Low-voltage cabinet distributed control method and system based on multi-source data fusion

Abstract

The invention discloses a low-voltage cabinet distributed control method and system based on multi-source data fusion, and relates to the technical field of distributed control. The method comprises the following steps: extracting topological structures of an equipment layer, a loop layer and a system layer from a distributed low-voltage cabinet set to obtain the topological structure of the equipment layer, the topological structure of the loop layer and the topological structure of the system layer; obtaining an equipment physical interface parameter set; obtaining a feeder loop state feature set; performing adjacency state analysis on the equipment physical interface parameter set to obtain an adjacency state matrix set; and taking the feeder loop state feature set and the adjacent state matrix set as real-time low-voltage cabinet working conditions, matching the real-time low-voltage cabinet working conditions with a pre-constructed working condition distributed control library, determining a target distributed control scheme, and performing distributed control on a distributed low-voltage cabinet set. The technical problems that in the prior art, low-voltage cabinet distributed control is poor in flexibility and poor in working condition adaptability are solved, and the technical effect of achieving intelligent self-adaptive control based on the real-time working condition is achieved.

Description

Technical Field

[0001] This invention relates to the field of distributed control technology, and specifically to a distributed control method and system for low-voltage switchgear based on multi-source data fusion. Background Technology

[0002] Low-voltage switchgear, as a core device in low-voltage power distribution systems, undertakes important functions such as power distribution, control, and protection, and is widely used in many scenarios such as industrial plants, commercial buildings, and residential communities. Traditional low-voltage switchgear control methods lack a comprehensive consideration of the interrelationships between devices and the overall system operating conditions. In distributed low-voltage switchgear systems, the individual switchgear units are not isolated but form an organic whole through electrical connections. Their operating states affect each other; for example, a fault in one switchgear unit may cause load changes in adjacent switchgear units, thus affecting the stability of the entire system. However, traditional control methods often only focus on the operating parameters of individual switchgear units, ignoring the adjacency relationships and synergistic effects between devices. This makes it impossible to achieve precise coordinated control based on the overall operating conditions, and when dealing with complex faults or abnormal situations, control errors or decision delays are prone to occur, affecting the normal operation of the system. Summary of the Invention

[0003] This application provides a distributed control method and system for low-voltage switchgear based on multi-source data fusion, which solves the technical problems of poor flexibility and insufficient adaptability to operating conditions in the distributed control of low-voltage switchgear in the prior art.

[0004] The first aspect of this application provides a distributed control method for low-voltage switchgear based on multi-source data fusion, the method comprising:

[0005] The distributed low-voltage switchgear assembly is subjected to equipment layer, loop layer, and system layer topology extraction to obtain equipment layer topology, loop layer topology, and system layer topology. Based on the equipment layer topology, equipment physical interface parameters are collected to obtain a set of equipment physical interface parameters. Based on the loop layer topology, feeder loop status is collected to obtain a set of feeder loop status features. Based on the system layer topology, adjacency status analysis is performed between adjacent distributed low-voltage switchgear based on the set of equipment physical interface parameters to obtain a set of adjacency status matrices. The set of feeder loop status features and the set of adjacency status matrices are used as real-time low-voltage switchgear operating conditions. Based on these real-time operating conditions, a pre-built distributed control database of operating conditions is matched to determine a target distributed control scheme. Distributed control is then performed on the distributed low-voltage switchgear assembly based on the target distributed control scheme.

[0006] A second aspect of this application provides a distributed control system for low-voltage switchgear based on multi-source data fusion, the system comprising:

[0007] Topology Extraction Module: Extracts the topology of the distributed low-voltage switchgear set at the equipment layer, loop layer, and system layer to obtain the equipment layer topology, loop layer topology, and system layer topology; Parameter Acquisition Module: Acquires equipment physical interface parameters based on the equipment layer topology to obtain a set of equipment physical interface parameters; Loop Status Acquisition Module: Acquires feeder loop status based on the loop layer topology to obtain a set of feeder loop status features; Analysis Module: Analyzes the adjacency status between adjacent distributed low-voltage switchgear based on the equipment physical interface parameter set using the system layer topology to obtain a set of adjacency status matrices; Control Module: Uses the set of feeder loop status features and the set of adjacency status matrices as real-time low-voltage switchgear operating conditions, matches these real-time operating conditions with a pre-built distributed control database of operating conditions, determines a target distributed control scheme, and performs distributed control on the distributed low-voltage switchgear set based on the target distributed control scheme.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0009] First, the topology of the distributed low-voltage switchgear assembly is extracted at the equipment layer, loop layer, and system layer, resulting in equipment layer topology, loop layer topology, and system layer topology. Next, based on the equipment layer topology, physical interface parameters of the equipment are collected, resulting in a set of equipment physical interface parameters; based on the loop layer topology, feeder loop states are collected, resulting in a set of feeder loop state features. Further, based on the system layer topology, adjacency state analysis is performed between adjacent distributed low-voltage switchgear based on the set of equipment physical interface parameters, resulting in a set of adjacency state matrices. Finally, the set of feeder loop state features and the set of adjacency state matrices are used as real-time low-voltage switchgear operating conditions. These real-time operating conditions are matched with a pre-built distributed control library to determine a target distributed control scheme. Distributed control of the distributed low-voltage switchgear assembly is then performed based on this target distributed control scheme. This approach solves the technical problems of poor flexibility and insufficient adaptability of distributed control for low-voltage switchgear in existing technologies, achieving the technical effect of intelligent adaptive control based on real-time operating conditions. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 A schematic diagram of the distributed control method for low-voltage switchgear based on multi-source data fusion provided in this application embodiment;

[0012] Figure 2 This is a schematic diagram of a distributed control system structure for low-voltage switchgear based on multi-source data fusion, provided in an embodiment of this application.

[0013] Figure labeling: Topology extraction module 11, parameter acquisition module 12, loop status acquisition module 13, analysis module 14, control module 15. Detailed Implementation

[0014] This application provides a distributed control method and system for low-voltage switchgear based on multi-source data fusion, which solves the technical problems of poor flexibility and insufficient adaptability to operating conditions in the existing distributed control of low-voltage switchgear.

[0015] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. 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.

[0016] It should be noted that the terms "comprising" and "having" are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to these processes, methods, products, or devices.

[0017] Example 1, as Figure 1 As shown, this application provides a distributed control method for low-voltage switchgear based on multi-source data fusion, wherein the method includes:

[0018] The equipment layer, loop layer, and system layer topologies of the distributed low-voltage switchgear assembly are extracted to obtain the equipment layer topology, loop layer topology, and system layer topology.

[0019] Specifically, firstly, a complete list of low-voltage switchgear equipment and their access information is obtained from the distributed low-voltage switchgear set, including physical equipment data such as switching elements, monitoring devices, acquisition units, and cable connection ports connected to each low-voltage switchgear. Based on this data, a device-layer topology is constructed with individual low-voltage switchgear as basic nodes. The device-layer topology uses the physical connection relationships within the low-voltage switchgear as edges, reflecting the structural layout of the equipment within each low-voltage switchgear and its physical interface connection methods. Next, the feeder circuit information connected to each low-voltage switchgear is identified, including the outgoing circuit number, the circuit type (e.g., lighting, power, fire protection), the load type connected to the circuit, and operating parameters. Using the feeder circuit as the basic unit, a circuit-layer topology is established according to the spatial and electrical connection relationships of each circuit in the low-voltage switchgear set, where each node represents a specific feeder circuit, and each edge represents the path relationship between circuits for electrical interaction or load coordination. Furthermore, based on the installation architecture of the entire low-voltage power distribution system, information on the electrical connection methods, communication paths, shared bus structures, and their locations between each low-voltage switchgear and adjacent low-voltage switchgear is collected to form a system-level topology with distributed low-voltage switchgear as nodes and system-level connection relationships as edges. This system-level topology characterizes the macroscopic logical relationships and communication coordination paths between low-voltage switchgear, serving as the foundational structure supporting subsequent adjacency state analysis and distributed control strategy generation.

[0020] Based on the device layer topology, the device physical interface parameters are collected to obtain the device physical interface parameter set.

[0021] Based on the device layer topology, each low-voltage switchgear in the distributed low-voltage switchgear set is traversed one by one, and the physical device unit corresponding to each node in the device layer topology is identified. The physical device unit includes, but is not limited to, circuit breakers, contactors, power metering devices, current / voltage sensors, communication interface modules, and wiring terminals.

[0022] During the traversal, the system identifies the interface level information of each physical device within a preset acquisition window and collects its current interface operating parameters using standard communication protocols (such as Modbus, CAN, IEC 61850, etc.). These interface operating parameters include at least voltage, current, power factor, frequency, interface temperature, interface load status, communication status, and interface type (input / output). Each parameter acquisition result is recorded as a timestamped data point and categorized by device number, forming an initial set of device physical interface parameter sequences. The system then performs data filtering on this initial set. Specifically, it removes incomplete, unreliable, or invalid data and retains high-confidence valid parameter data points through steps such as outlier removal, missing value imputation, time alignment, and parameter normalization. Finally, the system combines the processed physical interface parameters of each device into a unified dataset, forming the device physical interface parameter set.

[0023] Furthermore, based on the device layer topology, device physical interface parameters are collected to obtain a set of device physical interface parameters, including:

[0024] Within a preset acquisition window, the device layer topology is traversed to obtain the device physical interface parameters of each distributed low-voltage cabinet in the distributed low-voltage cabinet set, thus obtaining an initial set of device physical interface parameter sequences. The initial set of device physical interface parameter sequences is then filtered to determine the set of device physical interface parameters.

[0025] Within a set data acquisition window, the system traverses each node in the equipment layer topology, where each node corresponds to a specific low-voltage equipment unit within the distributed low-voltage switchgear set. For each low-voltage switchgear, the system extracts information from each physical interface based on its internal topology, including circuit breakers, transformers, terminals, energy meters, acquisition modules, and communication ports, and acquires the operating parameter data of the current interface through its communication interface. Each acquisition operation is timestamped, forming a time-series data, which is then aggregated to form an initial set of equipment physical interface parameter sequences. The system performs data filtering on the initial set of equipment physical interface parameter sequences to determine the final set of equipment physical interface parameters.

[0026] Furthermore, the initial set of device physical interface parameter sequences is filtered to determine the set of device physical interface parameters, including:

[0027] A coordinate system is constructed with time as the horizontal axis and parameters as the vertical axis. The initial set of device physical interface parameters is input into the coordinate system to obtain a set of data filtering coordinate systems, wherein each data filtering coordinate system has a set of filtering coordinate points. A first data filtering coordinate system and a first set of filtering coordinate points are extracted from the set of data filtering coordinate systems. The first set of filtering coordinate points is filtered in the first data filtering coordinate system using a multi-scale linear rotation filtering method to determine a first target filtering coordinate point, and the first target filtering coordinate point is used as the first device physical interface parameter. The data filtering coordinate system set is then filtered in the corresponding data filtering coordinate system using the multi-scale linear rotation filtering method to determine the set of device physical interface parameters.

[0028] Specifically, each initial sequence of device physical interface parameters is mapped to a two-dimensional coordinate system. Specifically, time is used as the horizontal axis, and the corresponding device physical parameters (such as current, voltage, power factor, temperature, etc.) are used as the vertical axis to construct a set of data filtering coordinate systems. Each device corresponds to an independent coordinate system. In this coordinate system, each parameter value in the sequence corresponds to a coordinate point, forming a set of filtering coordinate points that reflect the dynamic trajectory of the device within the acquisition window.

[0029] A set of coordinate systems is randomly selected from the data filtering coordinate system set as the initial analysis object, denoted as the first data filtering coordinate system. The corresponding parameter trajectory points constitute the first filtering coordinate point set. In the first data filtering coordinate system, a multi-scale straight-line rotation filtering method is applied to extract features from the first filtering coordinate point set. Specifically, an initial reference straight line passing through the origin of the coordinate system and parallel to the horizontal axis is used as the initial filtering baseline. An initial rotation direction (clockwise or counterclockwise) and rotation scale angle are set, and the filtering baseline is iteratively rotated to obtain rotating straight lines at different angles. After each rotation, the system calculates the fitting reliability coefficient of the rotating straight line relative to the coordinate point set. This coefficient can be evaluated based on indicators such as the average distance from the coordinate points to the line and linear correlation. By comparing the reliability coefficient of the current rotating straight line with that of the previous line, it is determined whether the rotation direction needs to be adjusted and the rotation scale reduced. When the rotating straight line obtained in a certain iteration reaches the optimal reliability coefficient or meets the preset upper limit of the number of iterations, it is considered that the filtering has converged, and this straight line is determined as the target data filtering line. Using the target data filtering line as a reference, a preset filtering bandwidth is set, and all coordinate points within the bandwidth are assigned to the first filtering neighborhood. The average or representative value of the coordinate points in the neighborhood is calculated and used as the representative point of the current parameter sequence, which is the first target filtering coordinate point. The first target filtering coordinate point constitutes the first physical interface parameter result of the device.

[0030] The above process is carried out sequentially in each data filtering coordinate system, thereby completing the filtering process of the entire initial set of device physical interface parameter sequences and finally obtaining the set of device physical interface parameters.

[0031] Furthermore, the multi-scale linear rotation filtering method is used to filter the first set of filtering coordinate points in the first data filtering coordinate system to determine the first target filtering coordinate point, and the first target filtering coordinate point is used as the first device physical interface parameter, including:

[0032] The initial data filtering line is a straight line passing through the origin of the first data filtering coordinate system and parallel to the horizontal axis of the first data filtering coordinate system. An initial rotation direction is obtained, and the initial data filtering line is rotated according to a preset rotation scale to obtain a first iterative data filtering line. The reliability coefficients of the first iterative data filtering line and the initial data filtering line are compared to determine the update rotation direction, update rotation scale, and stage data filtering line. The stage data filtering line is rotated according to the update rotation direction and update rotation scale to obtain a second iterative data filtering line. The reliability coefficients of the second iterative data filtering line and the stage data filtering line are compared to update the stage data filtering line, and the update rotation direction and update rotation scale are updated again until a preset number of updates is met to obtain a target data filtering line. Based on the target data filtering line and a preset filtering bandwidth, the first set of filtering coordinate points is filtered to determine the first target filtering coordinate point.

[0033] In the first data filtering coordinate system, a straight line passing through the origin and parallel to the horizontal axis (time axis) is defined as the initial data filtering line, serving as the starting baseline for data fitting. An initial rotation direction (e.g., clockwise) and a preset rotation scale (e.g., 1° each time) are set, and the initial data filtering line is rotated in the coordinate system according to the preset rotation scale to obtain the first iteration data filtering line. Next, the reliability coefficient of this first iteration data filtering line is calculated to measure the degree of fit of the line to the original data point set. The reliability coefficient is calculated as follows: the number of coordinate points in the first filtering coordinate point set whose distance to the current rotating line is less than the preset filtering bandwidth is divided by the total number of points in the set; the resulting ratio is the reliability coefficient.

[0034] The reliability coefficient of the line selected from the first iteration data is compared with the reliability coefficient of the line selected from the initial data: if the reliability coefficient of the line selected from the first iteration data is less than or equal to the reliability coefficient of the initial line, it means that the current rotation direction is not conducive to fitting optimization. At this time, the rotation direction is reversed and the rotation scale is halved; if the reliability coefficient of the line selected from the first iteration data is better than the original line, the current rotation direction is maintained and the iteration continues.

[0035] Based on the above judgment, the stage data filtering line is determined, the rotation direction and rotation scale are updated, and rotation iteration continues to generate the second iteration data filtering line. The reliability coefficient is calculated again and compared. The above process is executed cyclically until the preset maximum number of iterations is met or the change in the reliability coefficient tends to converge, finally obtaining an optimal target data filtering line. Taking the target data filtering line as the center, a preset filtering bandwidth threshold is set. All coordinate points whose vertical distance to the target data filtering line is less than the bandwidth are taken as valid filtering points. Statistical calculations (such as mean calculation) are performed on these points to determine the first target filtering coordinate point, which is used as the representative physical interface parameter corresponding to the device.

[0036] Furthermore, comparing the reliability coefficient of the first iterative data filtering line with the reliability coefficient of the initial data filtering line to determine the updated rotation direction, updated rotation scale, and stage data filtering line includes:

[0037] When the reliability coefficient of the first iterative data filtering line is less than or equal to the reliability coefficient of the initial data filtering line, the opposite direction of the initial rotation direction is taken as the updated rotation direction; half of the preset rotation scale is taken as the updated rotation scale, and the initial data filtering line is taken as the stage data filtering line.

[0038] When the reliability coefficient of the first iteration data selection line is less than or equal to the reliability coefficient of the initial data selection line, it indicates that the current iteration has not improved the data fitting effect. At this time, the opposite direction of the original initial rotation direction is set as the update rotation direction, and the original preset rotation scale is halved as the current update rotation scale. At the same time, the original initial data selection line is retained as the current stage data selection line for the reference of the next iteration.

[0039] Conversely, if the reliability coefficient of the first iteration data selection line is greater than that of the initial data selection line, it indicates that the current rotation iteration has effectively improved the fitting ability of the line to the data points. At this time, the current rotation direction and rotation scale remain unchanged, and the first iteration data selection line is used as the new stage data selection line for subsequent iterations.

[0040] Furthermore, the first filter coordinate points in the first set of filter coordinate points whose distance to the target data filter line is within the preset filter bandwidth range are added to the first filter neighborhood, the mean of the first filter coordinate points in the first filter neighborhood is calculated, and the first target filter coordinate point is determined.

[0041] In this embodiment, a more representative first target screening coordinate point is determined from the first set of screening coordinate points. This is done using a distance- and statistical feature extraction method. Specifically, this includes: screening coordinate points in the first set of screening coordinate points whose distance from the target data screening line is less than or equal to a preset screening bandwidth and assigning them to a first screening neighborhood. After obtaining the first screening neighborhood, the mean of all first screening coordinate points within the neighborhood is calculated, and this mean result is used as the final determined first target screening coordinate point.

[0042] Based on the aforementioned loop layer topology, the feeder loop status is acquired to obtain a set of feeder loop status features.

[0043] Based on the loop topology, the specific composition and connection relationships of each feeder loop in the distributed low-voltage switchgear assembly are identified, and the corresponding equipment nodes and their electrical connection paths for each feeder loop are determined. The system sets preset status monitoring indicators for each feeder loop, including at least key electrical parameters such as current, voltage, active power, reactive power, power factor, harmonic content, temperature, insulation resistance, and grounding resistance. Through sensors or intelligent measurement units installed on the feeder loops, combined with communication interfaces, the system collects the above status parameter data in real time, forming multi-dimensional feature data reflecting the operating status of the feeder loops. The collected feeder loop status data is categorized and synchronized according to the loop topology to construct a feeder loop status feature set, containing the dynamic operating status information of each feeder loop within the acquisition time window.

[0044] Furthermore, based on the aforementioned loop layer topology, feeder loop status is acquired to obtain a set of feeder loop status features, including:

[0045] Obtain preset feeder loop status indicators, wherein the preset feeder loop status indicators include at least current, power, power factor, harmonic content, temperature, insulation resistance, and grounding resistance; based on the preset feeder loop status indicators, perform feeder loop status acquisition on the loop layer topology to obtain a set of feeder loop status features.

[0046] The system sets status monitoring indicators for feeder circuits, pre-defined indicators including key electrical parameters such as current, power, power factor, harmonic content, temperature, insulation resistance, and grounding resistance. Based on the circuit topology, monitoring points and related equipment for each feeder circuit are located. Real-time operational data is collected item by item according to the pre-defined indicators using deployed sensors or measuring devices. During data collection, the system automatically synchronizes multi-dimensional status data for each feeder circuit, forming a structured set of feeder circuit status characteristics that reflects the overall operational status and health level of the feeder circuit within a specific time window.

[0047] Based on the system layer topology, the adjacency state analysis between adjacent distributed low-voltage cabinets is performed on the set of physical interface parameters of the equipment to obtain an adjacency state matrix set.

[0048] Based on the system-level topology, any two adjacent low-voltage switchgear units in the distributed low-voltage switchgear set are identified, and an adjacent low-voltage switchgear combination set is constructed. For each pair of low-voltage switchgear in the adjacent low-voltage switchgear combination, relevant interface parameters are extracted from the corresponding equipment physical interface parameter set to obtain the interface state data between the two low-voltage switchgears. Based on the equipment physical interface parameters, the parameter similarity or correlation between adjacent low-voltage switchgears is calculated. Commonly used calculation methods include Euclidean distance, cosine similarity, or correlation coefficient. The calculated similarity index is normalized and matrix-processed to form an adjacency state matrix set, which reflects the degree of state correlation and adjacency relationship strength between adjacent low-voltage switchgears.

[0049] Furthermore, based on the system layer topology, an adjacency state analysis is performed on the set of physical interface parameters of the equipment to obtain an adjacency state matrix set, including:

[0050] Based on the system layer topology, any two adjacent distributed low-voltage cabinets in the distributed low-voltage cabinet set are extracted to obtain a set of adjacent distributed low-voltage cabinet combinations; combined with the set of equipment physical interface parameters, the adjacency state of each adjacent distributed low-voltage cabinet combination in the set of adjacent distributed low-voltage cabinet combinations is analyzed to obtain the set of adjacency state matrices.

[0051] Based on the physical connection relationships, electrical connection paths, or communication link information of low-voltage switchgear represented in the system-level topology, any two adjacent distributed low-voltage switchgear in the distributed low-voltage switchgear set are extracted to construct a set containing all low-voltage switchgear combinations with adjacent relationships, denoted as the adjacent distributed low-voltage switchgear combination set. Combining the set of equipment physical interface parameters, adjacency state analysis is performed on each adjacent distributed low-voltage switchgear combination in the adjacent distributed low-voltage switchgear combination set. This includes extracting the interface electrical parameters (such as voltage, current, interface type, communication status, etc.) and physical layer connection parameters (such as port location, connection strength, connection type, etc.) of each pair of adjacent low-voltage switchgear, and calculating similarity or correlation based on a preset similarity measurement method (such as Euclidean distance, cosine similarity, or parameter difference threshold judgment). The calculated adjacency relationship values ​​are normalized and assembled into an adjacency state matrix according to a predetermined format to obtain the adjacency state matrix set.

[0052] Furthermore, by combining the set of physical interface parameters of the equipment, the adjacency state of each adjacent distributed low-voltage switchgear combination in the set of adjacent distributed low-voltage switchgear combinations is analyzed to obtain the adjacency state matrix set, including:

[0053] Extract the first adjacent distributed low-voltage switchgear combination from the set of adjacent distributed low-voltage switchgear combinations; extract the set of equipment physical interface parameters based on the first adjacent distributed low-voltage switchgear combination to obtain the first equipment physical interface parameter and the second equipment physical interface parameter; calculate the similarity between the first equipment physical interface parameter and the second equipment physical interface parameter to determine the parameter similarity set; normalize and matrix-process the parameter similarity set to obtain the first adjacency state matrix; perform adjacency state analysis on the adjacent distributed low-voltage switchgear combination set according to the set of equipment physical interface parameters to obtain the adjacency state matrix set.

[0054] Extract a first adjacent distributed low-voltage switchgear combination from the set of adjacent distributed low-voltage switchgear combinations. This combination consists of two physically adjacent distributed low-voltage switchgears or two that are directly connected electrically / communicationally. Based on this first adjacent distributed low-voltage switchgear combination, extract the interface parameter information of the two distributed low-voltage switchgears within the combination from the set of equipment physical interface parameters to obtain the first and second equipment physical interface parameters. These parameters include, but are not limited to, features such as voltage, current, power factor, interface number, interface status, and communication baud rate. Then, based on a preset similarity calculation function (such as vector cosine similarity, reciprocal Euclidean distance, normalized difference, etc.), analyze the first equipment physical interface parameters. The interface parameters and the physical interface parameters of the second device are evaluated for similarity to generate a parameter similarity set, which is used to represent the degree of connection or state matching between the pair of devices. Further, the similarity values ​​in the parameter similarity set are normalized, and the normalized similarity values ​​are sequentially filled into an adjacency state matrix with a preset structure to form a first adjacency state matrix. Each element of the adjacency state matrix is ​​used to characterize the degree of matching between the two devices in the corresponding parameter dimension. The above extraction and combination, interface parameter extraction, similarity calculation and matrix construction steps are repeated to process all adjacent device pairs in the adjacent distributed low-voltage cabinet combination set in sequence, and finally a complete adjacency state matrix set is obtained.

[0055] The feeder circuit state feature set and the adjacent state matrix set are used as real-time low-voltage switchgear operating conditions. Based on the real-time low-voltage switchgear operating conditions, they are matched with the pre-built operating condition distributed control library to determine the target distributed control scheme. Based on the target distributed control scheme, the distributed low-voltage switchgear set is subjected to distributed control.

[0056] Specifically, the feeder loop state feature set and the adjacent state matrix set are fused to form an integrated real-time low-voltage switchgear operating condition description vector. This vector comprehensively represents the current operating status of each distributed low-voltage switchgear and their interconnections. The real-time low-voltage switchgear operating condition is used as input features and matched with historical operating condition data in a pre-built distributed operating condition control library. This library contains multiple sets of historical operating scenarios, each containing a historical operating condition description vector and a corresponding verified distributed control strategy, including power switching logic, feeder reconfiguration logic, isolation operation logic, power balancing logic, and fault response strategies. Specifically, the similarity between the real-time low-voltage switchgear operating condition and historical operating conditions is calculated using vector matching methods (such as KNN, DTW, hash indexing, or deep embedding matching networks) to determine the optimal target distributed control scheme. Based on the target distributed control scheme, dynamic distributed control is implemented on the distributed low-voltage switchgear set according to the control logic and parameters defined in the scheme.

[0057] The feeder loop state feature set and the adjacent state matrix set are fused to form an integrated real-time low-voltage switchgear operating condition description vector. This process includes: First, the feeder loop state feature set is structured by aligning and uniformly encoding multiple feeder loop state indicators corresponding to each low-voltage switchgear to obtain a feeder loop state feature vector for each switchgear. Next, the adjacent state matrix set is partitioned, extracting the adjacent state matrix between each low-voltage switchgear and its neighboring switchgears as a separate adjacent feature vector. Then, the feeder loop state feature vector is concatenated and fused with the corresponding adjacent feature vector to obtain a composite operating condition description vector for the low-voltage switchgear at the current moment. The fusion method can be implemented using direct concatenation, attention fusion, principal component analysis compression fusion, or multimodal embedding mapping, to adapt to the feature coupling requirements under different control scenarios. Finally, the composite operating condition description vectors of all low-voltage switchgears are combined to form a unified real-time low-voltage switchgear operating condition description vector matrix, serving as the basis for representing the overall operating state of the distributed low-voltage system at the current moment.

[0058] In summary, the embodiments of this application have at least the following technical effects:

[0059] First, the topology of the distributed low-voltage switchgear assembly is extracted at the equipment layer, loop layer, and system layer, resulting in equipment layer topology, loop layer topology, and system layer topology. Next, based on the equipment layer topology, physical interface parameters of the equipment are collected, resulting in a set of equipment physical interface parameters; based on the loop layer topology, feeder loop states are collected, resulting in a set of feeder loop state features. Further, based on the system layer topology, adjacency state analysis is performed between adjacent distributed low-voltage switchgear based on the set of equipment physical interface parameters, resulting in a set of adjacency state matrices. Finally, the set of feeder loop state features and the set of adjacency state matrices are used as real-time low-voltage switchgear operating conditions. These real-time operating conditions are matched with a pre-built distributed control library to determine a target distributed control scheme. Distributed control of the distributed low-voltage switchgear assembly is then performed based on this target distributed control scheme. This approach solves the technical problems of poor flexibility and insufficient adaptability of distributed control for low-voltage switchgear in existing technologies, achieving the technical effect of intelligent adaptive control based on real-time operating conditions.

[0060] Example 2 is based on the same inventive concept as the low-voltage switchgear distributed control method based on multi-source data fusion in the previous examples, such as... Figure 2 As shown, this application provides a distributed control system for low-voltage switchgear based on multi-source data fusion, wherein the system includes:

[0061] Topology extraction module 11: Extracts the topology of the distributed low-voltage switchgear set at the equipment layer, loop layer, and system layer to obtain the equipment layer topology, loop layer topology, and system layer topology; Parameter acquisition module 12: Acquires equipment physical interface parameters based on the equipment layer topology to obtain the equipment physical interface parameter set; Loop status acquisition module 13: Acquires feeder loop status based on the loop layer topology to obtain the feeder loop status feature set; Analysis module 14: Analyzes the adjacency status between adjacent distributed low-voltage switchgear based on the equipment physical interface parameter set according to the system layer topology to obtain the adjacency status matrix set; Control module 15: Uses the feeder loop status feature set and the adjacency status matrix set as real-time low-voltage switchgear operating conditions, matches the real-time low-voltage switchgear operating conditions with a pre-built operating condition distributed control library, determines the target distributed control scheme, and performs distributed control on the distributed low-voltage switchgear set based on the target distributed control scheme.

[0062] Furthermore, the parameter acquisition module 12 is used to perform the following method:

[0063] Within a preset acquisition window, the device layer topology is traversed to obtain the device physical interface parameters of each distributed low-voltage cabinet in the distributed low-voltage cabinet set, thus obtaining an initial set of device physical interface parameter sequences. The initial set of device physical interface parameter sequences is then filtered to determine the set of device physical interface parameters.

[0064] Furthermore, the parameter acquisition module 12 is used to perform the following method:

[0065] A coordinate system is constructed with time as the horizontal axis and parameters as the vertical axis. The initial set of device physical interface parameters is input into the coordinate system to obtain a set of data filtering coordinate systems, wherein each data filtering coordinate system has a set of filtering coordinate points. A first data filtering coordinate system and a first set of filtering coordinate points are extracted from the set of data filtering coordinate systems. The first set of filtering coordinate points is filtered in the first data filtering coordinate system using a multi-scale linear rotation filtering method to determine a first target filtering coordinate point, and the first target filtering coordinate point is used as the first device physical interface parameter. The data filtering coordinate system set is then filtered in the corresponding data filtering coordinate system using the multi-scale linear rotation filtering method to determine the set of device physical interface parameters.

[0066] Furthermore, the parameter acquisition module 12 is used to perform the following method:

[0067] The initial data filtering line is a straight line passing through the origin of the first data filtering coordinate system and parallel to the horizontal axis of the first data filtering coordinate system. An initial rotation direction is obtained, and the initial data filtering line is rotated according to a preset rotation scale to obtain a first iterative data filtering line. The reliability coefficients of the first iterative data filtering line and the initial data filtering line are compared to determine the update rotation direction, update rotation scale, and stage data filtering line. The stage data filtering line is rotated according to the update rotation direction and update rotation scale to obtain a second iterative data filtering line. The reliability coefficients of the second iterative data filtering line and the stage data filtering line are compared to update the stage data filtering line, and the update rotation direction and update rotation scale are updated again until a preset number of updates is met to obtain a target data filtering line. Based on the target data filtering line and a preset filtering bandwidth, the first set of filtering coordinate points is filtered to determine the first target filtering coordinate point.

[0068] Furthermore, the parameter acquisition module 12 is used to perform the following method:

[0069] When the reliability coefficient of the first iterative data filtering line is less than or equal to the reliability coefficient of the initial data filtering line, the opposite direction of the initial rotation direction is taken as the updated rotation direction; half of the preset rotation scale is taken as the updated rotation scale, and the initial data filtering line is taken as the stage data filtering line.

[0070] Furthermore, the parameter acquisition module 12 is used to perform the following method:

[0071] The first filter coordinate points in the first set of filter coordinate points whose distance to the target data filter line is within the preset filter bandwidth are added to the first filter neighborhood. The mean value of the first filter coordinate points in the first filter neighborhood is calculated to determine the first target filter coordinate point.

[0072] Furthermore, the analysis module 14 is used to perform the following methods:

[0073] Based on the system layer topology, any two adjacent distributed low-voltage cabinets in the distributed low-voltage cabinet set are extracted to obtain a set of adjacent distributed low-voltage cabinet combinations; combined with the set of equipment physical interface parameters, the adjacency state of each adjacent distributed low-voltage cabinet combination in the set of adjacent distributed low-voltage cabinet combinations is analyzed to obtain the set of adjacency state matrices.

[0074] Furthermore, the analysis module 14 is used to perform the following methods:

[0075] Extract the first adjacent distributed low-voltage switchgear combination from the set of adjacent distributed low-voltage switchgear combinations; extract the set of equipment physical interface parameters based on the first adjacent distributed low-voltage switchgear combination to obtain the first equipment physical interface parameter and the second equipment physical interface parameter; calculate the similarity between the first equipment physical interface parameter and the second equipment physical interface parameter to determine the parameter similarity set; normalize and matrix-process the parameter similarity set to obtain the first adjacency state matrix; perform adjacency state analysis on the adjacent distributed low-voltage switchgear combination set according to the set of equipment physical interface parameters to obtain the adjacency state matrix set.

[0076] Furthermore, the loop status acquisition module 13 is used to perform the following method:

[0077] Obtain preset feeder loop status indicators, wherein the preset feeder loop status indicators include at least current, power, power factor, harmonic content, temperature, insulation resistance, and grounding resistance; based on the preset feeder loop status indicators, perform feeder loop status acquisition on the loop layer topology to obtain a set of feeder loop status features.

[0078] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0079] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0080] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A distributed control method for low-voltage switchgear based on multi-source data fusion, characterized in that, The method includes: The equipment layer, loop layer, and system layer topologies of the distributed low-voltage switchgear assembly are extracted to obtain the equipment layer topology, loop layer topology, and system layer topology. Based on the device layer topology, the device physical interface parameters are collected to obtain the device physical interface parameter set. Based on the aforementioned loop layer topology, the feeder loop status is acquired to obtain a set of feeder loop status features. Based on the system layer topology, the adjacency state analysis between adjacent distributed low-voltage cabinets is performed on the set of physical interface parameters of the equipment to obtain an adjacency state matrix set. The feeder circuit state feature set and the adjacent state matrix set are used as real-time low-voltage switchgear operating conditions. Based on the real-time low-voltage switchgear operating conditions, they are matched with the pre-built operating condition distributed control library to determine the target distributed control scheme. Based on the target distributed control scheme, the distributed low-voltage switchgear set is subjected to distributed control.

2. The low-voltage switchgear distributed control method based on multi-source data fusion as described in claim 1, characterized in that, Based on the device layer topology, device physical interface parameters are collected to obtain a set of device physical interface parameters, including: Within the preset acquisition window, the device physical interface parameters of each distributed low-voltage cabinet in the distributed low-voltage cabinet set are traversed through the device layer topology to obtain the initial set of device physical interface parameter sequences. The initial set of device physical interface parameter sequences is filtered to determine the set of device physical interface parameters.

3. The distributed control method for low-voltage switchgear based on multi-source data fusion as described in claim 2, characterized in that, The initial set of device physical interface parameter sequences is filtered to determine the set of device physical interface parameters, including: A coordinate system is constructed with time as the horizontal axis and parameters as the vertical axis. The set of initial device physical interface parameter sequences is input into the coordinate system to obtain a set of data filtering coordinate systems, wherein each data filtering coordinate system has a set of filtering coordinate points. Extract a first data filtering coordinate system and a first set of filtering coordinate points from the set of data filtering coordinate systems; Using a multi-scale linear rotation filtering method, the first set of filtering coordinate points is filtered in the first data filtering coordinate system to determine the first target filtering coordinate point, and the first target filtering coordinate point is used as the first device physical interface parameter; The data filtering coordinate system set is filtered in the corresponding data filtering coordinate system using the multi-scale linear rotation filtering method to determine the set of physical interface parameters of the device.

4. The distributed control method for low-voltage switchgear based on multi-source data fusion as described in claim 3, characterized in that, Using a multi-scale linear rotation filtering method, a first set of filtering coordinate points is filtered in the first data filtering coordinate system to determine a first target filtering coordinate point, and this first target filtering coordinate point is used as a first device physical interface parameter, including: The line that passes through the origin of the first data filtering coordinate system and is parallel to the horizontal axis of the first data filtering coordinate system is taken as the initial data filtering line. Obtain the initial rotation direction, and rotate the initial data filtering line according to a preset rotation scale to obtain the first iteration data filtering line; By comparing the reliability coefficient of the first iterative data filtering line with the reliability coefficient of the initial data filtering line, the updated rotation direction, the updated rotation scale, and the stage data filtering line are determined. The stage data filtering line is rotated according to the updated rotation direction and updated rotation scale to obtain the second iteration data filtering line; Compare the reliability coefficient of the second iterative data filtering line with the reliability coefficient of the stage data filtering line, update the stage data filtering line, and update the rotation direction and rotation scale again until the preset number of updates is met to obtain the target data filtering line; Based on the target data, the first set of filter coordinate points is filtered using a straight line and a preset filter bandwidth to determine the first target filter coordinate point.

5. The distributed control method for low-voltage switchgear based on multi-source data fusion as described in claim 4, characterized in that, Comparing the reliability coefficient of the first iterative data filtering line with the reliability coefficient of the initial data filtering line, and determining the updated rotation direction, updated rotation scale, and stage data filtering line, includes: When the reliability coefficient of the first iterative data filtering line is less than or equal to the reliability coefficient of the initial data filtering line, the opposite direction of the initial rotation direction is taken as the updated rotation direction. Half of the preset rotation scale is used as the updated rotation scale, and the initial data filtering line is used as the stage data filtering line.

6. The distributed control method for low-voltage switchgear based on multi-source data fusion as described in claim 4, characterized in that, The first filter coordinate points in the first set of filter coordinate points whose distance to the target data filter line is within the preset filter bandwidth are added to the first filter neighborhood. The mean value of the first filter coordinate points in the first filter neighborhood is calculated to determine the first target filter coordinate point.

7. The distributed control method for low-voltage switchgear based on multi-source data fusion as described in claim 1, characterized in that, Based on the system layer topology, an adjacency state analysis is performed on the set of physical interface parameters of the equipment between adjacent distributed low-voltage cabinets to obtain an adjacency state matrix set, including: Based on the system layer topology, any two adjacent distributed low-voltage cabinets in the distributed low-voltage cabinet set are extracted to obtain a set of adjacent distributed low-voltage cabinet combinations. By combining the set of physical interface parameters of the equipment, the adjacency state of each adjacent distributed low-voltage cabinet combination in the set of adjacent distributed low-voltage cabinet combinations is analyzed to obtain the set of adjacency state matrices.

8. The distributed control method for low-voltage switchgear based on multi-source data fusion as described in claim 7, characterized in that, Based on the set of physical interface parameters of the equipment, the adjacency state of each adjacent distributed low-voltage switchgear combination in the set of adjacent distributed low-voltage switchgear combinations is analyzed to obtain the adjacency state matrix set, including: Extract the first adjacent distributed low-voltage switchgear combination from the set of adjacent distributed low-voltage switchgear combinations; Based on the first adjacent distributed low-voltage switchgear combination, the set of physical interface parameters of the equipment is extracted to obtain the first physical interface parameters of the equipment and the second physical interface parameters of the equipment. Calculate the similarity between the physical interface parameters of the first device and the physical interface parameters of the second device, and determine the parameter similarity set; The parameter similarity set is normalized and matrixed to obtain the first adjacency state matrix; For each of the adjacent distributed low-voltage switchgear combinations, an adjacency state analysis is performed based on the set of equipment physical interface parameters to obtain the adjacency state matrix set.

9. The distributed control method for low-voltage switchgear based on multi-source data fusion as described in claim 1, characterized in that, Based on the aforementioned loop layer topology, feeder loop status is acquired to obtain a set of feeder loop status features, including: Obtain preset feeder circuit status indicators, wherein the preset feeder circuit status indicators include at least current, power, power factor, harmonic content, temperature, insulation resistance and grounding resistance; Based on the preset feeder loop status index, the feeder loop status of the loop layer topology is collected to obtain a set of feeder loop status features.

10. A distributed control system for low-voltage switchgear based on multi-source data fusion, characterized in that, For implementing the low-voltage switchgear distributed control method based on multi-source data fusion as described in any one of claims 1-9, the system comprises: Topology extraction module: Extracts the topology of the distributed low-voltage switchgear set at the equipment layer, loop layer, and system layer to obtain the equipment layer topology, loop layer topology, and system layer topology; Parameter acquisition module: Based on the device layer topology, it acquires device physical interface parameters to obtain a set of device physical interface parameters; Loop status acquisition module: Based on the loop layer topology, it acquires the feeder loop status and obtains a set of feeder loop status features; Analysis module: Based on the system layer topology, the adjacency state analysis of adjacent distributed low-voltage cabinets is performed on the set of physical interface parameters of the equipment to obtain an adjacency state matrix set; Control module: The feeder circuit state feature set and the adjacent state matrix set are used as real-time low-voltage switchgear operating conditions. Based on the real-time low-voltage switchgear operating conditions, the module matches them with the pre-built operating condition distributed control library to determine the target distributed control scheme. Based on the target distributed control scheme, the module performs distributed control on the distributed low-voltage switchgear set.

Images