A power distribution network power quality control method

Abstract

The application relates to the technical field of power quality monitoring and treatment, and discloses a power quality control method for a power distribution network and computer equipment. The method comprises the following steps: acquiring real-time operation data of the power distribution network and original power quality signals of a disturbance node, decomposing the signals and determining an optimal decomposition modal number with the minimum total variance as the target; determining a key time sequence feature set based on the optimal decomposition modal number and a preset disturbance type label; reasoning a disturbance propagation path from the disturbance node as the starting point in the topology structure of the power distribution network, and determining the voltage change rate of each node; screening candidate measures based on the voltage change rate, the feature set and a threshold value, combining a preset evaluation index to construct a judgment matrix; determining a priority weight based on the judgment matrix to determine optimal treatment equipment; and determining dynamic output parameters according to index measured values, reference values and real-time data, and controlling the optimal treatment equipment to operate according to the dynamic output parameters. The application can improve the accuracy of controlling power quality disturbances of the power distribution network.

Description

Technical Field

[0001] This application relates to the field of power quality monitoring and control technology, specifically to a power quality control method and computer equipment for power distribution networks. Background Technology

[0002] With the widespread application of power electronics technology, power quality problems in distribution networks are exhibiting new characteristics such as strong volatility, complex spectrum, and variable propagation paths. To ensure the high-quality and stable operation of the power grid, precise control and regulation of power quality disturbances has become a key focus of the industry.

[0003] In existing technological systems, the sensing and analysis of power quality disturbances typically employ conventional signal processing algorithms. These methods are mostly based on fixed mathematical basis functions, making it difficult to adaptively separate effective characteristic components when facing non-stationary or transiently distorted grid signals, leading to biases in the analysis of disturbance characteristics. Furthermore, existing monitoring methods are often limited to independent observation of single nodes, failing to fully consider the distribution network topology and the propagation patterns of disturbances at the network level, making it difficult to accurately locate the source and affected area of ​​disturbances.

[0004] When determining specific regulation measures or selecting compensation devices, existing technologies often rely on manual experience or make rough selections based solely on a single cost indicator. This selection method lacks a comprehensive quantitative evaluation of multiple factors such as technical indicators, economic input, and site environment, resulting in a low degree of matching between the selected regulation methods and the actual power grid operating conditions. Furthermore, in the equipment operation phase, a control mode with preset fixed parameters is typically used, lacking dynamic adaptive adjustment capabilities based on real-time operating data. When the power grid load undergoes transient changes or operating conditions switch, this static control method is prone to regulation lag or large steady-state errors. Therefore, existing technologies struggle to achieve high-precision control of power quality disturbances in distribution networks. Summary of the Invention

[0005] The purpose of this application is to provide a power quality control method and computer equipment for power distribution networks, so as to solve the problem of low accuracy in controlling power quality disturbances in existing technologies.

[0006] To achieve the above objectives, the first aspect of this application provides a power quality control method for a distribution network, the method comprising:

[0007] Acquire real-time operating data of the distribution network and raw power quality signals of disturbed nodes, and decompose the raw power quality signals into multiple modal components;

[0008] The variance of each modal component is determined based on the original power quality signal, and the optimal number of decomposed modes is determined with the goal of minimizing the total variance of all modal components.

[0009] The key time-series feature set is determined based on the optimal number of decomposed modes, the original power quality signal, the preset disturbance type label, and the preset feature correlation threshold.

[0010] In the topology of the distribution network, starting from the disturbance node, the disturbance propagation path is obtained by reasoning according to the preset constraints; the disturbance node is any node in the topology.

[0011] Determine the rate of voltage change at each node in the disturbance propagation path;

[0012] Based on voltage change rate, key time series feature set and preset voltage change rate threshold, multiple candidate measures are selected from the preset governance measures library, and a judgment matrix is ​​constructed by combining preset evaluation indicators;

[0013] The priority weights of each candidate measure are determined based on the judgment matrix and preset evaluation indicators, and the optimal governance equipment in the distribution network is determined from the measure with the highest priority weight.

[0014] Dynamic output parameters are determined based on the measured and reference values ​​of the power quality indicators of the optimal treatment equipment, as well as real-time operating data.

[0015] The optimal treatment equipment is controlled to operate according to dynamic output parameters.

[0016] In this embodiment, the step of determining the key time series feature set based on the optimal decomposition mode number, the original power quality signal, the preset disturbance type label, and the preset feature correlation threshold includes: decomposing the original power quality signal according to the optimal decomposition mode number to obtain multiple stationary mode components; extracting high-dimensional time series feature sets from each stationary mode component; and determining the key time series feature set based on the high-dimensional time series feature set, the preset disturbance type label, and the preset feature correlation threshold.

[0017] In this embodiment, determining the key time series feature set based on the high-dimensional time series feature set, the preset perturbation type label, and the preset feature correlation threshold includes: determining the information entropy of each high-dimensional time series feature in the high-dimensional time series feature set; determining the joint entropy of each high-dimensional time series feature with the preset perturbation type label; determining the mutual information entropy of each high-dimensional time series feature with the preset perturbation type label based on the information entropy of each high-dimensional time series feature, the joint entropy, and the information entropy of the preset perturbation type label; and selecting the key time series feature set from each high-dimensional time series feature based on the mutual information entropy and the preset feature correlation threshold.

[0018] In this embodiment of the application, the steps of selecting multiple candidate measures from a preset governance measure library based on voltage change rate, key time series feature set and preset voltage change rate threshold, and constructing a judgment matrix in combination with preset evaluation indicators include: determining the disturbance impact level of each node based on the voltage change rate of each node and preset voltage change rate threshold; determining the disturbance type based on key time series feature set; selecting multiple candidate measures from the preset governance measure library based on disturbance type and disturbance impact level, and constructing a judgment matrix in combination with preset evaluation indicators.

[0019] In this embodiment, determining the variance of each modal component based on the original power quality signal, and aiming to minimize the total variance of all modal components, determines the optimal number of decomposed modes by: determining the variance of each modal component according to the following formula:

[0020]

[0021] The total variance is determined based on the variance of each modal component, and the optimal number of decomposed modes is determined with the goal of minimizing the total variance.

[0022] in, For the first The variance of each modal component; The length of the original power quality signal; For the first Each modal component in The value at time; For the first The mean of each modal component.

[0023] In this embodiment of the application, the total variance is determined based on the variance of each modal component. Determining the optimal number of decomposed modes with the goal of minimizing the total variance includes: determining the optimal number of decomposed modes according to the following formula:

[0024]

[0025] in, The optimal decomposition mode number; This represents the total number of modal components.

[0026] In this embodiment of the application, the preset constraint conditions satisfy the following formula:

[0027]

[0028] in, For the first node directly connected to the disturbance node Voltage amplitude of each candidate node; The voltage amplitude at the disturbed node; For the perturbation node and the first The line current amplitude between candidate nodes; For the perturbation node and the first Line impedance between candidate nodes; This is the preset maximum voltage drop.

[0029] In this embodiment of the application, determining the priority weight of each candidate measure based on the judgment matrix and preset evaluation indicators includes: determining the priority weight according to the following formula:

[0030]

[0031] in, For the first The priority weight of each candidate measure; To determine the first element in the matrix The candidate measures in the first Quantitative evaluation value or membership degree under a preset evaluation index; To determine the first element in the matrix The candidate measures in the first Quantitative evaluation value or membership degree under a preset evaluation index; The total number of multiple candidate measures; The number of pre-set evaluation indicators.

[0032] In this embodiment of the application, determining the dynamic output parameters based on the measured values ​​and reference values ​​of the power quality indicators of the optimal treatment equipment, as well as real-time operating data, includes: determining the power quality indicator deviation of the optimal treatment equipment according to the following formula:

[0033]

[0034] in, for Dynamic output parameters at any given time; for The deviation of power quality indicators at any given time is obtained based on the measured values ​​and reference values ​​of power quality indicators. , and These are PID parameters that are self-tuned based on real-time operating data. It is the integral variable.

[0035] A second aspect of this application provides a computer device, including: a memory configured to store instructions; and a processor configured to retrieve instructions from the memory and to implement the above-described method when executing the instructions.

[0036] The above technical solution firstly utilizes the variance minimization objective to determine the optimal number of decomposed modes, achieving adaptive and refined decomposition of the original power quality signal. This avoids feature extraction bias caused by signal mode aliasing or noise interference at the source, establishing the data accuracy for disturbance identification. Secondly, by combining the distribution network topology and node voltage change rate to deduce the disturbance propagation path, the perception dimension of disturbance is expanded from isolated monitoring of a single node to full-domain path analysis on the network side, accurately defining the spatial distribution and impact level of the disturbance, ensuring the spatial accuracy of the governance scope. Furthermore, by constructing a judgment system containing multi-dimensional evaluation indicators... The matrix-based approach enables the scientific and objective identification of optimal governance equipment from a vast array of measures, avoiding the blindness of experience-based decision-making and ensuring a high degree of matching between governance methods and specific disturbance conditions, thereby improving decision-making accuracy. Finally, in the control execution phase, a dynamic parameter adjustment mechanism is introduced based on the deviation between the measured and reference values ​​of power quality indicators, constructing a real-time closed-loop control logic of "monitoring-feedback-correction." This logic can fine-tune output parameters in real time according to the transient changes in the grid's operating state, effectively eliminating the steady-state error and response lag present in traditional open-loop control, thus significantly improving the accuracy of controlling power quality disturbances in the distribution network across the entire chain.

[0037] Other features and advantages of the embodiments of this application will be described in detail in the following detailed description section. Attached Figure Description

[0038] The accompanying drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the following detailed description to explain the embodiments of this application, but do not constitute a limitation on the embodiments of this application. In the drawings:

[0039] Figure 1 A flowchart illustrating a power quality control method for a power distribution network according to an embodiment of this application is shown schematically.

[0040] Figure 2 The schematic diagram illustrates a structural diagram of a computer device according to an embodiment of this application. Detailed Implementation

[0041] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only for illustration and explanation of the embodiments of this application and are not intended to limit the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0042] It should be noted that if the embodiments of this application involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0043] Furthermore, if the embodiments of this application involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed in this application.

[0044] The acquisition, transmission, storage, use, and processing of data in this application comply with relevant laws and regulations. Furthermore, it should be noted that certain software, components, models, and other existing industry solutions may be mentioned in the embodiments of this application. These should be considered exemplary, intended only to illustrate the feasibility of implementing the technical solution of this application, and do not imply that the applicant has already used or necessarily used such solutions.

[0045] It should be noted that all data involved in this application (including but not limited to data used for analysis, data stored, data displayed, etc.) are information and data that have been fully authorized by all parties, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0046] Figure 1 A flowchart illustrating a power quality control method for a distribution network according to an embodiment of this application is shown schematically. Figure 1 As shown in the figure, this application provides a power quality control method for a power distribution network, which may include the following steps.

[0047] Step 101: Obtain real-time operating data of the distribution network and raw power quality signals of the disturbance nodes, and decompose the raw power quality signals into multiple modal components.

[0048] In this embodiment, the distribution network refers to a power network system used for distributing electrical energy, encompassing lines and transformers and switching equipment of different voltage levels, and is the physical carrier for power quality management. Real-time operational data refers to a set of electrical parameters characterizing the current physical operating state of the distribution network, including but not limited to voltage amplitude, current magnitude, power distribution, and load conditions, used to reflect the macroscopic operating scenario of the power grid. A disturbance node refers to a specific location in the distribution network topology where abnormal power quality indicators are detected or where special attention is needed; it can be any physical node in the topology or the location of a monitoring terminal. The raw power quality signal refers to the voltage or current time-series waveform data directly collected at the disturbance node by monitoring equipment. This data typically exhibits non-stationary characteristics and is mixed with fundamental, various transient or steady-state disturbance components, and random background noise. Modal components refer to the sub-signal sequences corresponding to different center frequencies or time scales obtained after processing the complex raw power quality signal through mathematical decomposition algorithms (such as variational mode decomposition). As the basic building blocks in the signal decomposition process, they are the direct objects for subsequent variance calculation to evaluate the decomposition effect. By acquiring the above data and performing preliminary signal decomposition, the complex multi-component signal mixed in the original waveform can be transformed into multi-dimensional sub-signals, thereby effectively reducing the masking effect of background noise on the effective signal and improving the signal-to-noise ratio and feature analysis accuracy of power quality disturbance signals.

[0049] Step 102: Determine the variance of each modal component based on the original power quality signal, and determine the optimal number of decomposed modes with the goal of minimizing the total variance of all modal components.

[0050] In this embodiment, variance refers to a statistical parameter used to measure the degree of fluctuation or dispersion of each modal component in the time domain. Its value reflects the energy concentration or signal stability contained in the component, specifically the average of the squares of the differences between the values ​​of each sampling point and the mean. Total variance refers to the sum of the variances of all decomposed modal components, used to evaluate the compactness or stability of the decomposition result as a whole. The optimal number of decomposed modes refers to the number of decomposition layers that minimizes the total variance. It represents the optimal number of sub-signals to divide the original signal into and is a core variable controlling the granularity of signal decomposition. By calculating the variance of each component and optimizing with minimizing the total variance as the constraint objective, the subjectivity and uncertainty of manually setting the number of decomposition layers can be overcome. This achieves adaptive and precise matching of decomposition parameters, ensuring that the decomposed signal components are statistically the most stable and compact, thereby effectively avoiding modal aliasing or the generation of spurious components and significantly improving the accuracy and reliability of power quality signal feature analysis.

[0051] Step 103: Determine the key time series feature set based on the optimal decomposition mode number, the original power quality signal, the preset disturbance type label, and the preset feature correlation threshold.

[0052] In this embodiment, the preset disturbance type label refers to a predefined classification identifier or category code stored in the system to distinguish different power quality anomaly modes. It covers various typical distribution network disturbance forms such as voltage sags, voltage swells, harmonic distortion, voltage flicker, and transient pulses, serving as a reference benchmark for evaluating the effectiveness of supervisory features. The preset feature correlation threshold refers to a numerical judgment standard or cutoff threshold set during the feature screening process to measure the strength of the correlation between a feature variable and the target disturbance category. This value is used to distinguish between effective strongly correlated features and ineffective weakly correlated redundant information in correlation measurement calculations such as mutual information entropy. The key time-series feature set refers to a set of data attributes that, after calculation and screening from the high-dimensional feature space, possess high sensitivity and high discriminative power quality disturbance identification capabilities. It eliminates noise interference and redundant terms from the analysis process, containing only core indicator vectors that can significantly characterize the essential physical laws of the disturbance. By refining the signal based on adaptively determined decomposition parameters and combining correlation threshold constraints to reduce the dimensionality and optimize the extracted high-dimensional features, the problem of high feature redundancy and strong noise masking effect in traditional methods can be effectively overcome. This method can accurately lock the most representative disturbance features and significantly improve the signal-to-noise ratio and accuracy of power quality disturbance signal feature extraction.

[0053] Step 104: In the topology of the distribution network, starting from the disturbance node, reason according to the preset constraints to obtain the disturbance propagation path; the disturbance node is any node in the topology.

[0054] In this embodiment, the topology refers to the architectural diagram or data model describing the physical connections and power transmission paths between electrical devices and lines in the distribution network, digitally mapping the node distribution and branch connection status of the power grid. Preset constraints refer to physical criteria or logical rules set according to electrical engineering principles to determine whether power quality disturbances can be effectively transmitted between candidate nodes. These include, but are not limited to, voltage drop threshold limits calculated based on line impedance and current, ensuring that path identification conforms to actual physical laws. The disturbance propagation path refers to the sequence of nodes or branch sets that the disturbance energy or abnormal electrical characteristics traverses from the source node along physical lines, intuitively representing the scope and direction of the disturbance's influence in the power grid space. Through path reasoning based on physical topology and electrical constraints, the dynamic diffusion trajectory of disturbances can be accurately reproduced from the complex network structure, thereby accurately delineating the specific areas and equipment affected by the disturbance, significantly improving the analytical accuracy of the propagation laws and influence range of power quality disturbances in the distribution network.

[0055] In one embodiment, firstly, propagation constraint rules are constructed. Based on the distribution network's topology data, line impedance parameters, and real-time power flow distribution, a set of propagation constraint rules is constructed to define the physical boundaries of disturbance propagation. These rules specifically include: voltage drop constraints to limit voltage attenuation, power transmission constraints to determine the direction and feasibility of energy flow, and topology connectivity constraints to confirm the validity of physical connections. Secondly, propagation path reasoning is performed. Starting from the identified initial disturbance monitoring point (i.e., the disturbance node), and considering the dynamic changes in real-time electrical parameters such as voltage, current, and power at each node in the distribution network, a breadth-first search (BFS) algorithm is used to search for paths. During the search, reasoning is strictly performed along nodes and lines that satisfy the propagation constraint rules, thereby determining the main propagation paths of disturbance energy or abnormal characteristics in the distribution network. Finally, an impact range assessment is conducted. Based on the variation amplitude of electrical parameters of each node along the propagation path and the tolerance threshold of various devices along the path, the impact of the disturbance on the power grid is divided into different levels such as slight impact, moderate impact, and severe impact. Based on this, the specific node range corresponding to each impact level is determined, thereby completing the accurate definition of the scope of the disturbance's impact.

[0056] Step 105: Determine the voltage change rate of each node in the disturbance propagation path.

[0057] In one embodiment, the rate of voltage change is determined to satisfy the following formula:

[0058]

[0059] in, The first node connected to the starting node in the disturbance propagation path The reference voltage of each node; For the first Real-time voltage of each node; The rated voltage of the power distribution network; For the first The rate of change of voltage at each node.

[0060] Step 106: Select multiple candidate measures from the preset governance measures library based on voltage change rate, key time series feature set and preset voltage change rate threshold, and construct a judgment matrix in combination with preset evaluation indicators.

[0061] In this embodiment, the voltage change rate refers to a quantitative indicator characterizing the deviation of the real-time voltage amplitude of a node from the rated or reference voltage; its magnitude directly reflects the severity of the disturbance. The preset voltage change rate threshold is a critical numerical standard set in the system for classifying voltage fluctuation levels, used to discretize continuously changing voltage data into specific impact levels such as slight, moderate, or severe. The preset mitigation measure library is a database storing various power quality mitigation technologies or equipment parameters, covering standardized mitigation methods corresponding to different disturbance types, such as reactive power compensation, filtering, and voltage recovery. Candidate measures refer to a set of alternative mitigation solutions that are technically applicable to the current fault scenario and are retained after dual logical screening based on disturbance type and impact level. Preset evaluation indicators are predefined criteria used to measure the comprehensive performance of mitigation solutions from multiple dimensions, specifically covering economic cost, technical gain, and on-site installation constraints. The judgment matrix is ​​a numerical array constructed by quantitatively scoring multiple candidate measures selected based on preset evaluation indicators, with the number of rows corresponding to the number of candidate measures and the number of columns corresponding to the number of preset evaluation indicators. By constructing this matrix, abstract multi-objective governance decisions can be transformed into matrix operations based on specific indicator data. The square root method is used to perform geometric averaging and normalization on the judgment matrix containing multi-dimensional scoring information, which effectively smooths out the differences in dimensions and orders of magnitude between different evaluation indicators. Thus, on the basis of ensuring technical feasibility, the comprehensive benefits of each candidate measure can be scientifically and objectively quantified, thereby significantly improving the scientificity and accuracy of controlling power quality disturbances in the distribution network.

[0062] Step 107: Determine the priority weight of each candidate measure based on the judgment matrix and preset evaluation indicators, and determine the optimal governance equipment in the distribution network from the measures with the highest priority weight.

[0063] In this embodiment, priority weight refers to a quantitative value representing the relative merits of each candidate measure after comprehensively considering all evaluation indicators, obtained by using a preset mathematical model (such as the square root method) to perform dimensionality reduction, aggregation, and normalization processing on the multidimensional evaluation data contained in the judgment matrix. This value directly serves as the mathematical basis for decision ranking. The optimal governance device refers to the physical execution unit that has a unique mapping relationship with the calculated highest priority weight in the candidate measure set and is ultimately selected to connect to the distribution network to perform power quality regulation tasks. Specifically, it covers various electrical devices with power regulation capabilities, such as static var generators, active power filters, or hybrid compensation devices. Through this weight-based quantitative decision-making mechanism, the execution entity with the best overall performance can be objectively identified from numerous technically feasible candidate solutions, ensuring that the allocation of governance resources is optimally matched with the current disturbance conditions, avoiding the blindness of human experience-based decisions, and thus improving the accuracy of controlling power quality disturbances in the distribution network.

[0064] Step 108: Determine the dynamic output parameters based on the measured values ​​of the power quality indicators of the optimal treatment equipment, the reference values ​​of the power quality indicators, and the real-time operating data.

[0065] In this embodiment, the measured power quality index refers to a specific numerical value characterizing the current physical level of power quality in the distribution network, obtained through real-time acquisition and quantification by a monitoring terminal. It encompasses instantaneous state quantities of key indicators such as voltage deviation, frequency deviation, and harmonic distortion rate, reflecting the actual operating conditions of the object to be addressed. The reference value of the power quality index refers to an ideal target value set according to power system operating standards or specific power supply quality requirements. It serves as a benchmark comparison object in the feedback control logic, used to define the gap between the governance target and the current state. Dynamic output parameters refer to time-varying control commands generated by processing the aforementioned index deviations and operating scenario data using a control algorithm. These commands drive physical equipment to perform specific compensation or adjustment actions. Specific forms include, but are not limited to, compensation current amplitude, phase commands, or voltage adjustment step size. By constructing a closed-loop feedback control mechanism based on index deviations and operating scenarios, it is possible to ensure that the output of the governance equipment accurately follows the instantaneous fluctuations in power quality and adapts to changes in macroscopic operating conditions, achieving rapid tracking and precise suppression of disturbance signals, thereby improving the accuracy of controlling power quality disturbances in the distribution network.

[0066] Step 109: Control the optimal treatment equipment to operate according to the dynamic output parameters.

[0067] The above technical solution firstly utilizes the variance minimization objective to determine the optimal number of decomposed modes, achieving adaptive and refined decomposition of the original power quality signal. This avoids feature extraction bias caused by signal mode aliasing or noise interference at the source, establishing the data accuracy for disturbance identification. Secondly, by combining the distribution network topology and node voltage change rate to deduce the disturbance propagation path, the perception dimension of disturbance is expanded from isolated monitoring of a single node to full-domain path analysis on the network side, accurately defining the spatial distribution and impact level of the disturbance, ensuring the spatial accuracy of the governance scope. Furthermore, by constructing a judgment system containing multi-dimensional evaluation indicators... The matrix-based approach enables the scientific and objective identification of optimal governance equipment from a vast array of measures, avoiding the blindness of experience-based decision-making and ensuring a high degree of matching between governance methods and specific disturbance conditions, thereby improving decision-making accuracy. Finally, in the control execution phase, a dynamic parameter adjustment mechanism is introduced based on the deviation between the measured and reference values ​​of power quality indicators, constructing a real-time closed-loop control logic of "monitoring-feedback-correction." This logic can fine-tune output parameters in real time according to the transient changes in the grid's operating state, effectively eliminating the steady-state error and response lag present in traditional open-loop control, thus significantly improving the accuracy of controlling power quality disturbances in the distribution network across the entire chain.

[0068] In this embodiment, the step of determining the key time series feature set based on the optimal decomposition mode number, the original power quality signal, the preset disturbance type label, and the preset feature correlation threshold includes: decomposing the original power quality signal according to the optimal decomposition mode number to obtain multiple stationary mode components; extracting high-dimensional time series feature sets from each stationary mode component; and determining the key time series feature set based on the high-dimensional time series feature set, the preset disturbance type label, and the preset feature correlation threshold.

[0069] In this embodiment, stationary modal components refer to narrow-band sub-signals with strict physical meaning and limited frequency bandwidth obtained after final decomposition of non-stationary power quality signals based on a determined optimal decomposition mode number. As a product of optimal decomposition, they eliminate mode aliasing to the greatest extent and can independently and stably characterize the intrinsic properties of disturbance signals in different frequency domains. The high-dimensional time-series feature set refers to the complete set of statistical feature vectors calculated and summarized from the aforementioned stationary components. It covers a vast number of time-domain and frequency-domain indicators such as mean, variance, skewness, kurtosis, and various entropy values. Serving as the initial data pool before feature selection, it aims to avoid overlooking any subtle changes that may reflect the nature of the disturbance. By performing signal decomposition and full feature extraction based on optimal parameters, and then combining labels and thresholds for dimensionality reduction and selection, complex time-domain waveforms can be transformed into a feature space with high signal-to-noise ratio and low redundancy. This ensures that subsequent analysis focuses only on sensitive features containing high-value information, thereby improving the accuracy of controlling power quality disturbances in the distribution network.

[0070] In one embodiment, to accurately extract and enhance weak disturbance features based on the temporal characteristics of power quality disturbances, and to provide high-quality feature support for subsequent identification and propagation analysis, the following steps are taken: First, adaptive temporal decomposition is performed. An improved variational mode decomposition (VMD) algorithm is used to adaptively determine the optimal number of decomposition modes based on the frequency distribution characteristics of the disturbance signal. Based on this optimal number of modes, the original power quality signal is decomposed into several stationary mode components, thereby effectively separating the main disturbance signal from background noise during signal preprocessing. Second, key temporal features are extracted. Multi-dimensional temporal features are extracted from each stationary mode component obtained after decomposition to form a high-dimensional temporal feature set. These multi-dimensional temporal features specifically include: time-domain features characterizing waveform statistical properties (including peak value, kurtosis, skewness, duration, etc.); frequency-domain features characterizing energy spectrum distribution (including fundamental frequency, harmonic content, spectral centroid, etc.); and temporal correlation features characterizing signal evolution patterns (including similarity between adjacent periods, rate of change of trend, etc.). Finally, feature selection and enhancement are performed. The high-dimensional time-series feature set is jointly evaluated and screened based on mutual information entropy and the ReliefF algorithm, retaining key features that are strongly correlated with disturbance identification and eliminating redundant feature information; further, through feature normalization and trend enhancement processing, the discriminative power of weak disturbance features in the feature space is improved, thus completing the construction of the key time-series feature set.

[0071] In one embodiment, three types of features are extracted for each stationary mode component obtained through optimal decomposition of mode number to construct a high-dimensional temporal feature set. These three types of features are time-domain features, frequency-domain features, and temporal correlation features. The time-domain features include calculating the peak value of the stationary mode component. , cliff skewness and duration Frequency domain characteristics include calculating the fundamental frequency of the stationary mode components. Harmonic content and the center of gravity of the spectrum Temporal correlation features include calculating the similarity between adjacent periods of stationary modal components. and trend change rate Finally, the above features of all stationary mode components are combined to form the high-dimensional temporal feature set. .

[0072] In this embodiment, determining the key time series feature set based on the high-dimensional time series feature set, the preset perturbation type label, and the preset feature correlation threshold includes: determining the information entropy of each high-dimensional time series feature in the high-dimensional time series feature set; determining the joint entropy of each high-dimensional time series feature with the preset perturbation type label; determining the mutual information entropy of each high-dimensional time series feature with the preset perturbation type label based on the information entropy of each high-dimensional time series feature, the joint entropy, and the information entropy of the preset perturbation type label; and selecting the key time series feature set from each high-dimensional time series feature based on the mutual information entropy and the preset feature correlation threshold.

[0073] In this embodiment, information entropy refers to a mathematical index that measures the uncertainty or information content of a single feature variable in a high-dimensional time-series feature set. Its numerical value characterizes the dispersion of the feature data distribution or the richness of information it contains. Joint entropy refers to the total uncertainty of a specific feature variable and a preset disturbance type label as a whole system. It reflects the average amount of information required for the feature value and disturbance category to occur simultaneously. Mutual information entropy refers to a quantitative value calculated based on the above entropy values, used to measure the degree of mutual dependence or statistical correlation between feature variables and disturbance labels. Its physical meaning is to clarify how much uncertainty about the disturbance type can be reduced by observing a specific feature, which represents the effective information contribution of the feature to accurately identify the disturbance category. By introducing the entropy calculation method, we can get rid of the dependence on the assumption of a specific data distribution, quantify the intrinsic relationship between features and targets from a statistical perspective, and thus accurately eliminate redundant or noisy features that are irrelevant to disturbance identification. This ensures that the selected feature set has the maximum classification and identification ability, significantly improves the convergence speed and judgment accuracy of subsequent analysis models, and thus improves the accuracy of controlling power quality disturbances in the distribution network.

[0074] In this embodiment of the application, the information entropy of each high-dimensional temporal feature is determined according to the following formula:

[0075]

[0076] in, For the first Information entropy of a high-dimensional temporal feature; Let the value of this high-dimensional time series feature be... The marginal probability; It is the set of values ​​for each high-dimensional time series feature.

[0077] In this embodiment, the joint entropy of each high-dimensional temporal feature and the preset perturbation type label is determined according to the following formula:

[0078]

[0079] in, For the first The joint entropy of a high-dimensional temporal feature and a preset perturbation type label; For the first The high-dimensional time-series features take values ​​of And the disturbance type label is The joint probability; This is a set of labels for all perturbation types.

[0080] In this embodiment, the mutual information entropy between each high-dimensional temporal feature and a preset perturbation type label is determined according to the following formula:

[0081]

[0082] in, For the first The mutual information entropy between a high-dimensional temporal feature and a preset perturbation type label.

[0083] In this embodiment, the key time series feature set is selected from each high-dimensional time series feature according to the following formula:

[0084]

[0085] in, A preset feature relevance threshold is set. In the first... When the mutual information entropy between a high-dimensional temporal feature and a preset perturbation type label is greater than or equal to a preset feature correlation threshold, the first feature will be... A high-dimensional temporal feature is added to the key temporal feature set.

[0086] In this embodiment of the application, the steps of selecting multiple candidate measures from a preset governance measure library based on voltage change rate, key time series feature set and preset voltage change rate threshold, and constructing a judgment matrix in combination with preset evaluation indicators include: determining the disturbance impact level of each node based on the voltage change rate of each node and preset voltage change rate threshold; determining the disturbance type based on key time series feature set; selecting multiple candidate measures from the preset governance measure library based on disturbance type and disturbance impact level, and constructing a judgment matrix in combination with preset evaluation indicators.

[0087] In this embodiment, the disturbance impact level refers to a graded state characterizing the degree of threat posed by the disturbance to the safety of power grid operation, based on the deviation of the voltage change rate of each node from a preset threshold. It maps continuous electrical quantity changes to discrete severity labels (e.g., minor, moderate, severe), serving as the basis for graded response of governance strategies. The disturbance type refers to the specific type of power quality anomaly currently occurring in the power grid, identified using a classifier or matching rules based on key time-series feature sets, such as voltage dips, swells, or specific subharmonic pollution, which clarifies the technical attributes of the governance target. Candidate measures refer to a set of preliminary governance schemes retrieved and matched from a preset governance measure library based on the determined disturbance type and disturbance impact level. These schemes are technically capable of effectively suppressing the current specific disturbance and their governance capabilities are compatible with the current impact level. This step first utilizes voltage change rate and key time-series characteristics to achieve qualitative classification and quantitative grading of disturbance conditions. This introduces dual constraints of physical characteristics and severity during the screening of mitigation measures, effectively eliminating ineffective solutions that are technically incompatible or have excessively high costs due to redundant mitigation capabilities, thus achieving efficient convergence of the candidate set. Second, by constructing a judgment matrix that includes candidate measures and preset evaluation indicators, the performance data of the originally discrete and heterogeneous mitigation solutions are transformed into a standardized matrix structure, providing a unified data foundation and computational carrier for subsequent objective and accurate priority calculation using mathematical models.

[0088] In one embodiment, the disturbance impact level satisfies the following formula:

[0089]

[0090] in, For the first The rate of change of voltage at each node; The rated voltage of the power distribution network; For the first The impact level of disturbance on each node; 1 indicates slight impact, 2 indicates moderate impact, and 3 indicates severe impact.

[0091] In this embodiment, determining the variance of each modal component based on the original power quality signal, and aiming to minimize the total variance of all modal components, determines the optimal number of decomposed modes by: determining the variance of each modal component according to the following formula:

[0092]

[0093] The total variance is determined based on the variance of each modal component, and the optimal number of decomposed modes is determined with the goal of minimizing the total variance.

[0094] in, For the first The variance of each modal component; The length of the original power quality signal; For the first Each modal component in The value at time; For the first The mean of each modal component.

[0095] In this embodiment of the application, the total variance is determined based on the variance of each modal component. Determining the optimal number of decomposed modes with the goal of minimizing the total variance includes: determining the optimal number of decomposed modes according to the following formula:

[0096]

[0097] in, The optimal decomposition mode number; This represents the total number of modal components.

[0098] In this embodiment of the application, the preset constraint conditions satisfy the following formula:

[0099]

[0100] in, For the first node directly connected to the disturbance node Voltage amplitude of each candidate node; The voltage amplitude at the disturbed node; For the perturbation node and the first The line current amplitude between candidate nodes; For the perturbation node and the first Line impedance between candidate nodes; This is the preset maximum voltage drop.

[0101] In the embodiments of this application, This refers to a node that is electrically connected to the disturbing node (the current reference node) in the physical topology, but whose status as a node in the propagation path has not yet been confirmed. This definition allows the system to move beyond being limited to the starting point and instead serve as a general judgment logic that can be repeatedly invoked to determine whether a disturbance can propagate from any confirmed disturbed node across the physical line to its directly connected downstream node. This enables the propagation path to extend gradually from a point to a surface within the distribution network.

[0102] In this embodiment of the application, determining the priority weight of each candidate measure based on the judgment matrix and preset evaluation indicators includes: determining the priority weight according to the following formula:

[0103]

[0104] in, For the first The priority weight of each candidate measure; To determine the first element in the matrix The candidate measures in the first Quantitative evaluation value or membership degree under a preset evaluation index; To determine the first element in the matrix The candidate measures in the first Quantitative evaluation value or membership degree under a preset evaluation index; The total number of multiple candidate measures; The number of pre-set evaluation indicators.

[0105] In one embodiment, the judgment matrix satisfies the following formula: .

[0106] In one embodiment, firstly, precise disturbance classification is performed. Based on the aforementioned key time-series feature set, an improved Support Vector Machine (SVM) algorithm is used for pattern recognition. This algorithm enables precise classification of specific disturbance types such as harmonics, voltage sags, voltage swells, and flicker, while simultaneously classifying the severity of disturbances (specifically, into mild, moderate, and severe), thereby clarifying the qualitative and quantitative attributes of the target disturbance. Secondly, a governance measure library is constructed. Existing power quality governance technologies are integrated, specifically covering reactive power compensation, harmonic filtering, voltage regulation, and load optimization scheduling. For each governance measure in the library, its specific applicable scenarios, governance cost indicators, and physical boundaries of governance effects are defined, thus constructing a pre-defined governance measure library containing multi-dimensional attribute information. Finally, precise matching and scheme generation are performed. Taking into account the identified disturbance type and severity level, as well as equipment constraints such as on-site installation space and capacity limitations, and the project's cost budget requirements, an evaluation model is constructed using the Analytic Hierarchy Process (AHP) to calculate the priority weight of each candidate remediation measure. Finally, the optimal targeted remediation scheme is determined based on the weight ranking results to achieve the optimization of economic costs while meeting the technical remediation requirements.

[0107] In this embodiment of the application, determining the dynamic output parameters based on the measured values ​​and reference values ​​of the power quality indicators of the optimal treatment equipment, as well as real-time operating data, includes: determining the power quality indicator deviation of the optimal treatment equipment according to the following formula:

[0108]

[0109] in, for Dynamic output parameters at any given time; for The deviation of power quality indicators at any given time is obtained based on the measured values ​​and reference values ​​of power quality indicators. , and These are PID parameters that are self-tuned based on real-time operating data. It is the integral variable.

[0110] In one embodiment, firstly, operational status is perceived. Multi-dimensional operational data of the distribution network is collected in real time, including load power, distributed power output, equipment switching status, and ambient temperature. Based on this data, an operational status evaluation index system is constructed, and the specific operational scenario of the distribution network is determined in real time. The operational scenario includes at least a stable scenario, a load fluctuation scenario, and a power output fluctuation scenario. Secondly, dynamic adjustment of governance parameters is performed. For different identified operational scenarios, a pre-established governance parameter adjustment rule library is invoked, and a proportional-integral-derivative (PID) optimization algorithm is applied for calculation. Based on the calculation results, the operating parameters of the governance equipment are adaptively adjusted, specifically including adjusting the compensation capacity of the reactive power compensation device and the filtering frequency band of the filtering device, to ensure stable governance effects under different operating conditions. Finally, real-time verification of governance effects is implemented. Key power quality indicators, including harmonic distortion rate, voltage deviation, and flicker value, are monitored in real time through monitoring terminals. The current governance effect is evaluated based on the monitoring results. If any indicator fails to meet the preset standard, a secondary parameter adjustment logic is immediately triggered, thereby forming a closed-loop optimization control to ensure that power quality indicators are always within the compliance range.

[0111] The above technical solution firstly utilizes the variance minimization objective to determine the optimal number of decomposed modes, achieving adaptive and refined decomposition of the original power quality signal. This avoids feature extraction bias caused by signal mode aliasing or noise interference at the source, establishing the data accuracy for disturbance identification. Secondly, by combining the distribution network topology and node voltage change rate to deduce the disturbance propagation path, the perception dimension of disturbance is expanded from isolated monitoring of a single node to full-domain path analysis on the network side, accurately defining the spatial distribution and impact level of the disturbance, ensuring the spatial accuracy of the governance scope. Furthermore, by constructing a judgment system containing multi-dimensional evaluation indicators... The matrix-based approach enables the scientific and objective identification of optimal governance equipment from a vast array of measures, avoiding the blindness of experience-based decision-making and ensuring a high degree of matching between governance methods and specific disturbance conditions, thereby improving decision-making accuracy. Finally, in the control execution phase, a dynamic parameter adjustment mechanism is introduced based on the deviation between the measured and reference values ​​of power quality indicators, constructing a real-time closed-loop control logic of "monitoring-feedback-correction." This logic can fine-tune output parameters in real time according to the transient changes in the grid's operating state, effectively eliminating the steady-state error and response lag present in traditional open-loop control, thus significantly improving the accuracy of controlling power quality disturbances in the distribution network across the entire chain.

[0112] Figure 2A schematic diagram illustrating the structure of a computer device according to an embodiment of this application is provided. Figure 2 As shown, this application provides a computer device that may include:

[0113] Memory 210 is configured to store instructions;

[0114] The processor 220 is configured to retrieve instructions from memory 210 and to implement the methods described above when executing instructions.

[0115] This application also provides a machine-readable storage medium storing instructions that cause a machine to perform the above-described method.

[0116] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0117] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0118] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0119] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0120] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0121] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0122] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0123] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0124] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A power quality control method for a power distribution network, characterized in that, The method includes: Acquire real-time operating data of the distribution network and raw power quality signals of disturbed nodes, and decompose the raw power quality signals into multiple modal components; Based on the original power quality signal, the variance of each modal component is determined, and the optimal number of decomposed modes is determined with the goal of minimizing the total variance of all modal components. The original power quality signal is decomposed according to the optimal decomposition mode number to obtain multiple stationary mode components; High-dimensional temporal feature sets are extracted from each stationary modal component; Determine the information entropy of each high-dimensional temporal feature in the high-dimensional temporal feature set; Determine the joint entropy of each high-dimensional temporal feature with the preset perturbation type label; The mutual information entropy between each high-dimensional time series feature and the preset perturbation type label is determined based on the information entropy, joint entropy of each high-dimensional time series feature and the information entropy of the preset perturbation type label. Based on the mutual information entropy and the preset feature correlation threshold, a set of key time series features is selected from each high-dimensional time series feature; In the topology of the distribution network, starting from the disturbance node, the disturbance propagation path is obtained by reasoning according to preset constraints; the disturbance node is any node in the topology. Determine the voltage change rate at each node in the disturbance propagation path; Based on the voltage change rate, the key time series feature set, and the preset voltage change rate threshold, multiple candidate measures are selected from the preset governance measure library, and a judgment matrix is ​​constructed by combining preset evaluation indicators. The priority weights of each candidate measure are determined based on the judgment matrix and the preset evaluation index, and the optimal governance equipment in the distribution network is determined from the measure with the highest priority weight. Dynamic output parameters are determined based on the measured and reference values ​​of the power quality indicators of the optimal treatment equipment and the real-time operating data. The optimal treatment equipment is controlled to operate according to the dynamic output parameters.

2. The method according to claim 1, characterized in that, The step of selecting multiple candidate measures from a preset governance measure library based on the voltage change rate, the key time-series feature set, and a preset voltage change rate threshold, and constructing a judgment matrix by combining preset evaluation indicators, includes: The disturbance impact level of each node is determined based on the voltage change rate of each node and the preset voltage change rate threshold. The type of disturbance is determined based on the key time-series feature set; Based on the disturbance type and the disturbance impact level, multiple candidate measures are selected from the preset governance measures library, and a judgment matrix is ​​constructed by combining preset evaluation indicators.

3. The method according to claim 1 or 2, characterized in that, The step of determining the variance of each modal component based on the original power quality signal, and determining the optimal number of decomposed modes with the objective of minimizing the total variance of all modal components, includes: The variance of each modal component is determined using the following formula: The total variance is determined based on the variance of each modal component, and the optimal number of decomposed modes is determined with the goal of minimizing the total variance. in, For the first The variance of each modal component; The length of the original power quality signal; For the first Each modal component in The value at time; For the first The mean of each modal component.

4. The method according to claim 3, characterized in that, The step of determining the total variance based on the variance of each modal component, and determining the optimal number of decomposition modes with the goal of minimizing the total variance, includes: The optimal number of decomposition modes is determined according to the following formula: in, The optimal decomposition mode number; The total number of the multiple modal components.

5. The method according to claim 1 or 2, characterized in that, The preset constraint conditions satisfy the following formula: in, The first node connected to the disturbance node Voltage amplitude of each candidate node; The voltage amplitude of the disturbed node; For the disturbance node and the first The line current amplitude between candidate nodes; For the disturbance node and the first Line impedance between candidate nodes; This is the preset maximum voltage drop.

6. The method according to claim 1 or 2, characterized in that, The determination of the priority weights of each candidate measure based on the judgment matrix and the preset evaluation index includes: The priority weight is determined according to the following formula: in, For the first The priority weight of each candidate measure; The first in the judgment matrix The candidate measures in the first Quantitative evaluation values ​​under preset evaluation indicators; The first in the judgment matrix The candidate measures in the first Quantitative evaluation values ​​under preset evaluation indicators; The total number of the multiple candidate measures; The number of the preset evaluation indicators.

7. The method according to claim 1 or 2, characterized in that, The step of determining the dynamic output parameters based on the measured and reference values ​​of the power quality indicators of the optimal treatment equipment and the real-time operating data includes: The power quality index deviation of the optimal treatment equipment is determined according to the following formula: in, for Dynamic output parameters at any given time; for The deviation of the power quality index at any given time is obtained based on the measured value of the power quality index and the reference value of the power quality index. , and These are the PID parameters that are self-tuned based on the real-time operating data; It is the integral variable.

8. A computer device, characterized in that, include: The memory is configured to store instructions; And a processor configured to retrieve the instructions from the memory and, when executing the instructions, to implement the method according to any one of claims 1 to 7.

Images