Photovoltaic special circuit breaker system and method

Abstract

The invention mainly relates to the technical field of direct current circuit breakers, and discloses a photovoltaic special circuit breaker system and method, and the system comprises a signal monitoring module, an abnormity recognition module, a path diagnosis module, a fault derivation module, an isolation control module, a path recovery module, and an action optimization module. According to the system, detailed parameters of path conduction are formed by extracting current amplitudes and direction changes of adjacent nodes, abnormal nodes can be accurately positioned, the conduction path range is marked, and the abnormal node fluctuation range, the amplitude change direction and the input and output deviation in the path are comprehensively calculated in combination with an isolated forest algorithm. According to the method, a specific path and an end point range of abnormal propagation can be quickly analyzed, so that the circuit breaker has quick response, multi-level diagnosis and optimization adjustment capabilities in a complex scene, and the safety, the reliability and the operation efficiency of a direct current system are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of direct current circuit breakers, and in particular to a photovoltaic dedicated circuit breaker system and method. Background Art

[0002] The field of DC circuit breaker technology aims to study, develop and apply switchgear suitable for DC circuits. The core goal is to achieve safe disconnection and control of DC current, especially under high voltage and high current conditions.

[0003] The purpose of photovoltaic dedicated circuit breakers is to achieve reliable circuit breaking, overload protection and short-circuit protection for DC circuits in photovoltaic systems. By using photovoltaic dedicated circuit breakers, abnormal current paths can be quickly cut off during power transmission and distribution, effectively preventing safety accidents caused by overload and short circuit.

[0004] The existing technology judges anomalies through simple current amplitude thresholds, lacks in-depth analysis of current fluctuation characteristics and time series data, and has difficulty distinguishing abnormal patterns in complex situations. The recognition delay of sudden current anomalies is high, resulting in some abnormal signals not being captured in the early stages, causing the fault area to expand. Path isolation mostly relies on fixed rules rather than dynamic adjustment, and does not fully consider the real-time impact of isolation operations on current distribution and path load. Some paths may be in a high-load state for a long time, reducing the overall service life and failing to fully guarantee the safety and operating efficiency of the DC circuit in the photovoltaic system. Summary of the invention

[0005] In view of the above-mentioned prior art problems, the present invention is proposed.

[0006] The purpose of the present invention is to provide a photovoltaic dedicated circuit breaker system, which aims to solve the problems of delayed current abnormality recognition, expanded fault area, inflexible path isolation, and insufficient safety and efficiency in existing DC circuit breakers.

[0007] To solve the above technical problems, the present invention provides the following technical solutions: A photovoltaic dedicated circuit breaker system, comprising a signal monitoring module, for real-time acquisition of current time series data in a DC circuit, and processing and generating current waveform data containing amplitude and timing information;

[0008] an abnormality identification module, which analyzes and marks abnormal features in the current fluctuation based on the current waveform data to form abnormal feature data;

[0009] A path diagnosis module, which determines the conduction path and position of the abnormal current in the circuit according to the abnormal characteristic data;

[0010] The fault derivation module is used to calculate the propagation end point of the abnormal signal using the isolation forest algorithm based on the abnormal path identification result;

[0011] Isolation control module, based on fault location information, selectively implements circuit isolation to ensure normal operation of non-fault areas;

[0012] A path recovery module adjusts the circuit load distribution after isolation and reconfigures the path to restore normal power transmission;

[0013] The action optimization module optimizes the action strategy of the circuit breaker to minimize the impact on the circuit.

[0014] As a preferred solution of the photovoltaic dedicated circuit breaker system of the present invention, the current waveform data set includes recorded current amplitude sequences, current cycle segments, and extreme point information of each fluctuation segment.

[0015] As a preferred solution of the photovoltaic dedicated circuit breaker system of the present invention, the abnormality identification module includes an abnormal value screening submodule, a jump interval marking submodule, and an abnormal amplitude marking submodule.

[0016] As a preferred solution of the photovoltaic dedicated circuit breaker system of the present invention, the path diagnosis module includes a node abnormality marking submodule, a path segmentation screening submodule, and an abnormal path generation submodule.

[0017] As a preferred solution of the photovoltaic dedicated circuit breaker system of the present invention, the fault derivation module includes an abnormal node extraction submodule, an abnormal propagation range submodule, and an end point range determination submodule.

[0018] As a preferred solution of the photovoltaic dedicated circuit breaker system of the present invention, wherein: the isolation forest algorithm is calculated according to a specific formula;

[0019]

[0020] The anomaly score of the node is evaluated by combining factors such as fluctuation amplitude difference and time offset trend.

[0021] As a preferred solution of the photovoltaic dedicated circuit breaker system of the present invention, the isolation control module includes an isolation point analysis submodule, a current distribution calculation submodule, and an isolation node configuration submodule.

[0022] As a preferred solution of the photovoltaic dedicated circuit breaker system of the present invention, wherein: the path recovery module includes a load distribution adjustment submodule, a path connectivity update submodule, and a new path marking submodule;

[0023] The action optimization module includes a load change analysis submodule, an action priority calculation submodule, and a timing path generation submodule.

[0024] Another object of the present invention is to provide a photovoltaic dedicated circuit breaker method, comprising the following steps:

[0025] Collect the current time series data in the DC circuit in real time and generate current waveform data containing amplitude and timing information;

[0026] Based on the current waveform data, analyzing and marking abnormal features in the current fluctuation to form abnormal feature data;

[0027] Determining the conduction path and position of the abnormal current in the circuit according to the abnormal characteristic data;

[0028] Combine path conduction information with system topology to predict and locate fault propagation range and endpoint;

[0029] Based on the fault location information, selectively implement circuit isolation to ensure normal operation of non-fault areas;

[0030] Adjust circuit load distribution after isolation and reconfigure paths to restore normal power delivery;

[0031] Optimize the circuit breaker operation strategy to minimize the impact on the circuit.

[0032] As a preferred solution of the photovoltaic dedicated circuit breaker method of the present invention, wherein: an isolation forest algorithm is used to analyze the fluctuation range, amplitude change direction and input-output deviation of abnormal nodes, and the specific range and end point of the abnormal signal transmission to the downstream path are quickly calculated;

[0033] The difference threshold is dynamically set to screen out points whose amplitude difference exceeds the threshold as preliminary abnormal points, and by checking the continuity of the amplitude change direction of the preliminary abnormal points, points with the same change direction are marked as abnormal amplitude increments, thereby achieving accurate identification of sudden current anomalies.

[0034] The beneficial effect of the photovoltaic dedicated circuit breaker system of the present invention is that by regularly evaluating the channel quality and automatically switching to the optimal channel with low interference, the risk of channel congestion and signal overlap is significantly reduced, ensuring the stability and efficiency of data transmission. The application layer data is quickly forwarded by the carrier node and arranged in chronological order, which improves the data processing speed and transmission efficiency. The status of all nodes is continuously monitored, and the network topology is adjusted by sending beacons when an isolated node is detected, so that the path interruption self-healing is realized, and the continuity and reliability of the network connection are ensured. When the node loses power or the communication is interrupted, it quickly switches to the backup channel to ensure that the data transmission is not interrupted, and generates a detailed fault report to assist maintenance, shortening the fault repair time. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings of the embodiments of the present invention are briefly introduced below. Obviously, the drawings described below only relate to some embodiments of the present invention, but are not intended to limit the present invention. Among them:

[0036] Figure 1 It is a schematic diagram of the photovoltaic dedicated circuit breaker system of the present invention.

[0037] Figure 2 It is a schematic diagram of the photovoltaic dedicated circuit breaker method of the present invention. DETAILED DESCRIPTION

[0038] In order to enable those skilled in the art to better understand the present invention, the present invention is further described in detail below in conjunction with specific implementation methods and drawings.

[0039] The terms used in the present invention are those general terms currently widely used in the art in consideration of the functions of the present invention, but these terms may vary according to the intention of a person of ordinary skill in the art, precedents, or new technologies in the art. In addition, specific terms may be selected by the applicant, and in this case, their detailed meanings will be described in the detailed description of the present invention. Therefore, the terms used in the specification should not be understood as simple names, but rather as a general description based on the meaning of the terms and the present invention.

[0040] Reference Figure 1 , this embodiment provides a photovoltaic dedicated circuit breaker system, including a signal monitoring module 1, which is used to collect current time series data in a DC circuit in real time, and process and generate current waveform data containing amplitude and timing information;

[0041] The abnormality identification module 2 analyzes and marks abnormal features in the current fluctuation based on the current waveform data to form abnormal feature data;

[0042] The path diagnosis module 3 determines the conduction path and position of the abnormal current in the circuit according to the abnormal characteristic data;

[0043] The fault derivation module 4 is used to calculate the propagation end point of the abnormal signal by using the isolation forest algorithm based on the abnormal path identification result;

[0044] The isolation control module 5 selectively implements circuit isolation based on the fault location information to ensure normal operation of the non-fault area;

[0045] a path recovery module 6, which adjusts the circuit load distribution after isolation and reconfigures the path to restore normal power transmission;

[0046] The action optimization module 7 optimizes the action strategy of the circuit breaker to minimize the impact on the circuit.

[0047] In this embodiment, please refer to Figure 1 The signal monitoring module 1 extracts the time series signal output by the current sensor based on the main circuit and branch line of the circuit breaker, records the amplitude and time difference point by point, separates the periodic fluctuation segment based on the timing characteristics, screens the continuous fluctuation range and determines the positive and negative direction changes of the current, obtains the extreme amplitude and corresponding time difference of each fluctuation segment, and generates a current waveform data set;

[0048] The abnormal identification module 2 analyzes the amplitude change of the mutation point in the waveform and the difference between the corresponding time points based on the current waveform data set, calculates the amplitude increment and direction change of the abnormal value in each continuous waveform, selects the amplitude jump interval that conforms to the specific change law, identifies the abnormal characteristics within the fluctuation range, combines the difference in the adjacent waveform changes of the continuous abnormal points, marks the abnormal waveform interval, and generates abnormal amplitude marking data;

[0049] Based on the abnormal amplitude marking data, the path diagnosis module 3 extracts the current fluctuation parameters and conduction direction of adjacent nodes in the path, calculates the amplitude difference of the abnormal current transmitted in the path, determines the abnormal nodes in the conduction path and marks the specific positions, analyzes the amplitude changes and direction consistency before and after the abnormal fluctuations, screens the segmented range of abnormal amplitude conduction in the path, and generates abnormal path identification results;

[0050] In this embodiment, the fault derivation module 4 uses the isolation forest algorithm based on the abnormal path identification result to extract the fluctuation range and amplitude change direction of the abnormal node in the path, and combines the input and output fluctuation difference and offset amplitude around the abnormal node to calculate the possible range of the abnormal signal transmission to the downstream of the path, extract the continuous abnormal fluctuation nodes downstream of the path, and judge the end point range of the abnormal propagation in combination with the topological distribution, and generate the fault propagation end point position data;

[0051] Based on the fault propagation endpoint location data, the isolation control module 5 analyzes the fluctuation current amplitude and direction change of the isolation point, calculates the current distribution change range of the nodes near the isolation point, determines the impact of the isolation operation on the fluctuation conduction of the adjacent path, extracts the path set that can remain connected after isolation, selects the path set with the lowest current offset as the isolation point range, and generates an isolation node configuration plan;

[0052] The path recovery module 6 calculates the current load distribution in the isolated path based on the isolated node configuration scheme, extracts the load change amplitude of each node in the remaining path, adjusts the connectivity relationship of the load path to form a new path distribution, calculates the current delay and load distribution of the new path, marks the new load path and node relationship, and generates a path recovery topology result;

[0053] Based on the path recovery topology results, the action optimization module 7 analyzes the load changes and remaining capacity distribution within the path, extracts the change amplitude of the load current during the circuit breaker action, calculates the delay effect of the action execution order on the path load distribution, determines the priority order and interval time of the action execution, screens the formed dynamic adjustment order and marks the action paths of each order, and generates a circuit breaker action timing list.

[0054] In one embodiment provided in the present application, the current waveform data set includes a recorded current amplitude sequence, current cycle segments, and extreme point information of each fluctuation segment.

[0055] As an optional embodiment, the abnormality identification module 2 includes an abnormal value screening submodule, a jump interval marking submodule, and an abnormal amplitude marking submodule.

[0056] As an optional embodiment, the path diagnosis module 3 includes a node abnormality marking submodule, a path segmentation screening submodule, and an abnormal path generation submodule.

[0057] As an optional embodiment, the fault deduction module 4 includes an abnormal node extraction submodule, an abnormal propagation range submodule, and an end point range determination submodule.

[0058] As an optional embodiment, the isolation forest algorithm is calculated according to a specific formula;

[0059] The anomaly score of the node is evaluated by combining factors such as fluctuation amplitude difference and time offset trend.

[0060]

[0061] As an optional embodiment, the isolation control module 5 includes an isolation point analysis submodule, a current distribution calculation submodule, and an isolation node configuration submodule.

[0062] As an optional embodiment, the path recovery module 6 includes a load distribution adjustment submodule, a path connectivity update submodule, and a new path marking submodule;

[0063] The action optimization module 7 includes a load change analysis submodule, an action priority calculation submodule, and a timing path generation submodule.

[0064] In this embodiment, please refer to Figure 1 The signal monitoring module 1 includes a time series separation submodule, an extreme value extraction submodule, and a waveform data generation submodule, wherein:

[0065] The time series separation submodule extracts the time series data from the signal based on the time series signal output by the current sensor, identifies the starting and ending points of the signal fluctuation by recording the amplitude and time difference of the current signal point by point, determines the periodic fluctuation range by comparing the amplitude change trend of adjacent recording points, and marks the continuous interval of the fluctuation segment by screening the amplitude change law to generate the fluctuation range marking data;

[0066] The extreme value extraction submodule is based on the fluctuation range marking data, and screens out the maximum and minimum amplitudes by analyzing the amplitude fluctuation characteristics in the fluctuation segment. It extracts the key points of amplitude change by associating the time points corresponding to the extreme values, and marks the extreme value features by establishing a mapping relationship between amplitude and time, thus generating fluctuation extreme value feature data.

[0067] The waveform data generation submodule is based on the fluctuation extreme value characteristic data. It constructs a complete time series waveform of the fluctuation segment by interpolating the time and amplitude sequences in the characteristic data. It generates complete current fluctuation data by connecting the interpolation results of continuous fluctuation segments, obtains the current waveform data set, and generates the current waveform data set.

[0068] In this embodiment, the time series separation submodule uses a signal processing algorithm to segment the original time series data based on the time series signal output by the current sensor, obtains the fluctuation characteristics of each point by calculating the time interval and amplitude difference between adjacent recording points, uses the sliding window method to analyze the amplitude change trend and mark the starting point and end point of the fluctuation segment, filters the fluctuation range by setting the amplitude change threshold, uses the amplitude difference method to calculate the change rate of adjacent recording points and mark the trend turning point, and generates the fluctuation range marking data;

[0069] The extreme value extraction submodule marks the data based on the fluctuation range, uses the local extreme value algorithm to set the local extreme value window to screen the maximum and minimum amplitudes in the fluctuation segment, calculates the extreme values ​​by traversing all the record points in the fluctuation segment and marks the extreme value time points, uses the time point interpolation method to fill in the key points of the amplitude change to ensure the integrity of the time series, establishes the mapping relationship between the amplitude and the time point, and stores the extreme value feature points through the mapping function to generate the fluctuation extreme value feature data;

[0070] The waveform data generation submodule uses a polynomial interpolation algorithm to perform smooth interpolation operations on the time series in the characteristic data based on the fluctuation extreme value characteristic data and generates time series data of continuous fluctuation segments. The smooth transition of the connection segments is ensured by calculating the amplitude change ratio of the connection points of the continuous fluctuation segments. The complete current waveform data is synthesized by connecting the interpolation results of all fluctuation segments. The generated waveform data is subjected to noise reduction processing using a Gaussian filter to generate a current waveform data set.

[0071] The current waveform data set includes the recorded current amplitude sequence, the extracted current cycle segments and the extreme point information of each fluctuation segment.

[0072] In this embodiment, the abnormality identification module 2 includes an abnormal value screening submodule, a jump interval marking submodule, and an abnormal amplitude marking submodule.

[0073] The outlier screening submodule is based on the current waveform data set. It records the characteristics of the mutation points by extracting the amplitude and time difference of the mutation points in the waveform sequence, screens the abnormal amplitude increment by analyzing the amplitude changes of continuous segments in the waveform, and marks the amplitude jump characteristic interval by marking the direction change law, generating the amplitude jump interval data.

[0074] The jump interval marking submodule is based on the amplitude jump interval data. It extracts the amplitude and time association of the abnormal points by analyzing the fluctuation characteristics in the jump interval, marks the abnormal interval by comparing the amplitude difference of adjacent fluctuation intervals, and marks the abnormal fluctuation range by integrating the changes in the continuous interval to generate abnormal waveform interval data.

[0075] The abnormal amplitude marking submodule is based on the abnormal waveform interval data. It records the amplitude change by analyzing the amplitude jump characteristics within the marking interval, marks the abnormal amplitude by associating the time point and the amplitude fluctuation data, obtains the abnormal fluctuation result by screening the jump characteristics and clarifying the marking range, and generates the abnormal amplitude marking data;

[0076] In this embodiment, the abnormal value screening submodule is based on the current waveform data set, and uses the mutation point detection algorithm to scan the waveform sequence point by point, extracts the mutation point characteristics by setting the amplitude change threshold and the time interval threshold, calculates the amplitude change rate of adjacent sampling points by using the point-by-point difference method and marks the points whose change rate is greater than the set threshold, screens the abnormal amplitude increment by extracting the cumulative difference of the amplitude change in the continuous segment, and uses the direction change analysis method to mark the jump characteristic interval by comparing the positive and negative directions of the amplitude change, and generates the amplitude jump interval data;

[0077] The jump interval marking submodule is based on the amplitude jump interval data. It uses the fluctuation analysis algorithm to calculate the amplitude and time series of the jump interval point by point. It establishes data comparison by extracting the amplitude and time association of each abnormal point. It uses the amplitude difference calculation method to analyze the amplitude difference of adjacent fluctuation intervals point by point and mark the intervals with differences greater than the set threshold. It integrates continuous intervals by accumulating the change trend of the marked points to generate abnormal waveform interval data.

[0078] The abnormal amplitude marking submodule is based on the abnormal waveform interval data and uses the jump characteristic analysis method to analyze the amplitude change data of the marked interval. It extracts the maximum change of the amplitude jump point by point and records the abnormal characteristics at the corresponding time point, and uses the correlation analysis method of the time point and the amplitude sequence to mark the abnormal amplitude. The characteristic screening algorithm is used to verify the amplitude change and direction of the jump point one by one and clarify the marking range to generate abnormal amplitude marking data;

[0079] The abnormal amplitude marking data includes the start and end time of the marked abnormal interval, the amplitude increment of each abnormal point, and the amplitude difference value of continuous abnormal waveforms.

[0080] Filter abnormal amplitude increments, extract amplitude data point by point from the time series waveform, generate a difference sequence by calculating the amplitude difference between two adjacent points and mark the positions where mutations may occur, dynamically set the difference threshold according to the overall amplitude change characteristics of the waveform and the background noise level, filter out points with amplitude differences exceeding the threshold as preliminary abnormal points, check the continuity of the amplitude change direction of the preliminary abnormal points, mark the points with the same change direction as abnormal amplitude increments, and filter out the fluctuation range with significant jump characteristics by analyzing the time interval and amplitude change difference of the abnormal points, and mark the complete abnormal amplitude increment range.

[0081] In this embodiment, the path diagnosis module 3 includes a node abnormality marking submodule, a path segmentation screening submodule, and an abnormal path generation submodule.

[0082] The node anomaly marking submodule extracts the current fluctuation parameters and conduction directions of adjacent nodes in the path based on the abnormal amplitude marking data, records the fluctuation characteristics between nodes by comparing the current amplitude changes between adjacent nodes, determines the amplitude abnormal nodes by calculating the amplitude gap between nodes, identifies the nodes with sudden changes in fluctuation characteristics by combining the continuity of the conduction direction, marks the specific locations of the abnormal nodes, and generates abnormal node marking data;

[0083] The path segmentation screening submodule extracts the change trend of the fluctuation amplitude and conduction direction before and after the marked node based on the abnormal node marking data, screens out the local segment range of the abnormal current by analyzing the fluctuation characteristics of adjacent nodes in the path, and generates abnormal conduction segmentation data by connecting continuous abnormal nodes to mark the complete fluctuation segmentation;

[0084] The abnormal path generation submodule is based on the abnormal conduction segmentation data. It determines the segmentation fluctuation characteristics by integrating the fluctuation amplitude records of each node in the path segment, builds the abnormal path by matching the data of the abnormal nodes within the segmentation range, summarizes the overall distribution of the abnormal area by marking the abnormal segments in the path, and generates the abnormal path identification result;

[0085] The node anomaly marking submodule uses the amplitude difference algorithm to analyze the current fluctuation parameters of adjacent nodes in the path based on the abnormal amplitude marking data, extracts the fluctuation characteristics by calculating the difference value of the current amplitude between adjacent nodes, calculates the continuity of the node conduction direction by using the direction consistency algorithm, marks the amplitude abnormal nodes by setting the amplitude difference threshold, and identifies the characteristic mutation nodes by comparing the amplitude change trend in combination with the fluctuation mutation detection method, and marks the specific location of the node point by point to generate abnormal node marking data;

[0086] In this embodiment, the path segmentation screening submodule uses a trend analysis algorithm to extract the fluctuation amplitude changes before and after the marked node point by point based on the abnormal node marking data, and screens the local segment range of the abnormal current by setting the amplitude change trend threshold. The segment connection algorithm is used to logically connect adjacent abnormal nodes to generate a continuous fluctuation interval, and the interval is marked by integrating the fluctuation amplitude trend of each segment to generate abnormal conduction segment data;

[0087] The abnormal path generation submodule is based on the abnormal conduction segmented data. It uses the path integration algorithm to analyze the node fluctuation amplitude records in each segment, extracts the amplitude characteristics of each segment and the node data to construct the segmented fluctuation characteristics, uses the abnormal matching algorithm to match the abnormal amplitude data of the nodes segment by segment to generate abnormal paths, and uses the path summary algorithm to mark the complete distribution of abnormal areas for each segment connection to generate abnormal path identification results.

[0088] Among them, the abnormal path identification results include the specific location of the abnormal node in the path, the transmission range of the fluctuation amplitude, and the starting and ending abnormal points in the path.

[0089] In this embodiment, the fault deduction module 4 includes an abnormal node extraction submodule, an abnormal propagation range submodule, and an end point range determination submodule.

[0090] The abnormal node extraction submodule uses the isolation forest algorithm based on the abnormal path identification results to extract the fluctuation amplitude and direction of each abnormal node in the path, records the node fluctuation deviation by comparing the fluctuation input and output differences between nodes, analyzes the correlation between nodes by calculating the amplitude change range of adjacent nodes, and generates abnormal node fluctuation data;

[0091] The abnormal propagation range submodule is based on abnormal node fluctuation data. It analyzes the fluctuation difference between nodes in the path to clarify the downstream propagation range, marks the downstream path by screening the node chain of continuous abnormal fluctuations, and analyzes the boundary of downstream propagation by combining the path topology structure to generate abnormal propagation range data.

[0092] The end range determination submodule determines the end of the propagation path based on the abnormal propagation range data by integrating the fluctuation amplitude and correlation characteristics of the downstream nodes, marks the propagation end point by analyzing the fluctuation characteristics of the terminal node, and records the end range data by extracting the fluctuation mark of the terminal path to generate the fault propagation end position data;

[0093] Based on the abnormal path identification results, the abnormal node extraction submodule uses the isolation forest algorithm to analyze the node fluctuation characteristics within the path point by point. By constructing a multidimensional data input set of abnormal node characteristics, including fluctuation amplitude, conduction direction, fluctuation input and output difference, and fluctuation amplitude range of adjacent nodes, the parameter values ​​of the isolation forest algorithm are set to the number of 100 trees and the size of 256 sample subsets. The training model is used to perform abnormal scoring on the fluctuation data of each node. By setting the scoring threshold, the abnormal nodes are extracted and the fluctuation deviation is recorded. The correlation index is calculated in combination with the input and output data of the abnormal nodes, and the fluctuation characteristics of the adjacent nodes are annotated to generate abnormal node fluctuation data.

[0094] The abnormal propagation range submodule is based on the abnormal node fluctuation data. It uses the propagation path analysis algorithm to calculate the fluctuation difference of adjacent nodes in the path. By calculating the fluctuation difference point by point and setting the downstream propagation threshold to 10% of the fluctuation difference, it selects the node chains that meet the conditions and marks them as continuous abnormal node chains. The fluctuation data in the node chain is segmented and accumulated to clarify the propagation range. The path topology analysis algorithm is combined to compare the path topology structure point by point. The boundary nodes of the propagation are marked by calculating the boundary fluctuation difference of the propagation path to generate abnormal propagation range data.

[0095] The end point range determination submodule is based on the abnormal propagation range data and uses the terminal analysis algorithm to integrate the fluctuation amplitude of the downstream nodes point by point. The end of the propagation path is determined by analyzing the fluctuation amplitude change rate. The propagation end point is marked by extracting the fluctuation direction change record of the terminal node. The fluctuation duration of the terminal path is analyzed using the time series data recorded by the fluctuation mark. The final node of the propagation is determined by calculating the fluctuation intensity and amplitude change range within the end point range, and the fault propagation end point position data is generated.

[0096] The fault propagation endpoint location data includes the propagation path range of the fault signal, the node location of the fault endpoint, and link distribution information related to the topology.

[0097] Isolation forest algorithm, according to the formula:

[0098]

[0099] Where: h(x) is the path length of node x in the isolated tree, c(n) is the normalization factor, and w a is the importance weight of path depth, wb is the weight of the fluctuation amplitude difference, f(x,y) is the fluctuation amplitude difference function between node x and adjacent node y, and w c is the importance weight of the time shift trend, g(t) is the time series fluctuation direction shift function, w d is the amplitude change impact weight, v(x) is the amplitude fluctuation of node x;

[0100] Execution process: First, determine the path length h(x) of node x, calculate the path depth, which represents the number of times the node is divided from the root node to the leaf node, and combine the path depth importance weight w a , adjust the impact ratio on the anomaly score, w a The value of can be determined by analyzing the skewness of node distribution and fitting through experiments, and then calculating the difference in fluctuation amplitude between node x and adjacent node y, and quantitatively expressing the difference in fluctuation amplitude through amplitude change function, and combining the difference in fluctuation amplitude to influence weight w b , controlling the contribution of the difference to the anomaly score, w b The value of is determined by comparing the normal fluctuations and abnormal fluctuations in the statistical historical data. Then the fluctuation direction offset g(t) in the time series is analyzed, and the fluctuation direction change trend based on the node x at time t is calculated, combined with the time offset importance weight w c , adjust the impact of fluctuation direction on anomaly score, weight w c Through the correlation analysis between the fluctuation direction and the consistency of abnormal annotation, the amplitude fluctuation amplitude v(x) of the node is quantified, and the weight w is affected by the amplitude change amplitude. d , adjust the abnormal detection ratio of the fluctuation range, weight w d The difference in amplitude distribution between normal nodes and abnormal nodes is statistically obtained, and finally the anomaly score is calculated based on the comprehensive parameters to generate abnormal node fluctuation data.

[0101] In this embodiment, the isolation control module 5 includes an isolation point analysis submodule, a current distribution calculation submodule, and an isolation node configuration submodule.

[0102] The isolation point analysis submodule extracts the fluctuation current amplitude and conduction direction of the nodes near the isolation point based on the fault propagation endpoint location data, identifies the fluctuation conduction boundary by comparing the current amplitude changes of adjacent nodes, determines the characteristics of the isolation point by recording the nodes with changes in conduction direction, and marks the isolation point position by integrating the node characteristics within the fluctuation boundary to generate the isolation point fluctuation data;

[0103] The current distribution calculation submodule extracts the current amplitude and direction changes of the nodes around the isolation point based on the isolation point fluctuation data, analyzes the impact of the isolation point on the fluctuation of the adjacent path by calculating the current change distribution range in the load conduction path, and records the isolation point adaptation range by integrating the load changes between nodes to generate the isolation point current distribution data;

[0104] The isolation node configuration submodule extracts a set of paths that can maintain connectivity based on the isolation point current distribution data, marks the configuration range of the isolation point by screening the path with the smallest load offset, generates key tags of the configuration nodes by recording the load data of key nodes in the path set, and generates an isolation node configuration plan;

[0105] The isolation point analysis submodule is based on the fault propagation endpoint location data and uses the fluctuation boundary detection algorithm to extract the current fluctuation amplitude and conduction direction of the nodes near the isolation point. The fluctuation conduction boundary is identified by calculating the current amplitude difference of adjacent nodes point by point. The direction change detection method is used to record the conduction direction change between nodes to mark the boundary characteristics. The node feature matrix is ​​established by integrating the node current change amplitude within the fluctuation boundary. The feature screening algorithm is used to mark the characteristics of the isolation point through matrix operation, and the isolation point is finally located in combination with the fluctuation range mark to generate the isolation point fluctuation data.

[0106] In this embodiment, the current distribution calculation submodule uses a path current distribution algorithm to analyze the current amplitude and direction changes of nodes around the isolation point based on the isolation point fluctuation data, calculates the load change range of nodes in the path point by point by constructing a current distribution matrix of the load conduction path, calculates the impact range on adjacent paths by combining the fluctuation characteristics of the isolation point using the path distribution model, extracts the load impact distribution of the isolation point by accumulating and analyzing the load data between nodes, and generates the isolation point current distribution data in combination with the node current change record;

[0107] The isolation node configuration submodule uses a path screening algorithm to extract a set of paths that can remain connected based on the isolation point current distribution data. The path set with the smallest load offset is screened by calculating the load offset point by point in the path to mark the configuration range. A key marking point set is established in combination with the load data of key nodes in the path set. A path connectivity matrix is ​​generated by recording the current load changes of each key marking point and the key marking of the configuration node is output, finally generating an isolation node configuration scheme.

[0108] The isolation node configuration scheme includes the selection range of the isolation node, the affected adjacent path set and the remaining load distribution information in the isolation state.

[0109] In this embodiment, the node path recovery module includes a load distribution adjustment submodule, a path connectivity update submodule, and a new path marking submodule.

[0110] The load distribution adjustment submodule extracts the current load distribution in the isolation path based on the isolation node configuration scheme, records the node load offset by comparing the load changes of each node before and after isolation, adjusts the load distribution records between nodes by integrating the path load data, and generates the isolation path load data;

[0111] The path connectivity update submodule extracts the load fluctuation range of the nodes in the remaining paths based on the isolated path load data, redistributes the load by adjusting the connectivity relationship between the nodes in the paths, and marks the load adjustment status in the paths by recording the changes in the current delay in the paths, thereby generating path connectivity adjustment data;

[0112] The new path labeling submodule is based on the path connectivity adjustment data. It labels the new path connectivity by analyzing the load distribution relationship of the nodes in the path, labels the node load status in each path by recording the current changes between the nodes, and forms a complete connectivity topology by integrating the node and load characteristics in the path, thus generating the path recovery topology result.

[0113] The load distribution adjustment submodule is based on the isolation node configuration scheme and uses a load offset calculation algorithm to extract the current load data in the isolation path point by point. The load offset is calculated by comparing the load changes before and after isolation node by node. The load offset of all nodes in the path is cumulatively analyzed using an offset integration algorithm and the load distribution data between nodes is recorded. The load transfer relationship between nodes is adjusted to balance the load distribution, and the isolation path load data is generated by integrating the current load data of the path.

[0114] In this embodiment, the path connectivity update submodule uses a dynamic path adjustment algorithm to extract the node load fluctuation range in the remaining path based on the isolated path load data, analyzes the load connectivity state point by point by establishing a connectivity relationship matrix between nodes, and gradually optimizes the node connectivity relationship in the path using the minimum path load distribution method. The path state after load adjustment is marked by recording the change trend of the current delay in the path, and the path connectivity adjustment data is generated in combination with the optimization record of the connectivity relationship.

[0115] The new path annotation submodule is based on the path connectivity adjustment data, uses the node load distribution analysis algorithm to calculate the distribution relationship of the node load data in the path, establishes the path connectivity state by annotating the load changes between nodes point by point, and uses the topology reconstruction algorithm combined with the node load distribution characteristics to generate a new path connectivity matrix. By integrating the nodes and load characteristics in the path, a complete connectivity topology structure is generated and the load status of each node is recorded to generate the path recovery topology result.

[0116] The path recovery topology result includes the load path allocation scheme after recovery, the node connection relationship in the new path, and the load delay information of each node in the path.

[0117] In this embodiment, the dynamic action optimization module includes a load change analysis submodule, an action priority calculation submodule, and a timing path generation submodule.

[0118] The load change analysis submodule extracts the load data and remaining capacity within the path based on the path recovery topology results, records the load change of each node by comparing the load current between nodes, calculates the overall capacity fluctuation range by analyzing the load distribution ratio between nodes, integrates the capacity distribution data of the nodes by marking the load change range, and generates load change capacity data;

[0119] The action priority calculation submodule extracts the load current change amplitude during the circuit breaker action process based on the load change capacity data, records the load offset range by gradually calculating the load delay distribution of the nodes in the path after the action is triggered, marks the action priority by analyzing the order of the action on the node load change, integrates the priority data by screening the time intervals between each action, and generates the action priority data;

[0120] The timing path generation submodule extracts the dynamic adjustment nodes corresponding to each action in the priority order rules based on the action priority data, records the action path by comparing the load change amplitude between nodes, marks the time node of each action by integrating the action path and dynamic adjustment rules, generates a complete action adjustment list by aggregating the dynamic adjustment process, and generates the circuit breaker action timing list;

[0121] In this embodiment, the load change analysis submodule uses a capacity fluctuation calculation algorithm to extract the load data and remaining capacity in the path point by point based on the path recovery topology result, calculates the load change of each node by comparing the recorded data of the load current node by node, calculates the overall capacity fluctuation range by using the distribution ratio analysis algorithm through the load current distribution ratio between nodes, uses the range marking method to mark the load change range of the node and integrates the node capacity distribution data, and finally generates the load change capacity data;

[0122] The action priority calculation submodule is based on the load change capacity data and uses a dynamic delay allocation algorithm to gradually extract the load current change amplitude during the circuit breaker action process. It calculates the path node load delay after the action is triggered node by node and records the load offset range of each node. It uses a sequential analysis algorithm to gradually analyze the order of the action's impact on the node load change and mark the action priority. It uses a time interval optimization algorithm to gradually screen the trigger interval of each action to generate a dynamic action sequence table, and finally generates action priority data.

[0123] The timing path generation submodule uses the action path aggregation algorithm to extract the dynamic adjustment nodes corresponding to each action in the priority order rules based on the action priority data, records the path sequence of each action by gradually analyzing the load change amplitude between nodes, and uses the rule annotation algorithm to integrate the action path and dynamic adjustment rules to annotate the action time points node by node. The process aggregation method is used to generate a complete adjustment list by analyzing the action rules and paths point by point, and generate the circuit breaker action timing list;

[0124] The circuit breaker action sequence list includes an action sequence list, path markings for each action, and a load adjustment interval corresponding to each action.

[0125] In summary, this system extracts and processes the time series data of the current signal point by point to form a basic data set of dynamic fluctuations, and combines the amplitude change law of the mutation point in the current fluctuation to accurately identify abnormal characteristics and continuous fluctuation ranges, laying the foundation for subsequent path diagnosis; by extracting the current amplitude and direction changes of adjacent nodes, detailed parameters of path conduction are formed, which can accurately locate abnormal nodes and mark the conduction path range. Combined with the isolation forest algorithm, the comprehensive calculation of the fluctuation range, amplitude change direction and input-output deviation of abnormal nodes in the path can quickly analyze the specific path and terminal range of abnormal propagation, so that the circuit breaker has the ability of rapid response, multi-level diagnosis and optimization adjustment in complex scenarios, thereby improving the safety, reliability and operation efficiency of the DC system.

[0126] This embodiment provides a method for manufacturing a photovoltaic circuit breaker system, which is characterized by comprising the following steps:

[0127] Collect the current time series data in the DC circuit in real time and generate current waveform data containing amplitude and timing information;

[0128] Based on the current waveform data, abnormal features in the current fluctuation are analyzed and marked to form abnormal feature data;

[0129] Determine the conduction path and position of the abnormal current in the circuit based on the abnormal characteristic data;

[0130] Combine path conduction information with system topology to predict and locate fault propagation range and endpoint;

[0131] Based on the fault location information, selectively implement circuit isolation to ensure normal operation of non-fault areas;

[0132] Adjust circuit load distribution after isolation and reconfigure paths to restore normal power delivery;

[0133] Optimize the circuit breaker operation strategy to minimize the impact on the circuit.

[0134] As an optional embodiment, an isolation forest algorithm is used to analyze the fluctuation range, amplitude change direction, and input-output deviation of abnormal nodes, and quickly calculate the specific range and end point of the abnormal signal transmission to the downstream path;

[0135] The difference threshold is dynamically set to screen out points whose amplitude difference exceeds the threshold as preliminary abnormal points, and by checking the continuity of the amplitude change direction of the preliminary abnormal points, points with the same change direction are marked as abnormal amplitude increments, thereby achieving accurate identification of sudden current anomalies.

[0136] In summary, this method significantly reduces the risk of channel congestion and signal overlap by regularly evaluating channel quality and automatically switching to the best channel with low interference, ensuring the stability and efficiency of data transmission. The application layer data is quickly forwarded by carrier nodes and arranged in chronological order, which improves the data processing speed and transmission efficiency. The status of all nodes is continuously monitored, and the network topology is adjusted by sending beacons when isolated nodes are detected, which realizes self-healing of path interruptions and ensures the continuity and reliability of network connections. When the node loses power or communication is interrupted, it quickly switches to the backup channel to ensure uninterrupted data transmission, and generates detailed fault reports to assist maintenance, shortening the fault repair time.

[0137] Finally, it should be pointed out that the methods and devices described in detail above are only embodiments, and those skilled in the art can modify these embodiments in different ways without departing from the scope of the present invention.

Claims

1. A photovoltaic circuit breaker system, characterized in that: include, A signal monitoring module (1) is used to collect the current time series data in the DC circuit in real time, and process and generate current waveform data containing amplitude and time series information; An abnormality identification module (2) analyzes and marks abnormal features in the current fluctuation based on the current waveform data to form abnormal feature data; A path diagnosis module (3) determines the conduction path and position of the abnormal current in the circuit according to the abnormal characteristic data; A fault derivation module (4), used for calculating the propagation end point of the abnormal signal by using an isolation forest algorithm based on the abnormal path identification result; An isolation control module (5) selectively implements circuit isolation based on the fault location information to ensure normal operation of non-fault areas; a path recovery module (6) that adjusts the circuit load distribution after isolation and reconfigures the path to restore normal power transmission; The action optimization module (7) optimizes the action strategy of the circuit breaker to minimize the impact on the circuit.

2. The photovoltaic circuit breaker system according to claim 1, characterized in that: The current waveform data set includes a recorded current amplitude sequence, current cycle segments, and extreme point information of each fluctuation segment.

3. The photovoltaic circuit breaker system according to claim 1, characterized in that: The anomaly identification module (2) comprises an anomaly screening submodule, a jump interval marking submodule, and an anomaly amplitude marking submodule.

4. The photovoltaic circuit breaker system according to claim 3, characterized in that: The path diagnosis module (3) comprises a node abnormality marking submodule, a path segmentation screening submodule, and an abnormal path generation submodule.

5. The photovoltaic circuit breaker system according to claim 4, characterized in that: The fault deduction module (4) comprises an abnormal node extraction submodule, an abnormal propagation range submodule, and an end point range determination submodule.

6. The photovoltaic circuit breaker system according to claim 5, characterized in that: The isolation forest algorithm is calculated according to a specific formula; The anomaly score of the node is evaluated by combining factors such as fluctuation amplitude difference and time offset trend.

7. The photovoltaic circuit breaker system according to claim 6, characterized in that: The isolation control module (5) comprises an isolation point analysis submodule, a current distribution calculation submodule, and an isolation node configuration submodule.

8. The photovoltaic circuit breaker system according to claim 7, characterized in that: The path recovery module (6) comprises a load distribution adjustment submodule, a path connectivity update submodule, and a new path marking submodule; The action optimization module (7) comprises a load change analysis submodule, an action priority calculation submodule, and a timing path generation submodule.

9. A photovoltaic circuit breaker method, characterized in that: The following steps are involved: Collect the current time series data in the DC circuit in real time and generate current waveform data containing amplitude and timing information; Based on the current waveform data, analyzing and marking abnormal features in the current fluctuation to form abnormal feature data; Determining the conduction path and position of the abnormal current in the circuit according to the abnormal characteristic data; Combine path conduction information with system topology to predict and locate fault propagation range and endpoint; Based on the fault location information, selectively implement circuit isolation to ensure normal operation of non-fault areas; Adjust circuit load distribution after isolation and reconfigure paths to restore normal power delivery; Optimize the circuit breaker operation strategy to minimize the impact on the circuit.

10. The photovoltaic circuit breaker method according to claim 9, characterized in that: Use the isolation forest algorithm to analyze the fluctuation range, amplitude change direction, and input-output deviation of abnormal nodes, and quickly calculate the specific range and end point of the abnormal signal transmission to the downstream path; The difference threshold is dynamically set to screen out points whose amplitude difference exceeds the threshold as preliminary abnormal points, and by checking the continuity of the amplitude change direction of the preliminary abnormal points, points with the same change direction are marked as abnormal amplitude increments, thereby achieving accurate identification of sudden current anomalies.

Images