuracy. The voltage amplitude and phase angle of the current iteration step are output as the three-phase power flow calculation result of the microgrid at that moment. In each iteration of the autonomous control strategy module, the length prediction algorithm is used to correct the state variables, and the correction equation group is solved in combination with the optimization calculation method to accelerate the convergence of the iterative process. At the same time, the voltage amplitude and phase angle of the previous moment are used as the initial value of the iteration at the current moment to reduce the number of iterations.

[0029] The power angle approximate decoupling approach ignores the partial derivatives of the power imbalance with respect to the voltage amplitude in the Jacobian matrix of the power flow model. This approach assumes that the node's active power injection depends only on the voltage phase angle, and the node's reactive power depends only on the voltage amplitude. This approximation is based on the physical fact that in high-voltage power grids, the line impedance angle is small (i.e., resistance is much smaller than reactance). Therefore, active power is primarily determined by the voltage phase angle difference, and reactive power is primarily determined by the voltage amplitude difference. The power angle approximate decoupling simplifies the Jacobian matrix into a block diagonal matrix, significantly reducing the iterative dimension and computational complexity of the power flow calculation. The voltage amplitude approximate decoupling approach further ignores the partial derivatives of the power imbalance with respect to the voltage phase angle in the Jacobian matrix of the power flow model. This approach assumes that the node's active power injection depends only on the voltage phase angle difference, and the node's reactive power injection depends only on the voltage amplitude difference. This approximation assumes that under normal operation of the high-voltage power grid, the voltage amplitudes at each node are close to the rated value (e.g., 1.0 per unit), and therefore the impact of voltage amplitude differences on active power is negligible. Based on the voltage amplitude approximate decoupling, the Jacobian matrix can be simplified into two independent sub-matrices, corresponding to the active power-phase angle sub-problem and the reactive power-voltage amplitude sub-problem, further improving the computational efficiency of the power flow calculation.

[0030] Furthermore, the supply and demand balance constraints include: the absolute value of the difference between the total active power generation and the total active power consumption in the microgrid is less than P1% of the rated capacity of the microgrid; the voltage amplitude deviation rate of the key load and general load nodes in the microgrid is less than P2%; the voltage amplitude deviation rate of the controllable load nodes in the microgrid is less than P3%.

[0031] P1 represents the allowable range for the ratio of the difference between the total active power generated and consumed within the microgrid to the microgrid's rated capacity. Generally, to ensure supply and demand balance within the microgrid, this ratio should be as close to zero as possible. However, given the uncertainties in microgrid operation and the flexibility of regulation, P1 can be a smaller value between 1% and 5%. A value that is too large will lead to significant power surpluses and deficits within the microgrid, impacting system stability; a value that is too small will place higher demands on the microgrid's regulation accuracy and response speed, increasing control difficulty. P2 represents the maximum allowable deviation rate of the voltage amplitude at key and general load nodes within the microgrid. To ensure the quality of power supply to important users, the voltage at these load nodes should be maintained within a narrow range of the nominal voltage. For power users with voltage levels of 380V and below, the acceptable range of supply voltage deviation is -10% to +7% of the nominal voltage. Therefore, P2 can be set to a more stringent value of 5% to 10% to keep the voltage deviation of key and general loads within a higher standard. P3 represents the maximum allowable deviation rate of the voltage amplitude at controllable load nodes within the microgrid. Compared to critical and general loads, controllable loads typically have a greater tolerance for supply voltage, allowing their operating states to be adjusted within a wider range. Therefore, P3 can take a relatively loose value of 10% to 20%, which not only meets the basic power supply quality requirements of controllable loads but also provides more margin for demand-side response control of the microgrid.

[0032] Furthermore, S21 uses the DBSCAN clustering algorithm to group the device nodes and outputs the division results of the three types of hierarchical control units, namely, microgrid, smart substation and regional distributed power supply, according to the clustering labels, including: the master station system extracts the geographical location coordinates, load type attributes and distributed power supply capacity attributes of the device nodes according to the established distributed power supply equipment model, and constructs a device node feature vector containing the extracted attributes; the master station system inputs the constructed device node feature vector into the DBSCAN clustering algorithm, and sets the minimum clustering parameter of the clustering algorithm to 3, which corresponds to the three types of hierarchical control units, namely, microgrid, smart substation and regional distributed power supply; based on the input device node feature vector, the density peak search algorithm is used to determine the cluster center, and the DB The SCAN clustering algorithm classifies feature vectors with cluster centers as the core, and outputs clustering results containing the cluster labels of each device node; the master station system identifies the type of hierarchical control unit to which each device node belongs based on the cluster label, and divides the device node with a cluster label of 1 into a microgrid, the device node with a cluster label of 2 into a smart substation, and the device node with a cluster label of 3 into a regional distributed power supply; when the established distributed power supply device model changes, the master station system extracts the characteristic attributes of the changed distributed power supply device node, updates the feature vector of the corresponding device node, and inputs the updated feature vector into the DBSCAN clustering algorithm. The above steps are repeated to receive the updated clustering results and adjust the hierarchical control unit division results of the device node.

[0033] Among them, load type attributes and distributed generation capacity attributes reflect the power consumption characteristics and power supply capabilities of device nodes. The load type attribute indicates the power consumption characteristics of the device node, such as industrial, commercial, and residential loads. Different load types have significantly different characteristics, such as daily load curves and power factors. The distributed generation capacity attribute indicates the rated capacity of the distributed generation configured on the device node, such as the installed capacity of a photovoltaic power plant or a wind farm. The capacity of the distributed generation directly determines the node's power supply capabilities. Load type and distributed generation capacity attributes are important criteria for classifying hierarchical control units such as microgrids and smart substations. The density peak search algorithm is a cluster center selection method based on the local density and distance of samples. This algorithm calculates the local density (i.e., the number of sample points in the neighborhood) and the minimum distance to points with higher local density for each sample point. Sample points with higher local density and farther distances to points with higher density are selected as cluster centers. Compared with other clustering algorithms, the density peak search algorithm does not require a preset number of cluster centers and can adaptively discover cluster structures of any shape. In this solution, the master station system uses a density peak search algorithm to determine the core of the device node clustering to improve the accuracy and efficiency of clustering. Cluster label: refers to the category identification obtained after the DBSCAN clustering algorithm classifies each sample point. Cluster labels are usually represented by natural numbers, and different cluster labels correspond to different cluster clusters. For noise points (that is, outliers that do not belong to any cluster cluster), their cluster labels are generally marked as -1. In this solution, the master station system uses the DBSCAN algorithm to obtain the cluster labels of the device nodes, and identifies the hierarchical control unit type of each node based on the cluster labels: nodes with a label of 1 are divided into microgrids, nodes with a label of 2 are divided into smart substations, and nodes with a label of 3 are divided into regional distributed power supplies. Cluster labels intuitively reflect the degree of similarity of device nodes in network topology and physical properties.

[0034] Furthermore, a density peak search algorithm is used to determine the cluster center, including: calculating the Euclidean distance between the feature vectors of device nodes to generate a node distance matrix; based on the node distance matrix, calculating the local density ρ of each device node i and the relative distance δ i ; where ρ i Indicates that the distance to node i is less than the cutoff distance d c The number of other nodes; δ i Represents the minimum distance between node i and the local density node; based on the calculated local density ρ of the device node i and the relative distance δ i , set the density threshold ρ min is the mean of the local density of all nodes, and the distance threshold δ is determined by an iterative method min , the number of cluster centers under the current threshold is not less than the preset minimum number of clusters; ρi >ρ min And δ i >δ min Node i is determined as the cluster center. Using the determined cluster center as the core, the distance between each non-cluster center node and each cluster center is calculated, and the corresponding node is classified into the category of the closest cluster center to form the clustering result. When a new device is connected or disconnected, the DBSCAN clustering algorithm extracts the characteristic attributes of the newly added or disconnected device node, updates the node feature vector, recalculates the node distance matrix, node local density, and relative distance, and adjusts the determined cluster center based on the updated node local density and relative distance. Then, the categories of the affected device nodes are automatically adjusted to dynamically update the clustering results.

[0035] Furthermore, S23, the coordinated control between the control units of the master station system includes: the master station system maps the logical relationship models of the three types of hierarchical control units of the microgrid, smart substation and regional distributed power supply constructed with the established distributed power supply equipment model respectively; the object association method is used to correspond the objects in the logical relationship model with the corresponding physical equipment entities in the distributed power supply equipment model one by one, and the microgrid, smart substation and regional distributed power supply hierarchical control unit model containing the actual equipment composition topology and internal logical control structure is generated by fusion; the master station system extracts the monitoring equipment objects inside each hierarchical control unit based on the generated hierarchical control unit model, and establishes a protocol data subscription relationship with the monitoring equipment. The system periodically receives the real-time operation measurement data of microgrids, smart substations and regional distributed power sources uploaded by monitoring equipment; the master station system inputs the received real-time operation measurement data of each hierarchical control unit into the operation status evaluation rules based on threshold values ​​and fuzzy rules predefined in the corresponding hierarchical control unit model, and uses threshold value judgment and fuzzy logic reasoning methods to evaluate the real-time operation status of each hierarchical control unit; the master station system uses the power flow calculation method to determine the real-time supply and demand balance status of each hierarchical control unit based on the generated hierarchical control unit operation status evaluation results, combined with the topological connection relationship of microgrids, smart substations and regional distributed power sources in the distribution network, and generates a power dispatch plan.

[0036] Specifically, the power dispatching plan is generated, including: when the operating status of the hierarchical control unit evaluated by the master station system is normal, the current control mode of the hierarchical control unit is maintained unchanged, and the default control parameters under normal working conditions are issued to the control equipment in the hierarchical control unit; when the operating status of the microgrid hierarchical control unit evaluated by the master station system is an alarm, the key measurement data in the microgrid is extracted, the fault diagnosis rules based on threshold values ​​predefined in the microgrid logical relationship model are input, and the fault type in the microgrid is determined by the threshold judgment method; according to the diagnosed fault type, the predefined fault handling rules are queried in the microgrid logical relationship model. The strategy is to send the control instructions such as load reduction, controllable load removal or distributed power supply decoupling of the microgrid to the corresponding execution unit through the communication protocol between the master station system and the microgrid control equipment, so as to realize the autonomous isolation of the internal fault of the microgrid; when the operating status of the smart substation or regional distributed power supply evaluated by the master station system is an alarm, the key measurement data is extracted, the fuzzy logic reasoning method is used to calculate the fault severity, and the calculation result is compared with the preset severity threshold; when the fault severity exceeds the threshold, the master station system formulates a fault removal plan based on the logical relationship model of the smart substation and regional distributed power supply. The scheme uses the communication protocol between the master station system and the control device to issue fault removal instructions to the corresponding switchgear to achieve fault isolation. When the master station system assesses the operating status of a hierarchical control unit as overloaded or underloaded, it extracts key measured quantities within the overloaded or underloaded hierarchical control unit and, combined with the topological connection relationship between the hierarchical control units, uses a power flow calculation method to calculate the real-time supply and demand balance status of each hierarchical control unit. Based on the calculated supply and demand balance status and the adjustment capacity of the controllable resources within the hierarchical control unit, a dynamic programming algorithm is used to generate the source-load coordination control strategy within the hierarchical control unit and the power exchange control strategy between the hierarchical control units. Through the communication protocol between the master station system and the control device, the source-load coordination control strategy is issued to the generation control unit and load control unit within the hierarchical control unit, and the power exchange control strategy is issued to the tie switch control unit connecting adjacent hierarchical control units, achieving autonomous coordinated control within the hierarchical control unit and between adjacent control units. The master station system receives feedback on the control instruction execution of the execution unit within the hierarchical control unit, evaluates the improvement of the control effect on fault isolation and supply and demand balance, and uses the evaluation results as input for the collaborative generation of power dispatching plans for microgrids, smart substations, and regional distributed power sources for the next period.

[0037] Real-time operational measurement data refers to real-time operating condition data collected by monitoring equipment within hierarchical control units, such as microgrids, smart substations, and regional distributed power sources. This data typically includes electrical parameters such as bus voltage, line current, active power, reactive power, and frequency, as well as external factors such as ambient temperature, light intensity, and wind speed. The master station system periodically receives and stores this real-time operational measurement data by establishing a data subscription relationship with the monitoring equipment, forming a dataset reflecting the dynamic operating status of each hierarchical control unit. Real-time operational measurement data is an important basis for evaluating the operating status of hierarchical control units and formulating control strategies. Operational status assessment rules based on thresholds and fuzzy rules refer to a series of logical judgment rules predefined by the master station system for evaluating the operating status of hierarchical control units. These operational status assessment rules primarily fall into two categories: hard judgment rules based on thresholds and soft judgment rules based on fuzzy rules. Hard judgment rules based on thresholds use clear numerical boundaries to classify the operating status of hierarchical control units into normal, warning, and fault levels. For example, they determine whether the bus voltage exceeds ±10% of the rated value or whether the line current exceeds the thermal stability limit. The advantages of hard judgment rules are clear logic and simple calculations, but the disadvantage is that it is difficult to fully reflect the complex system status. Soft judgment rules based on fuzzy rules use linguistic variables and membership functions to describe the operating status of hierarchical control units as qualitative levels such as good, general, and poor. For example, it determines whether the active power is at a "large" level, whether the voltage is "close" to the rated value, etc. Soft judgment rules use combined reasoning of multiple fuzzy propositions to obtain a qualitative assessment of the comprehensive operating status of the hierarchical control unit, and can better handle uncertainty and nonlinear problems. The master station system inputs real-time operation measurement data into predefined operating status evaluation rules, and uses threshold value judgment and fuzzy logic reasoning methods to perform comprehensive operations to obtain quantitative or qualitative hierarchical control unit operating status evaluation results, providing input for subsequent coordinated control of control units.

[0038] Beneficial effects: Compared with the prior art, the advantages of this application are:

[0039] By adopting a modeling method based on the CIM standard, unifying the distributed power supply equipment model and data interaction interface, real-time two-way data synchronization is achieved between the master station system and the external system, ensuring that the distributed power supply equipment model in the master station system is consistent with the actual operating status, providing a data basis for subsequent precise regulation.

[0040] By using the DBSCAN clustering algorithm to adaptively group device nodes, the division range of microgrids, smart substations and regional distributed power sources can be automatically adjusted according to the real-time topological changes of the distribution network, so that the division of hierarchical control units always matches the physical structure of the distribution network, improving the adaptability and flexibility of hierarchical control.

[0041] When constructing the hierarchical control unit model, object-oriented modeling and object association methods were used to map and fuse the logical relationship model with the physical device model, forming an integrated model that encompasses the actual device topology and internal logical control structure. This modeling approach enables rapid identification and precise positioning of monitoring and control devices within the hierarchical control unit, providing a convenient channel for subsequent data collection and control command issuance, and improving the real-time and reliability of hierarchical control.

[0042] Embedding an autonomous control strategy module within the microgrid's hierarchical control unit evaluates the microgrid's internal supply and demand balance through real-time power flow calculations. Based on the evaluation results, the system autonomously adjusts the power of critical, general, and controllable loads, and dispatches the output of distributed power sources and energy storage devices. This achieves self-generation and self-supply within the microgrid, balances supply and demand, and optimizes operation. Incorporating microgrids as independently controllable units into the hierarchical control system can reduce control pressure on the master station system and improve the power supply reliability and economic efficiency of the distribution network.

[0043] The autonomous control strategy module uses an improved Newton-Raphson method for iterative power flow calculations. By decoupling and simplifying the Jacobian matrix, it accelerates convergence. Furthermore, using actual measurements at each source and load node as the initial values ​​for the state variables avoids the convergence difficulties associated with flat-start calculations in conventional power flow calculations, improving the real-time and robustness of the autonomous control strategy.

[0044] Based on the real-time operational measurement data and evaluation results of each hierarchical control unit, and in combination with the topological connectivity of each unit within the distribution network, the master station system uses power flow calculation methods to optimize and generate power dispatch plans, enabling collaborative interaction among the hierarchical control units. Through the autonomous operation and regional coordination of control units at each level, the flexibility and adaptability of distributed control are leveraged while also balancing the integrity and optimality of global control, improving the economic efficiency and safety of distribution network operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is an exemplary flow chart of a method for regional autonomous hierarchical control of an active distribution network based on a master station system according to some embodiments of this specification;

[0046] Figure 2 is an exemplary flow chart for constructing a hierarchical regulatory unit model according to some embodiments of this specification;

[0047] Figure 3 is an exemplary flow chart of classification based on the DBSCAN clustering algorithm according to some embodiments of this specification;

[0048] Figure 4 This is an exemplary flow chart for adjusting the supply and demand balance of a microgrid according to some embodiments of this specification. DETAILED DESCRIPTION

[0049] The methods and systems provided in the embodiments of this specification are described in detail below with reference to the accompanying drawings.

[0050] Figure 1 This is an exemplary flowchart of an active distribution network regional autonomous hierarchical control method based on a master station system according to some embodiments of this specification, S1, adopting a modeling method based on the common information model CIM standard, and establishing a distributed power supply equipment model with the equipment model and equipment data of the distributed power supply that interacts with the external system; S2, the master station system divides a plurality of hierarchical control units according to the distribution network topology and the established distributed power supply equipment model, constructs a logical relationship model of the hierarchical control unit, maps and integrates the logical relationship model with the distributed power supply equipment model, and generates a hierarchical control unit model including monitoring equipment, control equipment and internal logical structure; S3, the master station system collects the operating status data of each control unit in real time through the monitoring equipment in the hierarchical control unit model, inputs the data into a pre-trained neural network evaluation model, and evaluates the operating status of each control unit; S4, the master station system collaboratively controls each control unit according to the evaluation results of the operating status of each control unit.

[0051] In a preferred embodiment of the present application, the master station system and the external system establish a public device model and a proprietary device model of the distributed power supply in accordance with the CIM (Common Information Model) specification defined in the IEC61970 / 61968 series of standards. The public device model adopts the Distributed Generator class in the CIM standard, which includes the basic properties of the distributed power supply, such as type (gen Type), capacity (max P) and geographic location coordinates (Geo Location). Among them, the type attribute defines the energy type of the distributed power supply, such as wind energy, solar energy, energy storage, etc.; the capacity attribute defines the maximum output of the distributed power supply; and the geographic location coordinate attribute defines the spatial position of the distributed power supply in the geographic information system. The proprietary device model adopts the extension mechanism of the CIM standard, and on the basis of the Distributed Generator class, adds properties that reflect the real-time operating status and control characteristics of the distributed power supply, such as real-time dispatchable capacity (available P), fault removal time (clearTime) and control mode (control Mode). Among them, the real-time dispatchable capacity attribute dynamically records the upper limit of the active power output that the distributed power source can currently participate in dispatching; the fault removal time attribute defines the maximum operating time of the distributed power source in the event of a fault; and the control mode attribute defines the output control method of the distributed power source, such as constant power output and reactive power priority.

[0052] Based on the CIM standard, the master system and external systems use SOAP (Simple Object Access Protocol)-based Web Services technology to establish an interactive interface for device models and device data. The external system uploads the CIM device model of the distributed power source (DG) to the master system by calling the Web Service interface published by the master system. Simultaneously, the external system also uploads the DG's nameplate parameters (such as rated capacity and rated voltage) and operating parameters (such as current active and reactive output and current voltage) in the form of device data to the master system via the Web Service interface. After receiving the device model and device data, the master system stores them in the unified distribution network model database to build a complete DG device model. When the device model or device data in the external system changes (such as adding or removing a DG or changing DG parameters), the external system pushes the change data in real time to the master system via the Web Service interface in the form of incremental data. Upon receiving the change data, the master system extracts the added, modified, or deleted device models and device data and, based on the data operation type, performs corresponding updates in the unified distribution network model database, enabling dynamic updates of the DG device model.

[0053] In a preferred embodiment of the present application, the master station system extracts key features of the device node based on the established distributed power device model, including geographic location coordinates, load type attributes, and distributed power capacity attributes. The geographic location coordinates are obtained from the Geo Location attribute of the device model, the load type attribute is obtained from the device archive data, and the distributed power capacity attribute is obtained from the max P attribute of the device model. The master station system constructs the extracted key features into a triple (x i ,y i ,t i ,g i ), as the feature vector of each device node i. Among them, x_i and y_i represent the horizontal and vertical coordinates of the geographical location of node i, t i represents the load type of node i (such as industrial load, commercial load, residential load, etc.), g i Represents the installed capacity of the distributed power supply at node i. Specifically, there are 10 device nodes in a distribution network, numbered from N1 to N10. The master station system extracts the key features of each node from the distributed power supply device model and constructs the feature vector [(x i ,y i ,t i ,g i ), i=1,2,......,10], as shown in Table 1 below:

[0054] Table 1

[0055] Node number Geographic location coordinates (x, y) Load type Distributed power generation capacity (kW) N1 (1.2，3.5) 1 500 N2 (2.7，1.8) 2 0 N3 (4.1，4.3) 1 800 N4 (3.6，2.9) 3 200 N5 (0.9，1.5) 2 0 N6 (2.4，4.7) 1 600 N7 (3.3，0.7) 3 150 N8 (1.8，2.6) 2 0 N9 (4.5，3.2) 1 1000 N10 (2.1，3.9) 2 0

[0056] Figure 2 This is an exemplary flow chart for constructing a hierarchical control unit model according to some embodiments of this specification. The master station system inputs the constructed device node feature vector into the DBSCAN (Density-based Spatial Clustering of Applications with Noise) clustering algorithm to adaptively group the device nodes. The DBSCAN algorithm is a density-based clustering algorithm that can automatically determine the number and shape of clusters based on the spatial distribution characteristics of the sample points. When inputting the device node feature vector, the master station system sets the minimum clustering parameter of the DBSCAN algorithm to 3, corresponding to the three types of hierarchical control units: microgrid, smart substation, and regional distributed power supply.

[0057] Figure 3 This is an exemplary flow chart of classification based on the DBSCAN clustering algorithm shown in some embodiments of this specification. The DBSCAN algorithm first calculates the Euclidean distance between the feature vectors of device nodes to generate an N×N node distance matrix D, where N is the total number of device nodes. Then, based on the distance matrix D, the algorithm calculates the local density ρ of each device node i. i and the relative distance δ i Among them, ρ i It means that the node i is the center and the truncation distance d c The number of other nodes in δ i Represents the distance between node i and the nearest node with higher density than it. The algorithm sets the density threshold ρ according to statistical laws. min is the mean of the local density of all nodes, and the distance threshold δ is determined by an iterative method min , so that the number of cluster centers under the current threshold is not less than the preset minimum number of clusters 3. i >ρ min And δ i >δ minNode i is determined as the cluster center. The DBSCAN algorithm outputs a clustering result matrix containing the cluster labels of each device node. The master station system identifies the type of hierarchical control unit to which each device node belongs based on the cluster label, classifying device nodes with a cluster label of 1 as microgrids, device nodes with a cluster label of 2 as smart substations, and device nodes with a cluster label of 3 as regional distributed power sources. When the established distributed power supply device model changes (such as adding or removing devices, device parameter changes, etc.), the master station system extracts the characteristic attributes of the changed distributed power supply device nodes, updates the feature vectors of the corresponding device nodes, re-inputs the updated feature vectors into the DBSCAN clustering algorithm, executes the above clustering process, and dynamically adjusts the hierarchical control unit division results of the device nodes.

[0058] Specifically, the master system inputs the feature vector into the DBSCAN algorithm, sets the minimum number of clusters to 3, and the cutoff distance d c The algorithm first calculates the Euclidean distance matrix D (10x10) between nodes. Based on the distance matrix, the algorithm calculates the local density ρ of each node i and the relative distance δ i , see Table 2 for details:

[0059] Table 2

[0060]

[0061]

[0062] The algorithm takes the determined cluster center as the core and uses the density peak search method to classify the remaining non-cluster center nodes. For each non-cluster center node, the Euclidean distance between it and each cluster center is calculated, and the node is classified into the category of the nearest cluster center. The above steps are iterated until all nodes are assigned to a category to form the final clustering result. Set the density threshold ρ min =2.5, and the distance threshold δ is determined by iteration min =1.0, the number of cluster centers under the current threshold is not less than 3. min And the relative distance is greater than δ min Nodes N1, N4, and N10 are identified as cluster centers, representing microgrids, smart substations, and regional distributed power supplies, respectively. With the cluster center as the core, the distances from other nodes to each center are calculated and classified: N2 is closest to cluster center N4 and is classified as a smart substation; N3, N6, and N9 are closest to cluster center N1 and are classified as microgrids; and N5, N7, and N8 are closest to cluster center N10 and are classified as regional distributed power supplies. The final clustering results are shown in Table 3:

[0063] Table 3

[0064] Cluster labels Corresponding layered units Contains device nodes 1 microgrids N1, N3, N6, N9 2 Smart station N2, N4 3 Regional distributed power supply N5, N7, N8, N10

[0065] At this time, if a new device node N11 is added to the distribution network, its eigenvector is (2.8, 2.3, 2, 0). The master station system re-inputs the updated eigenvector into the DBSCAN algorithm and obtains a new clustering result: node N11 is classified as a smart substation, and the original clustering result remains unchanged. Based on the hierarchical control unit division results, the master station system uses an object-oriented modeling method to construct a logical relationship model within each hierarchical control unit. Specifically, in the microgrid logical relationship model, key loads, general loads, controllable loads, distributed power sources, and energy storage device objects are defined, and control associations between these objects are established based on control mode attributes. In the smart substation logical relationship model, switchgear, distribution transformer terminals, and user smart meter objects are defined, and a hierarchical structure between these objects is established based on topological relationships. In the regional distributed power supply logical relationship model, power generation control voltage and grid connection control unit objects are defined.

[0066] Specifically, in a distribution network, the master station system uses the DBSCAN clustering algorithm to divide the device nodes in the distribution network into three microgrids, two smart substations, and one regional distributed power supply unit. The device nodes in each unit are as follows: Microgrid 1: N1, N3, N6, N9; Microgrid 2: N12, N15, N18; Microgrid 3:

[0067] N22, N24, and N27; Smart Substation 1: N2, N4, and N11; Smart Substation 2: N14, N16, and N20; and Regional Distributed Generation (RDGs): N5, N7, N8, and N10. In S22, the master station system uses an object-oriented modeling approach to construct logical relationship models within each hierarchical control unit: the microgrid logical relationship model. For microgrid 1, the master station system defines the following objects: critical loads: hospital (N1) and emergency command center (N6); general loads: residential electricity (N3); controllable loads: electric vehicle charging stations (N9); distributed generation: rooftop photovoltaics (N1 and N3) and wind power (N9); and energy storage devices: battery energy storage (N6). Based on the control mode attributes of each object, the master station system establishes the following control relationships: Prioritizing power supply for critical loads, with energy storage and distributed generation providing backup power. Controllable loads adjust their power based on the microgrid's supply-demand balance through a demand-side response mechanism. Distributed generation and energy storage are coordinated and controlled through the microgrid's energy management system to maintain power balance in the microgrid. The logical relationship models for Microgrids 2 and 3 are similar and will not be further detailed here.

[0068] The logical relationship model for smart substations. For Smart Substation 1, the master station system defines the following objects: switchgear: feeder switches (N2), ring main unit switches (N4, N11); distribution transformers: distribution transformers (N4); and user smart meters: residential users (N2) and industrial and commercial users (N11). Based on the topological relationships of these devices, the master station system establishes the following hierarchical structure: feeder switches are located at the starting point of the power supply line and control the total power supply for the entire smart substation; ring main unit switches are located on each branch within the substation and control the power supply to each branch; distribution transformers monitor the operating status of distribution transformers and work with ring main unit switches to achieve fault isolation and load transfer within the substation; user smart meters are located at each user, collecting electricity usage information and implementing demand-side management. The logical relationship model for Smart Substation 2 is similar and will not be repeated here. The master station system defines the following objects for the regional distributed power supply (DG) logical relationship model: the power generation control unit (wind turbine controller (N5), photovoltaic controller (N7), and energy storage controller (N10); and the grid connection control unit (microgrid connection point (N8)). Based on the electrical connection relationships of these objects, the master station system establishes the following logical relationships: the power generation control unit controls the active / reactive output and operating status of each type of DG; the grid connection control unit controls the active / reactive power exchange between the regional DG and the distribution network trunk line via the microgrid connection point.

[0069] Figure 4 This is an exemplary flow chart for adjusting the supply and demand balance of a microgrid, as described in some embodiments of this specification. The autonomous control strategy module establishes a data subscription relationship with monitoring devices within the microgrid, periodically receiving the following real-time measurement data: active and reactive power of critical loads, general loads, and controllable loads; active and reactive output of distributed power sources and energy storage devices; and power flow at the microgrid's connection point with the main grid. Through the data subscription interface, the autonomous control strategy module obtains real-time power flow measurement datasets for each source-load node and connection point within the microgrid.

[0070] The autonomous control strategy module inputs a real-time power flow measurement dataset into a pre-established three-phase power flow model for the microgrid and uses the Newton-Raphson method for iterative power flow calculations. Compared to conventional power flow algorithms, the Newton-Raphson method introduces the following technical features: Using the measured voltage amplitudes and phase angles at the source-load nodes and connection points as the initial values ​​of the state variables improves the convergence speed and accuracy of the iterative calculations; Based on approximate decoupling of power angles and voltage amplitudes, it simplifies the Jacobian matrix of the power flow model and reduces the computational complexity. After each iteration, the autonomous control strategy module calculates the power imbalance at each node and compares it with a preset convergence threshold. When the imbalances at each node are less than the threshold, convergence is considered met, and the voltage amplitudes and phase angles at each node for the current iteration are output as the optimal power flow calculation result for the microgrid.

[0071] The autonomous control strategy module compares the microgrid's optimal power flow calculation results with preset supply and demand balance constraints to comprehensively assess the microgrid's supply and demand balance status. These supply and demand balance constraints include: the absolute value of the difference between the total active power generated and consumed within the microgrid is less than P1% of the microgrid's rated capacity; the voltage amplitude deviation rate at critical and general load nodes within the microgrid is less than P2%; and the voltage amplitude deviation rate at controllable load nodes within the microgrid is less than P3%. P1, P2, and P3 are preset evaluation thresholds based on the microgrid's operational requirements. If the power flow calculation results meet all of these constraints, a supply and demand balance assessment is output; otherwise, a supply and demand imbalance assessment is output. By comparing quantitative indicators, the microgrid's supply and demand balance status can be automatically determined.

[0072] When the assessment results indicate an imbalance between supply and demand within the microgrid, the autonomous control strategy module takes the following measures, generating and issuing control instructions to address the imbalance: It generates power adjustment curves for critical loads, general loads, and controllable loads and issues them to the corresponding load control devices; and it generates active power output dispatch instructions for distributed power sources and energy storage devices and issues them to the corresponding power generation control devices. These load power adjustment curves and power output dispatch instructions can be generated using intelligent algorithms, such as genetic algorithms and particle swarm optimization, based on the microgrid's operational safety constraints and economic optimization objectives. By autonomously regulating the power of both sources and loads, the microgrid achieves supply and demand balance. Specifically, a distribution microgrid with a rated capacity of 500 kW consists of one critical load node, two general load nodes, one controllable load node, two distributed power generation nodes (one each for photovoltaic and wind power generation), one energy storage node, and one connection point to the main grid. The microgrid's autonomous control strategy module establishes a data subscription relationship with the monitoring devices at each node, receiving real-time power flow measurement data every five minutes. At a certain moment, the autonomous control strategy module receives the following real-time power flow measurement data as shown in Table 4:

[0073] Table 4

[0074]

[0075] The autonomous control strategy module inputs the above real-time power flow measurement data into the microgrid three-phase power flow model and uses the Newton-Raphson method to calculate the power flow. The convergence threshold is set to 0.01. After four iterations, the power imbalance of each node is less than the threshold, and the power flow calculation converges. The optimal power flow calculation results are shown in Table 5:

[0076] Table 5

[0077]

[0078]

[0079] The autonomous control strategy module compares the optimal power flow calculation results with the preset supply and demand balance constraints: the total active power generation is 219.8kW, the total active power consumption is 280kW, and the absolute value of the difference between the two is 60.2kW, accounting for 12.04% of the rated capacity of the microgrid, which is greater than the constraint threshold P1 (set to 5%); the voltage amplitude deviation rates of critical load and general load nodes are 1.8%, 2.7% and 2.5% respectively, all less than the constraint threshold P2 (set to 3%); the voltage amplitude deviation rate of controllable load nodes is 4.5%, which is greater than the constraint threshold P3 (set to 4%).

[0080] Comprehensive comparison results indicate an imbalance in supply and demand in the microgrid. The autonomous control strategy module generates the following control instructions: maintain the current power levels of critical and general loads; reduce the active power of controllable loads from 60kW to 40kW; increase the active output of the photovoltaic power source from 150.1kW to its rated output of 180kW; increase the active output of the wind power source from 99.8kW to its rated output of 120kW; and adjust the active output of the energy storage device from -30.1kW (charging) to 20kW (discharging). After the control instructions were issued, the total active power generation in the microgrid reached 320kW, and the total active power consumption reached 320kW, achieving a balanced supply and demand. The voltage amplitudes at each load node also returned to a reasonable range. Through the adjustments made by the autonomous control strategy module, the supply and demand imbalance in the microgrid was eliminated.

[0081] The master station system maps the logical relationship models of the three hierarchical control units (microgrids, smart substations, and regional distributed power sources) to the established distributed power source equipment models. The logical relationship models describe the logical control relationships between objects within each hierarchical control unit (such as loads, power sources, and switchgear); the distributed power source equipment models describe the physical device composition and topological connectivity of distributed power sources in the actual distribution network. The master station system uses an object association method to map the objects in the logical relationship models to the corresponding physical device entities in the distributed power source equipment models. This fusion generates a hierarchical control unit model that incorporates the actual device composition topology and internal logical control structure. This model fusion method organically combines the logical and physical models, enabling the hierarchical control unit model to fully reflect the actual operating conditions of the distribution network. Based on the generated hierarchical control unit models, the master station system extracts the monitoring device objects within each unit (such as smart meters, micro-meteorological stations, and load monitoring terminals) and establishes protocol data subscription relationships with the monitoring devices. On this basis, the master station system periodically receives real-time operational measurement data from monitoring equipment on microgrids, smart substations, and regional distributed power sources. This data includes active / reactive power, voltage amplitude, and phase angle at load nodes; active / reactive output, grid-connected voltage, and frequency at distributed power source nodes; and the on / off status and conduction current of switchgear nodes. By subscribing to data from monitoring equipment, the master station system obtains comprehensive, real-time operational information for the hierarchical control units, providing data support for subsequent operational status assessments.

[0082] The master station system receives the real-time operating measurement data of each hierarchical control unit and inputs it into the predefined operating status evaluation rules in the corresponding unit model. The evaluation rules are constructed based on threshold value judgment and fuzzy logic reasoning methods: Threshold value judgment determines whether the actual operating indicators exceed the limit by setting the threshold value of key indicators (such as voltage deviation, current overlimit, etc.), and quickly diagnoses obvious abnormal conditions; fuzzy logic reasoning sets the fuzziness of key indicators and formulates fuzzy reasoning rules for multiple indicator combinations to comprehensively analyze the fuzziness of indicators and evaluate complex or hidden abnormal conditions. The master station system uses rule engine technology to automatically execute predefined evaluation rules and generate real-time operating status evaluation results (such as "normal", "warning", "serious alarm", etc.) for microgrids, smart substations, and regional distributed power sources.

[0083] Based on the operational status assessment results and the topological connections between microgrids, smart substations, and regional distributed power sources in the distribution network, the master station system uses a three-phase power flow calculation method to determine the real-time supply and demand balance of each hierarchical control unit and generate a corresponding power dispatch plan. Specifically, for units with balanced supply and demand, the current operating mode is maintained and no dispatch instructions are issued. For units with imbalanced supply and demand, the load power adjustment curve and power output dispatch instructions are optimized to guide them back to balance. For units with continuous imbalanced supply and demand, the power flow between them and other units is coordinated to achieve regional supply and demand balance through joint dispatch.

Claims

1. A method for regional autonomous hierarchical control of an active distribution network based on a master station system, comprising: S1, using a modeling method based on the Common Information Model (CIM) standard, the distributed power supply device model and device data that interact with the external system to establish a distributed power supply device model; S2: The master station system divides the distribution network topology and the established distributed power supply equipment model into multiple hierarchical control units, constructs a logical relationship model of the hierarchical control units, and maps and integrates the logical relationship model with the distributed power supply equipment model to generate a hierarchical control unit model that includes monitoring equipment, control equipment, and internal logical structure, including: The master station system embeds an autonomous control strategy module in the microgrid control layer. The autonomous control strategy module establishes data subscription with the microgrid monitoring equipment and periodically receives real-time measurement data from the source and load nodes in the microgrid. The autonomous control strategy module uses the power flow calculation method based on the received real-time source and load measurement data to evaluate the supply and demand balance status within the microgrid: The autonomous control strategy module receives the real-time flow measurement data set of the microgrid, extracts the three-phase active and reactive power, voltage amplitude and phase angle measurement values ​​of each source point, load point and tie point, and forms the input vector of the flow iterative calculation; the autonomous control strategy module constructs the node power imbalance equation based on the microgrid three-phase flow model, adopts Taylor series expansion and ignores the second-order and above terms, and linearizes it into a modified equation group; introduces the power angle approximate decoupling hypothesis, and decouples the modified equation group into the power-voltage amplitude sub-equation group and the power-voltage phase angle sub-equation group; the autonomous control strategy module calculates the two decoupled sub-equations respectively. Solve the problem, calculate the correction value through LU decomposition, and update the state variables; the autonomous control strategy module iteratively calculates the corrected voltage amplitude and phase angle of each node until the remainder of the correction equation group is less than the set convergence accuracy; output the voltage amplitude and phase angle of the current iteration step as the microgrid three-phase power flow calculation result at the corresponding moment; during each iteration of the autonomous control strategy module, the length prediction algorithm is used to correct the state variables, and the correction equation group is solved in combination with the optimization calculation method to accelerate the convergence of the iterative process; at the same time, the voltage amplitude and phase angle of the previous moment are used as the initial value of the current iteration to reduce the number of iterations; S3, the master station system collects the operating status data of each control unit in real time through the monitoring equipment in the hierarchical control unit model, inputs the data into the pre-trained neural network evaluation model, and evaluates the operating status of each control unit; S4, the master station system coordinates and controls each control unit according to the evaluation results of the operation status of each control unit.

2. The method for regional autonomous hierarchical control of active distribution networks based on a master station system according to claim 1, characterized in that: S1, establish a distributed power supply equipment model, including: The master station system and external systems define the public equipment model and proprietary equipment model of the distributed power source in accordance with the CIM standard. The public equipment model includes the type, capacity, and geographic location coordinates of the distributed power source; the proprietary equipment model includes the real-time dispatchable capacity, fault removal time, and control mode of the distributed power source. The master station system and external system establish an interactive interface for device models and device data based on the CIM standard. The device data includes the nameplate parameters and operating parameters of the distributed power supply. When the device model or device data of the external system changes, the changes are pushed to the main system through the interactive interface.

3. The method for regional autonomous hierarchical control of active distribution network based on master station system according to claim 1 is characterized in that: S2, generates a hierarchical control unit model that includes monitoring equipment, control equipment, and internal logical structure, including: In step S21, the master station system extracts the geographical coordinates, load type, and distributed power capacity of the device nodes as key features based on the established distributed power device model. Based on the extracted key features, the DBSCAN clustering algorithm is used to group the device nodes and output the division results of the three hierarchical control units: microgrid, smart substation, and regional distributed power according to the cluster labels. S22, the master station system uses an object-oriented modeling method to construct a logical relationship model within each hierarchical control unit based on the hierarchical control unit division results; wherein, in the microgrid logical relationship model, key loads, general loads, controllable loads, distributed power sources, and energy storage device objects are defined, and control associations between the objects are established based on control mode attributes; in the smart substation logical relationship model, switchgear, distribution transformer terminals, and user smart meter objects are defined, and a hierarchical structure between the objects is established based on topological relationships; in the regional distributed power supply logical relationship model, power generation control voltage and grid connection control unit objects are defined; S23, the master station system adopts the object association method to merge the logical relationship model objects with the device entity mapping of the distributed power supply equipment model, and generates a hierarchical control unit model that includes the device composition topology and internal logic; extracts the monitoring device and control device objects in the hierarchical control unit model, and establishes data subscription and control instruction issuance channels respectively for collaborative control between control units.

4. The method for regional autonomous hierarchical control of active distribution networks based on a master station system according to claim 3 is characterized in that: S2, generating a hierarchical control unit model including monitoring equipment, control equipment and internal logical structure, and also including: When the supply and demand within the microgrid are unbalanced, the autonomous control strategy module generates power adjustment curves for key loads, general loads, and controllable loads defined within the microgrid, as well as active output dispatch instructions for distributed power sources and energy storage devices within the microgrid; The generated power adjustment curve and active output dispatch instructions are respectively sent to the corresponding control devices to eliminate the supply and demand imbalance within the microgrid.

5. The method for regional autonomous hierarchical control of active distribution network based on master station system according to claim 4 is characterized in that: The autonomous control strategy module uses the power flow calculation method based on the received real-time source and load measurement data to evaluate the supply and demand balance within the microgrid, including: The autonomous control strategy module periodically receives the real-time active and reactive power measurements of the microgrid's key loads, general loads, and controllable loads, the real-time active and reactive output measurements of distributed power sources and energy storage devices, and the flow measurements of the microgrid's connection points with the main grid through the data subscription interface with the monitoring equipment within the microgrid, thus forming a real-time flow measurement data set for each source point, load point, and connection point within the microgrid. The autonomous control strategy module inputs the acquired real-time power flow measurement data set into the pre-established microgrid three-phase power flow model. The measured voltage amplitude and phase angle at each source, load, and connection point are used as the initial values ​​of the state variables, and the Newton-Raphson method is used to perform power flow iterative calculations. After each power flow iteration, the autonomous control strategy module calculates the power imbalance of each node in the microgrid's three-phase power flow model and compares the calculated power imbalance with the convergence threshold. When the power imbalance of each node is less than the convergence threshold, the iterative calculation is stopped and the voltage amplitude and phase angle of each source point, load point, and connection point at the current iteration step are output as the microgrid optimal power flow calculation result. The autonomous control strategy module compares the microgrid optimal power flow calculation results with the preset microgrid supply and demand balance constraints; if the supply and demand balance constraints are met at the same time, it outputs the microgrid supply and demand balance status evaluation result, otherwise it outputs the supply and demand imbalance status evaluation result.

6. The method for regional autonomous hierarchical control of active distribution network based on master station system according to claim 5, characterized in that: Supply and demand balance constraints include: The absolute value of the difference between the total active power generated and the total active power consumed in the microgrid is less than P1% of the rated capacity of the microgrid; The voltage amplitude deviation rate of key load and general load nodes in the microgrid is less than P2%; The voltage amplitude deviation rate of the controllable load nodes in the microgrid is less than P3%.

7. The method for regional autonomous hierarchical control of active distribution network based on master station system according to claim 3 is characterized by: S21 uses the DBSCAN clustering algorithm to group device nodes and outputs the division results of three types of hierarchical control units: microgrid, smart substation, and regional distributed power supply according to cluster labels, including: The master station system extracts the geographical coordinates, load type attributes, and distributed power capacity attributes of the device node based on the established distributed power device model, and constructs a device node feature vector containing the extracted attributes; The master station system inputs the constructed device node feature vector into the DBSCAN clustering algorithm and sets the minimum clustering parameter of the clustering algorithm to 3, corresponding to the three types of hierarchical control units: microgrid, smart substation, and regional distributed power supply; Based on the input device node feature vector, the density peak search algorithm is used to determine the cluster center. The feature vector is classified with the cluster center as the core through the DBSCAN clustering algorithm, and the clustering result containing the cluster label of each device node is output; The master station system identifies the type of hierarchical control unit to which each device node belongs based on the clustering label. It divides the device node with a clustering label of 1 into a microgrid, the device node with a clustering label of 2 into a smart substation, and the device node with a clustering label of 3 into a regional distributed power supply. When the established distributed power supply equipment model changes, the master station system extracts the characteristic attributes of the changed distributed power supply equipment node, updates the characteristic vector of the corresponding equipment node, inputs the updated characteristic vector into the DBSCAN clustering algorithm, repeats the iterative update, receives the updated clustering results, and adjusts the hierarchical control unit division results of the equipment node.

8. The method for regional autonomous hierarchical control of active distribution network based on master station system according to claim 7 is characterized in that: The density peak search algorithm is used to determine the cluster center, including: Calculate the Euclidean distance between the feature vectors of device nodes and generate a node distance matrix; Based on the node distance matrix, calculate the local density of each device node and relative distance ;in, Indicates that the distance to node i is less than the cutoff distance The number of other nodes; represents the minimum distance between node i and the local density node; Based on the calculated local density of device nodes and relative distance , set the density threshold is the mean of the local density of all nodes, and the distance threshold is determined by an iterative method , the number of cluster centers under the current threshold is not less than the preset minimum number of clusters; and The node i is determined as the cluster center; Taking the determined cluster center as the core, the distance between each non-cluster center node and each cluster center is calculated, and the corresponding node is classified into the category of the cluster center closest to it to form a clustering result.

9. The method for regional autonomous hierarchical control of active distribution network based on master station system according to claim 8, characterized in that: S23, coordinated control between control units of the master station system, including: The master station system maps the logical relationship models of the three types of hierarchical control units, namely microgrid, smart substation and regional distributed power supply, with the established distributed power supply equipment models. The object association method is used to establish a one-to-one correspondence between the objects in the logical relationship model and the corresponding physical device entities in the distributed power supply equipment model, and the hierarchical control unit models of microgrid, smart substation and regional distributed power supply are generated by fusion, which include the actual device composition topology and internal logical control structure. Based on the generated hierarchical control unit model, the master station system extracts the monitoring device objects within each hierarchical control unit, establishes a protocol data subscription relationship with the monitoring device, and periodically receives the real-time operation measurement data of the microgrid, smart substation, and regional distributed power source uploaded by the monitoring device; The master station system inputs the received real-time operation measurement data of each hierarchical control unit into the operation status evaluation rules based on threshold values ​​and fuzzy rules predefined in the corresponding hierarchical control unit model, and uses threshold value judgment and fuzzy logic reasoning methods to evaluate the real-time operation status of each hierarchical control unit; Based on the generated hierarchical control unit operation status evaluation results, the master station system combines the topological connection relationship of microgrids, smart substations and regional distributed power sources in the distribution network, uses the power flow calculation method to determine the real-time supply and demand balance status of each hierarchical control unit and generate a power dispatch plan.

Images