Power distribution network regulation and control method, system, equipment, medium and product

Abstract

The invention discloses a power distribution network regulation and control method, system and device, a medium and a product, and belongs to the technical field of power grids, and the method comprises the steps: periodically obtaining real-time electric parameter data in a power distribution network, coding the real-time electric parameter data based on a preset coding sequence, and obtaining an operation state vector; inputting the operation state vector into a fault identification model to obtain a fault type, a fault probability and a fault position; if the fault probability is smaller than or equal to a preset threshold value, obtaining a cooperative control instruction for each adjustable resource based on the operation state vector, the adjustable quantity of the multi-source adjustable resource and a preset multi-target optimization scheduling model, and executing the cooperative control instruction; if the fault probability is larger than a preset threshold value, the hierarchical control instructions are determined according to the fault type and the fault position and are executed in sequence to achieve intelligent regulation and control of the power distribution network, and therefore the operation economical efficiency, stability and power supply reliability of a power system can be improved.

Description

Technical Field

[0001] This invention relates to the field of power grid technology, and in particular to a distribution network control method, system, equipment, medium, and product. Background Technology

[0002] With the large-scale integration of distributed power sources, energy storage, and adjustable loads into low-voltage distribution networks, the power flow distribution, operating characteristics, and fault modes of these networks are becoming increasingly complex. In this complex environment, the regulation and control of the distribution network must simultaneously consider multiple objectives, including economic efficiency, security, and power supply reliability, to support its stable operation.

[0003] Current low-voltage distribution network control methods typically use minimizing resource scheduling costs as the objective function to schedule adjustable resources. However, the actual operating state of the power grid is determined by multiple related key indicators. Existing methods fail to consider the interrelationships between these key indicators in a coordinated manner, which can easily lead to oscillations or instability in the power grid's operating state, resulting in the inability of the power grid to operate safely, economically, and reliably. Summary of the Invention

[0004] This invention provides a distribution network control method, system, equipment, medium, and product that can solve the problem of how to optimize the control of the distribution network for multiple operational objectives.

[0005] This invention provides a distribution network control method, comprising: Based on a preset control cycle, real-time electrical parameter data in the distribution network is periodically acquired, and each real-time electrical parameter data is encoded according to a preset encoding order to obtain an operating state vector. The operating state vector is input into the fault identification model to obtain the identification result, wherein the identification result includes the fault type, fault probability and fault location; If the fault probability is less than or equal to a preset threshold, then based on the operating state vector, the adjustable amount of the multi-source adjustable resources, and the preset multi-objective optimization scheduling model, a cooperative control instruction for each of the adjustable resources is obtained, and the cooperative control instruction is executed. The objective function of the multi-objective optimization scheduling model is used to calculate the minimum sum between the line loss weighting term, the voltage deviation weighting term, the resource adjustment cost weighting term, and the switching operation number weighting term. If the fault probability is greater than the preset threshold, then based on the fault type and the fault location, hierarchical control commands are determined and executed sequentially to achieve intelligent control of the power distribution network.

[0006] This invention unifies dispersed measurements such as voltage, current, and switch status into standardized vectors, providing structured and time-consistent input for subsequent models and avoiding misjudgments or optimization failures caused by inconsistent data formats. It utilizes a machine learning model to perform a three-tiered fault determination—whether a fault exists, what type of fault it is, and where it occurred—improving the speed and accuracy of fault identification. Based on fault probability and threshold judgments, different paths are executed, constructing an intelligent switching mechanism for normal optimization and emergency self-healing modes. This ensures that the distribution network is no longer passively responding or operating in a single mode, but actively selects the most suitable global strategy based on real-time risk assessment, achieving an integrated control strategy of optimization and self-healing. In the first path, a multi-objective optimization scheduling model generates collaborative control commands, achieving comprehensive balancing and collaborative optimization of multiple key operating indicators such as line loss (economy), voltage quality (safety), and equipment operating costs (reliability), thereby enabling the system to operate in an overall optimal state under normal conditions. In the second path, hierarchical control commands are determined and executed sequentially, providing a structured and correctly ordered self-healing operation process. This ensures rapid fault isolation, maximum power restoration, and support for system stability after restoration through resource scheduling, significantly shortening outage time and improving power supply reliability. Overall, this embodiment provides an intelligent distribution network control method. By constructing a unified state perception, probability-based intelligent fault identification, and a decision-making mechanism that dynamically switches between economic optimization and safety self-healing modes, it achieves multi-dimensional, adaptive, and coordinated control of the distribution network operation status. This method can improve operational economy, voltage quality, and power supply reliability while ensuring system safety and stability.

[0007] Furthermore, the hierarchical control instructions sequentially include fault isolation instructions, power restoration instructions, and resource scheduling instructions, and the determination and sequential execution of the hierarchical control instructions specifically involves: Based on the fault location, determine the fault isolation command used to control the closing of the fault section boundary switch; Execute the fault isolation command to determine the non-fault power loss section, and based on the fault location and the topology of the distribution network, reconstruct the power supply path for the non-fault power loss section to obtain the power supply restoration command; Execute the power restoration command and obtain the current operating status, and determine the resource scheduling command based on the current operating status.

[0008] This invention concretizes hierarchical control into three clear and orderly steps: isolation, recovery, and scheduling. It defines a standardized operating procedure for self-healing control, ensuring logical correctness and engineering feasibility. Furthermore, by determining the isolation command based on the fault location, it achieves precise fault localization and minimal-range isolation, minimizing the power outage area and preventing fault escalation. After isolation, non-faulty power loss sections are identified and path reconstruction is performed, clarifying the target objects for power restoration and the means of topology reconstruction. After restoration, resource scheduling commands are determined based on the new state, embodying the idea of ​​combining rigid operation (switching) and flexible adjustment (resources) in the self-healing process. After completing the physical topology switch, controllable resources are used to quickly mitigate potential secondary problems such as overload and voltage exceedances, ensuring the system immediately enters a stable operating state after self-healing and avoiding secondary instability. Overall, this embodiment ensures that the self-healing process starts with the minimum power outage area, restores the maximum power supply range as quickly as possible, and ensures immediate stability after restoration through flexible adjustment, thereby systematically and quantifiably improving the fault self-healing speed and post-restoration operational quality of the distribution network.

[0009] Furthermore, the power distribution network control method further includes: Obtain the operating data of the power distribution network within the historical period; The average power restoration time, line loss reduction rate, and switch operation reduction rate are calculated based on the power distribution network operation data to obtain the evaluation results of the multi-objective optimization scheduling model. The model parameters of the historical multi-objective optimization scheduling model are dynamically adjusted based on the evaluation results to obtain the multi-objective optimization scheduling model, or the historical probability threshold is dynamically adjusted based on the evaluation results to obtain the preset threshold.

[0010] This invention establishes a quantitative evaluation system for control effectiveness by calculating performance evaluation indicators based on operational data. It dynamically adjusts model parameters or probability thresholds according to the evaluation results, automatically correcting the weights of the optimized model or adjusting the sensitivity of fault detection based on historical performance. This allows the entire control system to adapt to changes in the power grid and continuously improve its performance. Overall, through this closed-loop adaptive mechanism, experience is continuously accumulated, and decision parameters are optimized, thereby maintaining a high level of operational economy, safety, and self-healing performance in a long-term and stable manner.

[0011] Furthermore, the electrical parameter data includes voltage data, resistance data, current data, and switch status data; the distribution network control method further includes: For each branch, an initial loss weighting term is calculated based on the resistance data, the current data, and the first weighting coefficient. The initial loss weighting term is then normalized based on a preset capacity value to obtain the loss weighting term. For each node, an initial voltage deviation weighting term is calculated based on the voltage data, a preset voltage value, and a second weighting coefficient. The initial voltage deviation weighting term is then normalized based on the preset voltage value to obtain the loss weighting term. For each adjustable resource, an initial adjustment cost weighting term is calculated based on the adjustable amount, the preset cost coefficient, and the third weighting coefficient. The initial adjustment cost weighting term is then normalized based on the preset cost value to obtain the resource adjustment cost weighting term. For each switch, an initial switch operation count is calculated based on the switch state data and the fourth weighting coefficient. The initial switch operation count is then normalized based on a preset number of operations to obtain the switch operation count.

[0012] This invention describes in detail the composition and calculation of each term in the multi-objective function, clarifying how to transform physically incomparable dimensional indicators (such as power, voltage, currency, and frequency) into dimensionless scalars that can be mathematically compared and optimized. This is a key technical guarantee for the engineering implementation of the multi-objective optimization model. By introducing normalization processing by dividing each term by the corresponding benchmark value (such as capacity benchmark value, voltage benchmark value, etc.), the numerical comparability of each objective term is ensured, making the weighted summation have clear mathematical and physical meaning, thereby obtaining a reasonable and balanced comprehensive optimization solution. Overall, this embodiment ensures the mathematical rigor and engineering practicality of the optimization model, enabling the multi-objective function of co-optimizing line loss, voltage, cost, and switching frequency to be accurately and reliably implemented at the algorithm level.

[0013] Furthermore, the power distribution network control method further includes: The multi-objective optimization scheduling model is solved under the constraints of power balance, node voltage, branch current, and resource adjustment.

[0014] The embodiments of this invention clearly define the four major categories of engineering constraints that must be satisfied when solving the optimization model, ensuring that the optimal solution obtained by the optimization scheduling model is an engineering-executable and feasible solution. This closely integrates theoretical optimization with the actual safe operation of the power grid, ensuring the safety and reliability of the intelligent control system.

[0015] Furthermore, the fault identification model is a machine learning model used to extract the spatiotemporal features of the operating state vector, wherein the machine learning model includes a convolutional model, a long short-term memory model, and a model combined with a gated recurrent unit structure.

[0016] This approach defines the fault identification model as a machine learning model capable of extracting spatiotemporal features. By leveraging the powerful feature extraction capabilities of machine learning models (especially CNN, LSTM, etc.), the model can automatically learn the deep and abstract features of faults from the time-series data of the operating state vector. This significantly improves the accuracy of identifying complex and hidden faults (such as high-resistance grounding and early-stage faults) and enhances early warning capabilities. Consequently, it provides higher-quality information input for subsequent optimization or self-healing decisions, thereby improving the overall system's intelligence level and response efficiency.

[0017] Another embodiment of the present invention provides a power distribution network control system, including: a data acquisition module, a fault identification module, a first control module, and a second control module; The data acquisition module is used to periodically acquire real-time electrical parameter data in the distribution network based on a preset control cycle, and encode each real-time electrical parameter data according to a preset encoding order to obtain an operating state vector. The fault identification module is used to input the operating state vector into the fault identification model to obtain the identification result, wherein the identification result includes fault type, fault probability and fault location; The first control module is used to, if the fault probability is less than or equal to a preset threshold, obtain a coordinated control instruction for each of the adjustable resources based on the operating state vector, the adjustable amount of the multi-source adjustable resources, and a preset multi-objective optimization scheduling model, and execute the coordinated control instruction. The objective function of the multi-objective optimization scheduling model is used to calculate the minimum sum between the line loss weighting term, the voltage deviation weighting term, the resource adjustment cost weighting term, and the switching operation number weighting term. The second control module is used to determine and execute hierarchical control commands according to the fault type and the fault location if the fault probability is greater than the preset threshold, so as to realize intelligent control of the power distribution network.

[0018] Another embodiment of the present invention provides a terminal device, including: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, it implements the steps of the power distribution network control method of the present invention.

[0019] Another embodiment of the present invention provides a computer-readable storage medium item, including: a stored computer program, which, when the computer program is running, controls the device where the computer-readable storage medium is located to perform the steps of the power distribution network control method of the present invention.

[0020] Another embodiment of the present invention provides a computer program product stored in a storage medium, the computer program product being executed by at least one processor to implement the steps of the power distribution network control method as described in any one of the first aspects. Attached Figure Description

[0021] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0022] Figure 1 This is a schematic flowchart of a power distribution network control method provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of a power distribution network control system provided in an embodiment of the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0025] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0026] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0027] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0028] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0029] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0030] See Figure 1 To address the problem of optimizing and controlling a distribution network with multiple operational objectives in existing technologies, an embodiment of the present invention provides a distribution network control method, comprising steps 101-104, specifically including: Step 101: Based on a preset control cycle, periodically acquire real-time electrical parameter data in the distribution network, and encode each real-time electrical parameter data according to a preset encoding order to obtain an operating state vector.

[0031] In this embodiment, operational data from key nodes of the distribution network are periodically collected and organized into a predefined, uniformly formatted data structure to characterize the overall operational status of the system. Specifically, within each preset period, raw data is acquired through an electrical parameter sensor group and a switch actuator. The electrical parameter sensor group is used to collect voltage, current, power, and leakage current signals of low-voltage lines in real time, and the switch actuator is used to perform opening, closing, or flexible load limiting actions within millisecond time scales. The sampling sequence output by the electrical parameter sensor group undergoes data preprocessing operations such as filtering, calibration, and feature extraction. The processed sampling sequence and switch status are encapsulated into an operational state vector x(t) according to a unified encoding rule. For example, the voltage, current, and switch quantities from each node at time t are uniformly arranged in a preset order to form the operational state vector, which is represented as follows: ; Where x(t) is the running state vector, u M (t) is the voltage value of node M at time t, i M (t) is the current value at node M at time t, s L (t) is the switching quantity of switch L at time t.

[0032] Step 102: Input the operating state vector into the fault identification model to obtain the identification result, wherein the identification result includes the fault type, fault probability and fault location.

[0033] In this embodiment, the pre-trained fault identification model automatically identifies normal and various fault events based on the running state vector, and outputs an identification result including fault type, fault probability, and fault location. In the fault identification model, the probability of each type of event occurring at time t is calculated to obtain a probability vector, represented as: ; in, These represent the probabilities of the normal, short circuit, grounding, and open circuit events occurring at time t, respectively.

[0034] Based on the maximum probability in the above probability vector, the recognition result of the model is determined, and the recognition result can be defined as: .

[0035] As an example of an embodiment of the present invention, the fault identification model is a machine learning model for extracting the spatiotemporal features of the operating state vector, wherein the machine learning model includes a convolutional model, a long short-term memory model, and a model combined with a gated recurrent unit structure.

[0036] In this embodiment, different model structures can be selected and configured according to actual needs. For example, in a one-dimensional convolutional neural network (1D-CNN) model, each sampling period of the running state vector x(t) is taken as a feature frame, and data from multiple consecutive periods (time windows) are stacked into a two-dimensional data structure (time step × feature dimension). A one-dimensional convolutional kernel is used to slide along the time axis to automatically extract the local spatiotemporal dependence features (such as sudden rise, sudden fall, and harmonic modes) of voltage, current and other measurements within a short time window. Then, fault identification is performed based on the extracted features to obtain the identification result. This model is suitable for fault types such as short circuits and lightning strikes that have obvious transient characteristics and whose waveforms change drastically in a short time and have high identification sensitivity. For example, in Long Short-Term Memory (LSTM) or Gated Recurrent Unit (GRU) models, the running state vector x(t) is input into the recurrent neural network in chronological order. Utilizing the "gating" mechanism (input gate, forget gate, output gate) of the LSTM or GRU unit, information is selectively memorized and forgotten, thereby learning and extracting the dynamic evolution and deep temporal dependencies of electrical parameter data over long time series. This model is suitable for fault types with relatively slow development processes and strong temporal continuity, such as grounding faults, intermittent arcs, and equipment insulation aging. Alternatively, a combined model, such as the CNN-LSTM model, can be used. This model employs a cascaded structure, first using a 1D-CNN layer to quickly extract local high-level features from the original running state vector. Then, these feature sequences are input into an LSTM layer, which further learns the long-range temporal contextual relationships of these features. This model structure combines the local feature extraction advantages of CNNs with the temporal modeling advantages of LSTMs, making it suitable for comprehensive scenarios with extremely high requirements for recognition accuracy and robustness, and complex fault characteristics. It can simultaneously capture the instantaneous characteristics and development process of faults.

[0037] Step 103: If the fault probability is less than or equal to a preset threshold, then based on the operating state vector, the adjustable amount of the multi-source adjustable resources, and the preset multi-objective optimization scheduling model, a collaborative control instruction for each of the adjustable resources is obtained, and the collaborative control instruction is executed. The objective function of the multi-objective optimization scheduling model is used to calculate the minimum sum among the line loss weighting term, voltage deviation weighting term, resource adjustment cost weighting term, and switching operation number weighting term.

[0038] Adjustable resources include distributed power sources, energy storage resources, and adjustable load resources.

[0039] In this embodiment, if the fault probability is less than or equal to a preset threshold, the power grid is determined to be in a low-risk operating state, and an economic collaborative optimization process is initiated. The core of this process is based on a preset decision model to uniformly schedule controllable resources in the power grid to optimize overall operational efficiency. Specifically, the power grid structure is abstracted into a set of nodes and a set of branches, and distributed power sources, energy storage devices, and adjustable loads are defined as a set of schedulable resources. An objective function is constructed that can simultaneously quantify four key operational indicators: line loss level, node voltage deviation, resource regulation cost, and the number of critical switching operations. This function assigns weight coefficients to each indicator and sums them into a single comprehensive cost indicator; the optimization objective is to minimize this comprehensive cost. To ensure the safety and feasibility of the optimization results, the model can be restricted to solve under preset physical and engineering constraints. For example, power balance constraints are used to ensure the net power balance between all controllable resources and uncontrollable loads and power sources after regulation, or operational constraints are used to limit the voltage of each node and the current of each branch within a preset safety range. The real-time measured values ​​of the voltage and branch current of each node are extracted from the current operating state vector x(t) and used as the state input of the model. At the same time, the current output and adjustable margin reported by each schedulable resource are summarized. The constrained multi-objective optimization problem defined above is solved by using optimization algorithms (such as particle swarm optimization, genetic algorithm, etc.). The solution process is to search for the combination of decision variables that minimizes the comprehensive objective function value in the solution space that satisfies all constraints. This combination includes the target output (or adjustment amount) of each schedulable resource in the next control cycle. The optimal combination of decision variables obtained by the solution is converted into a specific set of coordinated control instructions that can be identified by each resource terminal. This set of instructions is sent to the corresponding distributed power inverter, energy storage management system and load control terminal through the communication network.

[0040] Step 104: If the fault probability is greater than the preset threshold, then based on the fault type and the fault location, determine the hierarchical control instructions and execute them sequentially to achieve intelligent control of the power distribution network.

[0041] In this embodiment, when the fault probability exceeds a preset threshold, the distribution network is determined to be in a high-probability fault state. A preset self-healing control process, with the core objectives of fault isolation and rapid power restoration, is immediately initiated. This process employs a hierarchical, progressive instruction sequence, prioritizing safety and maximizing power restoration. After the self-healing control is completed and the power grid stabilizes, the updated operating state vector is re-input into the fault identification model until the fault probability falls below the preset threshold. Then, the system automatically switches back to the economic optimization mode of step 103 for power grid regulation.

[0042] It should be noted that, in this embodiment of the invention, the fault identification model is encapsulated with a unified interface, and the encapsulated model is embedded as a functional module into the entire chain of data acquisition-event identification-decision control, and is integrated with the resource collaborative optimization in step 103 and the self-healing control process in step 104.

[0043] As an example of an embodiment of the present invention, the hierarchical control instructions sequentially include fault isolation instructions, power restoration instructions, and resource scheduling instructions. The process of determining and executing the hierarchical control instructions sequentially specifically involves: determining the fault isolation instruction for controlling the closing of the boundary switch of the fault section based on the fault location; executing the fault isolation instruction to determine the non-fault power loss section; reconstructing the power supply path for the non-fault power loss section based on the fault location and the topology of the distribution network to obtain the power restoration instruction; executing the power restoration instruction and obtaining the current operating status; and determining the resource scheduling instruction based on the current operating status.

[0044] In this embodiment, based on the fault type and location information output by the fault identification model and combined with the pre-stored distribution network topology map, the faulty section (i.e., the smallest fault range) is automatically located. Control commands are generated and immediately issued to drive the smart switches at the boundary of the faulty section to perform a millisecond-level "opening" operation, physically isolating the section from the power grid and preventing the fault's impact from spreading. After the fault is effectively isolated, the non-faulty upstream or downstream areas that experienced power outages due to the isolation operation are automatically identified. Based on the power grid topology and pre-set tie / section switch information, the optimal backup power supply path is automatically calculated and selected. Control commands are generated and issued to drive the relevant tie switches or section switches on the backup path to perform a "closing" operation, reconstructing the power grid topology and restoring power to the non-faulty sections that lost power. After completing the switching operations for fault isolation and power restoration, a rapid assessment is performed based on the reconstructed new grid topology and real-time operating status (reflected by the updated operating status vector). If the assessment finds risks such as local overload or voltage exceeding limits in the new operating status, a second type of control command is quickly generated. This type of command aims to adjust one or more schedulable resources (such as commanding nearby energy storage to discharge, adjusting the output of distributed power sources, or reducing some adjustable loads) to provide power support or reduce loads, ensuring instantaneous stable operation of the grid after power restoration.

[0045] As an example of an embodiment of the present invention, the distribution network control method further includes: acquiring the operation data of the distribution network in a historical period; calculating the average restoration time, line loss reduction rate, and switch operation reduction rate based on the distribution network operation data to obtain the evaluation result of the multi-objective optimization scheduling model; dynamically adjusting the model parameters of the historical multi-objective optimization scheduling model based on the evaluation result to obtain the multi-objective optimization scheduling model, or dynamically adjusting the historical probability threshold based on the evaluation result to obtain the preset threshold.

[0046] In this embodiment, relevant historical operating data of the distribution network within a historical period (such as the past 24 hours, week, or month) can be calculated. This data includes fault event logs, operating indicator time series, and historical control commands. The fault event logs include timestamps for each fault event (e.g., the time of fault occurrence and termination), fault type, fault location, self-healing control activation time, and the time of complete power restoration. The operating indicator time series includes key operating indicators of the distribution network within the historical period, such as total active power loss of the total feeder / transformer area (used to calculate line loss), node voltage amplitude, and the operation counter readings of all key switches. Historical control commands include all issued coordinated control commands and hierarchical control commands and their timestamps. The average power restoration time is calculated based on the historical operating data. Line loss reduction rate and the rate of reduction in the number of switching operations The multi-objective optimization scheduling model is evaluated, and its calculation methods are as follows: ; ; ; in, The number of failure events within the historical period. and These represent the start time of the power outage and the completion time of the power restoration for the k-th fault, respectively. and These represent the line loss power (or line loss energy) under the historical period and the current period, respectively. and These represent the total number of key switch operations in the historical and current periods, respectively. Based on the calculated evaluation results, the system employs a closed-loop feedback mechanism to automatically adjust core parameters and achieve self-optimization. This includes adjusting the multi-objective optimization scheduling model and the fault identification probability threshold. Specifically, the adjustment of the multi-objective optimization scheduling model includes weight coefficients in the objective function (such as line loss weight coefficient, voltage deviation weight coefficient, and switch operation count weight coefficient). For example, if the rate of reduction in switch operation count is negative (increased switch operations), the switch operation count weight coefficient is increased to increase the penalty for switch operations in subsequent optimizations, prompting the strategy to favor resource adjustment rather than switch actions. Another example is that if the average restoration time exceeds the preset restoration time threshold, the fault location accuracy can be improved by adjusting model parameters, thereby reducing the restoration time.

[0047] As an example of an embodiment of the present invention, the electrical parameter data includes voltage data, resistance data, current data, and switch status data. The distribution network control method further includes: for each branch, calculating an initial loss weighting term based on the resistance data, the current data, and a first weighting coefficient, and normalizing the initial loss weighting term based on a preset capacity value to obtain the loss weighting term; for each node, calculating an initial voltage deviation weighting term based on the voltage data, a preset voltage value, and a second weighting coefficient, and normalizing the initial voltage deviation weighting term based on a preset voltage value to obtain the loss weighting term; for each adjustable resource, calculating an initial adjustment cost weighting term based on the adjustable amount, a preset cost coefficient, and a third weighting coefficient, and normalizing the initial adjustment cost weighting term based on a preset cost value to obtain the resource adjustment cost weighting term; for each switch, calculating an initial switch operation count term based on the switch status data and a fourth weighting coefficient, and normalizing the initial switch operation count term based on a preset operation count to obtain the switch operation count term.

[0048] In this embodiment, the multi-objective optimization scheduling model transforms four key operational indicators with different physical dimensions into dimensionless scalars that can be weighted and summed through initial weighting term calculation and normalization. The objective function can be expressed as follows: ; in, and Let be the resistance and current of branch (i,j), respectively. Let k be the voltage magnitude at node k. Rated voltage, Let m be the adjustment amount for the m-th controllable resource. To correspond to the adjustment cost coefficient, ns represents the number of operations of the s-th critical switch during the optimization period. , , , n is a weighting factor for line loss, voltage deviation, regulation cost, and number of switching operations. s This represents the number of times switch s is closed and opened within the scheduling cycle. , , , It consists of a set of branches, a set of nodes, a set of controllable resources, and a set of key switches.

[0049] To ensure dimensional consistency when performing the weighted summation of the above four terms, preset capacity values ​​are used in the actual optimization solution. Preset voltage value Preset cost value and preset number of operations To each , , , After normalization, the multi-objective optimization algorithm described above internally calculates the corresponding dimensionless index. and The objective function in the above equation is only symbolically represented and the normalization factor is not explicitly given. The normalization process is represented as follows: ; ; ; ; In the objective function of this embodiment, the branch current and node voltage The real-time measured values ​​in the operating state vector x(t) uploaded by each low-voltage intelligent switchgear are obtained directly through power flow calculation or equivalent network model; controllable resource adjustment quantity Calculated based on the current output and adjustable margin reported by distributed power sources, energy storage, and adjustable load terminals; number of switching operations n s The action counter inside the switch actuator accumulates and counts data within a preset time window, and uploads it along with the status data.

[0050] As an example of an embodiment of the present invention, the power distribution network control method further includes: solving the multi-objective optimization scheduling model under power balance constraints, node voltage constraints, branch current constraints, and resource regulation constraints.

[0051] In this embodiment, within a given preset control period, the optimization problem corresponding to the objective function must satisfy power balance constraints, node voltage constraints, branch current thermal stability constraints, and capacity and ramp-up rate constraints of various controllable resources to ensure that the optimization results are feasible in engineering. Specifically, in the power balance constraints, the baseline active power output of the m-th controllable resource before adjustment is denoted as... The corresponding adjustment amount is After adjustment, the active power output is Let the total active power demand of the non-adjustable load within the control area be . The total active power output of non-adjustable distributed power sources is Then, the active power balance constraint during the scheduling period can be written as: ; In node voltage constraints, the method used to constrain the voltage amplitude of each node within a preset voltage range is represented as: ; Branch current constraint, used to constrain the regulation of each branch current within its thermal stability limit, is expressed as: ; The controllable resource capacity and ramp rate constraints, used to constrain the adjustment amount of the m-th controllable resource to meet the upper and lower limits of output and the single adjustment step size requirements, are expressed as: ; in, This is the adjustment amount from the previous scheduling time. This represents the maximum allowable ramp rate for the m-th controllable resource between adjacent scheduling times.

[0052] By incorporating the output of distributed power sources, the charging and discharging power of energy storage, and the adjustable load reduction into the same optimization framework through the above objective function and constraints, instead of controlling them separately, we can overcome the problem in the existing technology that the overall operating indicators are difficult to optimize in a coordinated manner due to the step-by-step and object-by-object tuning of various resources, and the tendency to have "local optimum and global suboptimal".

[0053] like Figure 2 As shown, based on the above-mentioned method embodiments, an embodiment of the present invention provides a power distribution network control system 200, including: a data acquisition module 201, a fault identification module 202, a first control module 203, and a second control module 204; The data acquisition module 201 is used to periodically acquire real-time electrical parameter data in the distribution network based on a preset control cycle, and encode each real-time electrical parameter data according to a preset encoding order to obtain an operating state vector. The fault identification module 202 is used to input the operating state vector into the fault identification model to obtain the identification result, wherein the identification result includes fault type, fault probability and fault location; The first control module 203 is used to, if the fault probability is less than or equal to a preset threshold, obtain a coordinated control instruction for each of the adjustable resources based on the operating state vector, the adjustable amount of the multi-source adjustable resources, and a preset multi-objective optimization scheduling model, and execute the coordinated control instruction. The objective function of the multi-objective optimization scheduling model is used to calculate the minimum sum between the line loss weighting term, the voltage deviation weighting term, the resource adjustment cost weighting term, and the switching operation number weighting term. The second control module 204 is used to determine and execute hierarchical control commands according to the fault type and the fault location if the fault probability is greater than the preset threshold, so as to realize intelligent control of the power distribution network.

[0054] It is understood that the above system embodiments correspond to the method embodiments of the present invention, and can implement the power distribution network control method provided by any of the above method embodiments of the present invention.

[0055] It should be noted that the system embodiments described above are merely illustrative, and some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the system embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can be specifically implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0056] For ease of description and brevity, the system embodiments of the present invention include all the implementation methods described in the above-described power distribution network control method embodiments, and will not be repeated here.

[0057] Based on the above embodiments of the power distribution network control method, another embodiment of the present invention provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the power distribution network control method of any embodiment of the present invention.

[0058] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the terminal device.

[0059] The terminal device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0060] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting all parts of the terminal device via various interfaces and lines.

[0061] Based on the above-described method embodiments, another embodiment of the present invention provides a computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the power distribution network control method described in any of the above-described method embodiments of the present invention.

[0062] Based on the above-described method embodiments, this invention also provides a computer program / program product, which is stored in a storage medium and executed by at least one processor to implement the various processes of any of the above-described method embodiments, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0063] The modules / units integrated in the device / terminal equipment, if implemented as software functional units and sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0064] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A distribution network control method, characterized in that, include: Based on a preset control cycle, real-time electrical parameter data in the distribution network is periodically acquired, and each real-time electrical parameter data is encoded according to a preset encoding order to obtain an operating state vector. The operating state vector is input into the fault identification model to obtain the identification result, wherein the identification result includes the fault type, fault probability and fault location; If the fault probability is less than or equal to a preset threshold, then based on the operating state vector, the adjustable amount of the multi-source adjustable resources, and the preset multi-objective optimization scheduling model, a cooperative control instruction for each of the adjustable resources is obtained, and the cooperative control instruction is executed. The objective function of the multi-objective optimization scheduling model is used to calculate the minimum sum between the line loss weighting term, the voltage deviation weighting term, the resource adjustment cost weighting term, and the switching operation number weighting term. If the fault probability is greater than the preset threshold, then based on the fault type and the fault location, hierarchical control commands are determined and executed sequentially to achieve intelligent control of the power distribution network.

2. The power distribution network control method as described in claim 1, characterized in that, in, The hierarchical control commands sequentially include fault isolation commands, power restoration commands, and resource scheduling commands. The specific steps for determining and executing the hierarchical control commands sequentially are as follows: Based on the fault location, determine the fault isolation command used to control the closing of the fault section boundary switch; Execute the fault isolation command to determine the non-fault power loss section, and based on the fault location and the topology of the distribution network, reconstruct the power supply path for the non-fault power loss section to obtain the power supply restoration command; Execute the power restoration command and obtain the current operating status, and determine the resource scheduling command based on the current operating status.

3. The power distribution network control method as described in claim 1, characterized in that, The power distribution network control method further includes: Obtain the operating data of the power distribution network within the historical period; The average power restoration time, line loss reduction rate, and switch operation reduction rate are calculated based on the operational data to obtain the evaluation results of the historical multi-objective optimization scheduling model. The model parameters of the historical multi-objective optimization scheduling model are dynamically adjusted based on the evaluation results to obtain the multi-objective optimization scheduling model, or the historical probability threshold is dynamically adjusted based on the evaluation results to obtain the preset threshold.

4. The power distribution network control method as described in claim 1, characterized in that, in, The electrical parameter data includes voltage data, resistance data, current data, and switch status data. The distribution network control method further includes: For each branch, an initial loss weighting term is calculated based on the resistance data, the current data, and the first weighting coefficient. The initial loss weighting term is then normalized based on a preset capacity value to obtain the loss weighting term. For each node, an initial voltage deviation weighting term is calculated based on the voltage data, a preset voltage value, and a second weighting coefficient. The initial voltage deviation weighting term is then normalized based on the preset voltage value to obtain the loss weighting term. For each adjustable resource, an initial adjustment cost weighting term is calculated based on the adjustable amount, the preset cost coefficient, and the third weighting coefficient. The initial adjustment cost weighting term is then normalized based on the preset cost value to obtain the resource adjustment cost weighting term. For each switch, based on the switch state data and the fourth weighting coefficient, an initial switch operation count is calculated, and the initial switch operation count is normalized based on a preset number of operations to obtain the switch operation count.

5. The power distribution network control method as described in claim 1, characterized in that, The power distribution network control method further includes: The multi-objective optimization scheduling model is solved under the constraints of power balance, node voltage, branch current, and resource adjustment.

6. The power distribution network control method as described in claim 1, characterized in that, The fault identification model is a machine learning model used to extract the spatiotemporal features of the operating state vector, wherein the machine learning model includes a convolutional model, a long short-term memory model, and a model combined with a gated recurrent unit structure.

7. A power distribution network control system, characterized in that, include: Data acquisition module, fault identification module, first control module, and second control module; The data acquisition module is used to periodically acquire real-time electrical parameter data in the distribution network based on a preset control cycle, and encode each real-time electrical parameter data according to a preset encoding order to obtain an operating state vector. The fault identification module is used to input the operating state vector into the fault identification model to obtain the identification result, wherein the identification result includes fault type, fault probability and fault location; The first control module is used to, if the fault probability is less than or equal to a preset threshold, obtain a coordinated control instruction for each of the adjustable resources based on the operating state vector, the adjustable amount of the multi-source adjustable resources, and a preset multi-objective optimization scheduling model, and execute the coordinated control instruction. The objective function of the multi-objective optimization scheduling model is used to calculate the minimum sum between the line loss weighting term, the voltage deviation weighting term, the resource adjustment cost weighting term, and the switching operation number weighting term. The second control module is used to determine and execute hierarchical control commands according to the fault type and the fault location if the fault probability is greater than the preset threshold, so as to realize intelligent control of the power distribution network.

8. A terminal device, characterized in that, The method includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, it implements the power distribution network control method as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, include: A stored computer program, wherein, when the computer program is executed, it controls the device containing the computer-readable storage medium to perform the power distribution network control method as described in any one of claims 1-6.

10. A computer program product, characterized in that, include: Computer instructions, when executed by a processor, implement the steps in the power distribution network control method as described in any one of claims 1 to 6.

Images