Distributed collaborative modularized rapid detection method and device for box-type substation

Abstract

The invention discloses a distributed collaborative modularized rapid detection method and device for a box-type substation, and the method comprises the following steps: enabling each distributed module to judge whether the distributed module has a fault possibility or not based on collected operation data; the distributed module with the fault possibility initiates an abnormal signaling; after the other distributed modules receive the abnormal signaling, whether correction characteristic parameters of the other distributed modules are filled in corresponding positions in the abnormal signaling or not is confirmed, and the abnormal signaling is sequentially transmitted among all the other distributed modules; and summarizing all the obtained correction characteristic parameters, calculating a posterior fault probability by the distributed module which initiates the abnormal signaling, and judging that a fault exists when the posterior fault probability exceeds a set threshold value. According to the distributed architecture, the bottleneck of a centralized system is avoided, the modules are processed in parallel, and the detection efficiency is improved. Through the feature correlation model, interference of redundant information is avoided, and the accuracy of fault judgment is improved.

Description

Technical Field

[0001] The present invention belongs to the field of detection, and in particular relates to a distributed collaborative modular rapid detection method and device for a box-type substation. Background Art

[0002] With the expansion of the scale and increase of complexity of the power system, box-type substations are widely used in distribution networks and new energy access systems due to their advantages such as convenient installation, flexible expansion, and convenient maintenance. However, due to its distributed and modular characteristics, there is a complex correlation between the status and performance of each module. The traditional centralized fault diagnosis method is difficult to timely and accurately discover and locate faults, and cannot fully utilize the correlation information between modules. For example, patent CN117741305A discloses a fault detection system and detection method for a substation, which relates to the field of power testing technology. The system includes: a distributed data acquisition module, which is arranged in a key area of ​​the substation, including a plurality of electromagnetic induction sensors, for collecting real-time working status information of the substation; a data processing module, which is connected to the distributed data acquisition module through a preset anti-interference communication line, for receiving the real-time working status information, and generating fault detection results according to the real-time working status information. Therefore, there is an urgent need for a fault diagnosis method based on distributed architecture and feature association to achieve fast, efficient and accurate detection and fault diagnosis of box-type substations. Summary of the invention

[0003] In view of the defects existing in the above-mentioned prior art, the present invention provides a distributed collaborative modular rapid detection method for a box-type substation, the method comprising the following steps: Each distributed module determines whether it has the possibility of failure based on the collected operating data; A distributed module with a possibility of failure initiates abnormal signaling; After receiving the abnormal signaling, other distributed modules confirm whether to fill in their own correction characteristic parameters in the corresponding position in the abnormal signaling, and the abnormal signaling is transmitted in sequence among all other distributed modules; All the correction characteristic parameters obtained are summarized, and the distributed module that initiates the abnormal signaling calculates the a posteriori fault probability, and when the a posteriori fault probability exceeds a set threshold, it is determined that the distributed module has a fault.

[0004] Wherein, a distributed detection network is constructed for each of the distributed modules, including deploying intelligent detection nodes and building self-organizing network connections; Each distributed module acquires the collected data after time synchronization, and determines whether there is a possibility of failure of itself based on the collected data.

[0005] Among them, each distributed module independently collects its own operating data, and after preprocessing, calculates and corrects the key characteristic parameters to obtain the corrected characteristic parameters; Each distributed module determines whether it has a possibility of failure based on the correction characteristic parameters; A distributed module with the possibility of failure sets itself as the head node, initiates abnormal signaling, and requests other distributed modules to provide correction characteristic parameters.

[0006] Among them, the distributed modules include transformer module, insulation performance module, dielectric performance module, high-voltage switch module, and lightning arrester module; The key characteristic parameters of the transformer module include DC resistance and transformation ratio; the key characteristic parameters of the insulation performance module include insulation resistance; the key characteristic parameters of the dielectric performance module include dielectric loss factor; the key characteristic parameters of the high-voltage switch module include operation time, speed, acceleration; the key characteristic parameters of the lightning arrester module include reference voltage; The corrected characteristic parameters obtained by calculating and correcting the key characteristic parameters satisfy the following relationship: In the formula, : Corrected DC resistance, : Measured DC resistance, : Temperature coefficient of resistance, : reference temperature, : Measured temperature, : Transformation ratio after correction, : Measured transformation ratio, : Transformation ratio temperature coefficient, : Corrected insulation resistance, : Measured insulation resistance, : Insulation resistance temperature influence coefficient, : Insulation resistance humidity influence coefficient, : Insulation resistance atmospheric pressure influence coefficient, : Measured and reference relative humidity, : measured and reference atmospheric pressure, : Corrected dielectric loss factor, : Temperature influence coefficient of dielectric loss factor, : Measured dielectric loss factor, : Actual operation time, : Signal transmission and device response delay, : The time when the operation command is issued, : Contact action time, : Corrected reference voltage, : Measured voltage, : They are the influence coefficients of voltage temperature, voltage atmospheric pressure and voltage humidity respectively.

[0007] After receiving the abnormal signaling, the other distributed modules confirm their own association with the head node; Filling the correction characteristic parameters of the distributed modules whose correlation degree exceeds the correlation threshold into the corresponding positions in the abnormal signaling; The abnormal signaling is transmitted in sequence among other distributed modules whose association degree with the head node exceeds the association threshold, and finally returned to the head node by the tail node.

[0008] Among them, the other distributed modules are based on the internal feature association matrix Confirm the association between itself and the head node.

[0009] Among them, other distributed modules search for the feature association matrix based on the feature association model stored in their own internal storage The correlation between itself and the head node ; if is greater than the predetermined correlation threshold, indicating that the module With the head node module There is a strong feature correlation, and it participates in the data filling of abnormal signaling; Otherwise, it is considered to have no direct correlation with the fault and does not participate in the data filling of abnormal signaling.

[0010] Among them, the head node summarizes all the correction feature parameter sets Then, the Bayesian formula is used to combine the prior failure probability and the observed data to calculate the posterior failure probability ; The head node whose posterior failure probability exceeds the set threshold is judged to be faulty and an alarm is triggered.

[0011] Wherein, the head node is based on the a posteriori failure probability Determine whether there is a fault, including the following steps: Determine prior probability ; Calculate weights for calibration feature parameters ; Calculate the conditional probability density function of the correction feature parameter; Calculate the weighted log-likelihood function; Calculate likelihood ratios and apply Bayesian formula to calculate posterior failure probability ; Setting the posterior probability threshold , like , then determine the distributed module If a fault occurs; Then determine the distributed module Normal operation.

[0012] Among them, the weight The calculation formula is: , : Correction characteristic parameters The weight of : Distributed modules and calibration characteristic parameters Distributed modules The feature correlation of : Correction characteristic parameters The importance coefficient of The conditional probability density function of the correction characteristic parameter is calculated, including determining: In case of fault conditions: ; Under normal conditions: ; in: : Correction characteristic parameters The conditional probability density function PDF under fault and normal conditions respectively; : actual measured characteristic parameter value; : Correction characteristic parameters mean and standard deviation under fault conditions; : Correction characteristic parameters mean and standard deviation under normal conditions; : Natural exponential function, that is ; Among them, calculating the weighted log-likelihood function includes determining: Log-likelihood function under fault conditions: ; Log-likelihood function under normal conditions: ; in: : Observed under fault conditions The log-likelihood value of ; : Observed under normal conditions The log-likelihood value of ; : Correction characteristic parameters The weight of and : PDF value of characteristic parameters; Among them, the likelihood ratio is calculated and the posterior failure probability is calculated by applying the Bayesian formula , specifically including: Calculate likelihood ratio : , in: : Likelihood ratio, which represents the ratio of the likelihood of observing the data under fault conditions to the likelihood of observing the data under normal conditions; Calculate the a posteriori failure probability : , in: : Given observation data After that, the distributed module the a posteriori probability of failure; : likelihood ratio; : Distributed modules The prior probability of failure; : Distributed modules A priori probability of normal operation; Among them, the likelihood function Indicated in the parameter Under, the observed data possibility.

[0013] The present invention discloses a distributed collaborative modular rapid detection device for a box-type substation, and adopts the distributed collaborative modular rapid detection method for the box-type substation. The device: A possibility judgment module is used to judge whether there is a possibility of a fault based on the collected operation data; A signaling initiation module, which is used to initiate abnormal signaling when there is a possibility of failure; A signaling determination module, which is used to confirm whether to fill in its own correction characteristic parameters at the corresponding position in the abnormal signaling; The fault determination module is used to summarize all the corrected characteristic parameters, calculate the a posteriori fault probability, and determine that a fault exists when the a posteriori fault probability exceeds a set threshold.

[0014] The present invention also discloses a box-type substation rapid detection system, the system comprises a communication module, a memory and a processor, the memory stores a computer program, and the processor is used to execute the computer program to implement the above method.

[0015] Each module of the present invention independently performs data collection and preprocessing, calculates key characteristic parameters in real time, reduces the centralized data processing time, and speeds up the detection speed. Each module immediately determines whether there is a possibility of a fault based on its own corrected characteristic parameters, without waiting for global information, and responds quickly. When the possibility of a fault is detected, the module actively initiates abnormal signaling and communicates only with associated modules, reducing communication delays and data transmission volume, and quickly acquiring related data. Using the Bayesian formula, combined with the corrected characteristic parameters of itself and the associated modules, the fault probability is quickly calculated to achieve efficient fault judgment.

[0016] The distributed architecture of the present invention avoids the bottleneck of the centralized system, and each module is processed in parallel, which improves the detection efficiency. Through the feature association model, only the data of the relevant modules is focused on, which avoids the interference of redundant information and improves the accuracy of fault judgment. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present disclosure will become readily understood. In the accompanying drawings, several embodiments of the present disclosure are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein: Figure 1 The flowchart shows a distributed collaborative modular rapid detection method of a box-type substation according to an embodiment of the present invention.

[0018] Figure 2 It is a schematic diagram showing a distributed collaborative modular rapid detection device of a box-type substation according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical scheme and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] The terms used in the embodiments of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. The singular forms "a", "said" and "the" used in the embodiments of the present invention and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings, and "multiple" generally includes at least two.

[0021] It should be understood that although the terms first, second, third, etc. may be used to describe ... in the embodiments of the present invention, these ... should not be limited to these terms. These terms are only used to distinguish .... For example, without departing from the scope of the embodiments of the present invention, the first ... may also be referred to as the second ..., and similarly, the second ... may also be referred to as the first ....

[0022] It should be understood that the term "and / or" used in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0023] As used herein, the words "if" and "if" may be interpreted as "at the time of" or "when" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrases "if it is determined" or "if (stated condition or event) is detected" may be interpreted as "when it is determined" or "in response to determining" or "when detecting (stated condition or event)" or "in response to detecting (stated condition or event)", depending on the context.

[0024] It should also be noted that the term "includes", "comprising" or any other variation thereof is intended to cover non-exclusive inclusion, so that a commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprising a ..." do not exclude the existence of other identical elements in the commodity or device including the elements.

[0025] The purpose of the present invention is to provide a distributed collaborative modular rapid detection method for a box-type substation. By constructing a distributed detection network, utilizing the correlation between characteristic parameters of each distributed module, and designing a complex distributed fault diagnosis algorithm, rapid and accurate diagnosis and positioning of faults in each module can be achieved, thereby solving the problems of low detection efficiency and insufficient accuracy in the prior art.

[0026] like Figure 1 As shown, the present invention provides a distributed collaborative modular rapid detection method for a box-type substation, comprising the following steps: Building a distributed detection network, including deploying intelligent detection nodes on distributed modules and building self-organizing network connections; Obtain the original data of each distributed module after time synchronization; Each distributed module determines whether it has the possibility of failure based on the collected operating data; If the distributed module finds that there is a possibility of failure, it initiates abnormal signaling; After receiving the abnormal signaling, other distributed modules confirm whether to fill in their own correction characteristic parameters in the corresponding position in the abnormal signaling, and the abnormal signaling is transmitted in sequence among all other distributed modules; The head node summarizes all the corrected characteristic parameters, and the distributed module that initiates the abnormal signaling calculates the a posteriori fault probability. If the a posteriori fault probability exceeds the set threshold, the distributed module is determined to be faulty. The distributed module refers to each module in the box-type substation, that is, the box-type substation includes multiple distributed modules.

[0027] The present invention makes full use of the high efficiency of the distributed architecture and the accuracy of the Bayesian method to achieve rapid fault detection of box-type substations, reflecting the rapid response and intelligent diagnosis capabilities of the system.

[0028] Among them, the distributed modules are transformer comprehensive test module, insulation performance test module, dielectric performance test module, high-voltage switch comprehensive test module, and lightning arrester AC / DC test module.

[0029] Among them, intelligent detection nodes are deployed on each module of the box-type substation (including transformer module, insulation performance module, dielectric performance module, high-voltage switch module, lightning arrester module, etc.), and each intelligent detection node has data collection, preprocessing, feature extraction and communication functions.

[0030] Each intelligent detection node is connected to each other through wireless communication (such as ZigBee, LoRa, WiFiMesh) or wired communication to form an ad hoc network connection, that is, a self-organizing distributed detection network, to achieve real-time transmission and sharing of data.

[0031] Among them, each intelligent detection node collects on-site raw data, including voltage, current, temperature, humidity, signal timestamp, etc.

[0032] Each intelligent detection node uses local sensors to collect the operating data of the corresponding distributed module in real time, and uses the Network Time Protocol (NTP) or Precision Time Protocol (PTP) to achieve time synchronization between nodes, and output the original data (original operating data) of each distributed module after time synchronization.

[0033] Among them, data preprocessing includes: filtering and denoising the original data to remove interference signals; detecting and eliminating abnormal data points through statistical methods to ensure data quality; normalizing the data to unify the data format and dimension.

[0034] Each distributed module independently collects its own operating data, performs preprocessing, and calculates the corrected key characteristic parameters (i.e., corrected characteristic parameters). Based on these key characteristic parameters, each distributed module determines whether it has the possibility of failure. If a distributed module has the possibility of failure, it sets itself as the head node, initiates abnormal signaling, and requests other distributed modules to provide corrected characteristic parameters. After receiving the abnormal signaling, other distributed modules will calculate the corrected characteristic parameters based on the internal characteristic association matrix. , confirming its own correlation with the head node. If the correlation is high, fill in its own correction feature parameters in the corresponding position in the signaling. The abnormal signaling is transmitted in turn between other distributed modules whose correlation with the head node exceeds the correlation threshold, and finally returned to the head node by the tail node. The head node summarizes all the correction feature parameter sets Then, the Bayesian formula is used to combine the prior failure probability and the observed data to calculate the posterior failure probability If the a posteriori fault probability exceeds the set threshold, the fault is determined to be established and an alarm is triggered.

[0035] Each distributed module of the present invention independently performs data collection and preprocessing, calculates key characteristic parameters in real time, reduces the centralized data processing time, and speeds up the detection speed. Each distributed module immediately determines whether there is a possibility of a fault based on its own corrected characteristic parameters, without waiting for global information, and responds quickly. When a possible fault is detected, the distributed module actively initiates abnormal signaling and communicates only with the associated distributed modules, reducing communication delays and data transmission volume, and quickly acquiring related data. Using the Bayesian formula, combined with the corrected characteristic parameters of itself and the associated modules, the fault probability is quickly calculated to achieve efficient fault judgment.

[0036] The distributed architecture of the present invention avoids the bottleneck of the centralized system, and each distributed module processes in parallel, thereby improving the detection efficiency. Through the feature association model, only the data of the relevant modules is focused on, thus avoiding the interference of redundant information and improving the accuracy of fault judgment.

[0037] The present invention utilizes the high efficiency of the distributed architecture and the accuracy of the Bayesian method to achieve rapid fault detection of box-type substations, reflecting the rapid response and intelligent diagnosis capabilities of the system.

[0038] Among them, each intelligent detection node calculates the key characteristic parameters of this distributed module according to the preprocessed data, and the corresponding key characteristic parameters of this module are: Transformer module: DC resistance , Transformation ratio .

[0039] Insulation performance module: insulation resistance .

[0040] Dielectric Properties Module: Dielectric Loss Factor .

[0041] High voltage switch module: operating time ,speed , acceleration .

[0042] Arrester module: reference voltage

[0043] Afterwards, the influence of environmental factors such as temperature, humidity, and atmospheric pressure on the characteristic parameters is considered and correction is performed to obtain the corrected characteristic parameters.

[0044] Optionally, the key characteristic parameter determination and calibration method of each distributed module includes: For transformer module DC resistance Calculation and correction of The measurement method of the measured DC resistance: Use a DC resistance tester to measure the DC resistance of the transformer winding to obtain the measured DC resistance

[0045] Effect of temperature on resistance: Resistance changes with temperature and can be corrected using the temperature coefficient. The correction method is: , in: : Corrected DC resistance

[0046] : Measured DC resistance

[0047] : Temperature coefficient of resistance ( ) : Reference temperature (take ) : Measured temperature.

[0048] For the transformer module ratio Calculation and correction of The actual measurement method of the transformation ratio is: Use a transformation ratio tester to measure the primary voltage and secondary voltage , calculate the measured transformation ratio: .

[0049] in: : Measured transformation ratio :Primary voltage

[0050] :Secondary voltage (V) Temperature may affect the size and material properties of the windings, and thus affect the transformation ratio. Correction method: , in: : Transformation ratio after correction : Measured transformation ratio : Transformation ratio temperature coefficient (determined according to transformer material and structure) : Reference temperature (take ) : Measured temperature.

[0051] Insulation resistance for insulation performance modules Calculation and correction of The actual insulation resistance is measured by applying the test voltage using an insulation resistance tester (megohmmeter) to measure the actual insulation resistance. .

[0052] The influencing factor of environmental factor correction is temperature ,humidity , atmospheric pressure The calibration method is: , in: : Corrected insulation resistance

[0053] : Measured insulation resistance ( ) : Temperature influence coefficient

[0054] : Humidity influence coefficient

[0055] : Atmospheric pressure influence coefficient

[0056] : Measured and reference relative humidity (%) : Measured and reference atmospheric pressure

[0057] : Reference temperature (take ) : Measured temperature.

[0058] For the dielectric properties module, the dielectric loss factor Calculation and correction of The dielectric loss factor represents the loss characteristics of dielectric materials and is defined as the ratio of dielectric loss current to capacitive current. The measurement method is to use a dielectric loss tester to measure the active power of the test sample under AC voltage. and voltage , Current . Calculation formula: or .

[0059] in: : Measured dielectric loss factor :Power loss(W) :Loss current (A) : angular frequency ( ,in For frequency ) :Capacitance(F) :Test voltage

[0060] Temperature will affect the loss characteristics of the medium, correction formula: .

[0061] in: : Corrected dielectric loss factor : Temperature influence coefficient of dielectric loss factor

[0062] : Measured dielectric loss factor : Reference temperature (take ) : Measured temperature.

[0063] Operating time of high voltage switch module The calculation method is to record the time when the operation command is issued. and contact action time .

[0064] Calculation formula:

[0065] in: :Actual operation time(s) : Signal transmission and device response delay (s), which can be obtained through pre-calibration.

[0066] speed and acceleration The motion travel data is obtained by using a displacement sensor to obtain the displacement change of the switch contact over time. .

[0067] Speed ​​calculation formula: .

[0068] Discrete data calculation: .

[0069] Acceleration calculation formula:

[0070] Discrete data calculation:

[0071] The mechanical motion characteristics of high-voltage switches are generally less affected by ambient temperature, so no environmental correction is performed.

[0072] Reference voltage of the arrester module Calculation and correction of the standard voltage is applied to the arrester in the measurement, the current passing through the arrester is measured, and the voltage at a specific current is found .

[0073] Factors affecting environmental correction: Temperature , atmospheric pressure ,humidity .

[0074] The correction formula is

[0075] in: : Corrected reference voltage

[0076] : Measured voltage

[0077] : They are the voltage temperature, voltage atmospheric pressure, and voltage humidity influence coefficients respectively. Other parameters are the same as above.

[0078] Each module collects and corrects the characteristic parameters to form its own characteristic parameter set : Transformer module: ; Insulation performance module: ; Dielectric Properties Module: ; High voltage switch module: ; Lightning arrester module: ; All environmental parameters , , All of these are obtained through real-time measurement by environmental sensors.

[0079] Reference conditions , Generally take standard conditions: temperature ; Relative humidity ; Atmospheric pressure . Influence coefficient , , , , , , , Etc., need to be determined based on equipment characteristics, material properties and experimental data.

[0080] Through the above calculation formula and wherein, each detection node can calculate and correct the key characteristic parameters of this module according to the preprocessed data, and obtain the corrected characteristic parameter set These parameters not only reflect the operating status of the equipment, but also eliminate the influence of environmental factors, providing an accurate and reliable data basis for subsequent feature correlation analysis and fault diagnosis.

[0081] Among them, each distributed module (including transformer module, insulation performance module, dielectric performance module, high-voltage switch module, lightning arrester module, etc.) uses local sensors to collect its own operating data in real time, such as voltage, current, temperature, humidity, etc., and achieves time synchronization through the Network Time Protocol (NTP) or Precision Time Protocol (PTP).

[0082] Preprocess the collected raw data, including filtering, noise reduction, removal of abnormal data points, normalization, etc., to ensure data quality and consistency. Calculate and correct characteristic parameters, including correcting the preprocessed data in combination with environmental factors (temperature, humidity, atmospheric pressure, etc.), calculating the key characteristic parameters of this module, and obtaining the corrected characteristic parameter set .

[0083] Then judge the possibility of fault, including each module according to its own correction characteristic parameters , to determine whether there is a possibility of failure. The specific method is to compare the correction characteristic parameters with the preset normal operating threshold or standard range. If all characteristic parameters are within the normal range, it is considered that there is no possibility of failure and normal operation continues. If some characteristic parameters exceed the normal range and exceed the preset fault trigger threshold, it is considered that there is a possibility of failure.

[0084] And, if there is a possibility of failure, it is necessary to initiate an abnormal signaling request, including: Determine the faulty module as the head node: When a distributed module (referred to as a distributed module ) When it determines that there is a possibility of failure, it sets itself as the "head node" and constructs abnormal signaling.

[0085] The created exception signaling message contains: Head node identifier: indicates that the signaling is sent by the distributed module Initiated.

[0086] Anomaly type and description: Indicates the detected anomaly characteristic parameters and out-of-range conditions.

[0087] Data filling structure: A predefined data structure (such as an array or list) used by other distributed modules to fill in correction feature parameters. This data structure contains The associated other distributed module locations.

[0088] Send abnormal signaling request to other distributed modules. Distributed module Through the distributed detection network, the abnormal signaling is sent to other distributed modules along a certain path.

[0089] After the abnormal signaling message is constructed, the abnormal signaling is sent to other distributed modules. Other distributed modules (referred to as modules ) After receiving the abnormal signaling, perform the following processing: Read head node information: identify the head node identifier in the signaling .

[0090] Feature correlation judgment: Find the feature association matrix based on the internally stored feature association model The corresponding correlation .

[0091] if Greater than a predetermined correlation threshold, indicating that the distributed module Distributed modules with head nodes There is a strong feature correlation.

[0092] If the correlation is insufficient, it is considered to have no direct correlation with the fault and will not be included in the data filling.

[0093] Fill in the calibration characteristic parameters: For distributed modules associated with the head node , and its own correction characteristic parameters Fill in the corresponding position in the exception signaling.

[0094] After processing the abnormal signaling, the distributed module Pass the signal to the next distributed module and continue to pass the exception signal: The abnormal signaling can be transmitted in the following ways: Sequential delivery: Delivered in sequence according to the pre-set distributed module serial number or topological order.

[0095] Route-based delivery: The next-hop distributed module is selected based on the network topology and routing protocol.

[0096] The termination condition of the abnormal signaling transmission is that all distributed modules associated with the head node receive the abnormal signaling and fill in the data.

[0097] When the signaling is transmitted, the last distributed module (tail node) that processes the abnormal signaling returns the abnormal signaling to the head node module. , that is, the tail node returns abnormal signaling.

[0098] The return method can be: Direct return: sent directly to the head node through the network.

[0099] Or, reverse transmission: return in the reverse order of signal transmission.

[0100] Head node corresponds to distributed module After receiving the returned abnormal signaling, extract the correction characteristic parameters filled in by all the associated distributed modules to form a correction characteristic parameter set , that is, the head node receives and summarizes the data.

[0101] The head node comprehensively analyzes the correction feature parameter set based on the correction feature parameters of itself and the associated distributed modules. , including using feature association models to further analyze the fault and calculate the updated fault probability .

[0102] If the updated failure probability If the fault exceeds the set fault confirmation threshold, the distributed module is judged to be A failure occurred.

[0103] If the probability of failure decreases, it may be a false alarm and you can continue monitoring or cancel the fault alarm.

[0104] If the fault is confirmed, the head node sends a fault alarm message to the system or operation and maintenance tester, including the fault type, impact scope, and recommended handling measures.

[0105] For example, assuming that the transformer module (distributed module ) detects its own DC resistance If the value is abnormally high, it is determined that there is a possibility of failure. Perform the following steps in sequence: 1. Distributed modules Initiate abnormal signaling: Set itself as the head node, create an exception signal, and describe Abnormal situation.

[0106] 2.Signaling sending: Broadcast abnormal signaling to other modules.

[0107] 3. Insulation performance module (distributed module ) Receive signaling: Read the head node identification as a distributed module .

[0108] Find the Incidence Matrix , and found that it has a high correlation with the transformer module.

[0109] Fill in its own insulation resistance Equal correction characteristic parameters.

[0110] 4. Dielectric performance module, high voltage switch module, etc. handle abnormal signaling: According to the degree of association, fill in the respective correction characteristic parameters.

[0111] 5. Signaling returns to the head node: The last module returns the exception signal to the distributed module .

[0112] 6. Distributed modules Conduct fault analysis: The summed correction characteristic parameter set .

[0113] Combined with the characteristic correlation model, the insulation resistance is found Further confirm the source of the fault.

[0114] Calculate the failure probability of the update , exceeds the threshold.

[0115] 7. Fault determination and handling: Confirm that the transformer module is faulty.

[0116] Send an alarm message and recommend checking the insulation condition of the transformer.

[0117] Among them, the feature association model is used to describe the correlation between feature parameters of different distributed modules. Elements The larger the value, the better the module With module The higher the correlation between the characteristics, the higher the correlation between the characteristics. The correlation can be determined based on historical data statistics, physical mechanism analysis or expert experience.

[0118] For example, the following is a specific correlation matrix In the form of: Among them, each distributed module and its key characteristic parameters are as follows: Transformer module (distributed module 1): DC resistance , ratio .

[0119] Insulation performance module (distributed module 2): insulation resistance .

[0120] Dielectric Properties Module (Distributed Module 3): Dielectric Loss Factor .

[0121] High voltage switch module (distributed module 4): operating time ,speed , acceleration .

[0122] Arrester module (distributed module 5): reference voltage .

[0123] Feature Correlation Matrix is a A matrix with elements Represents a distributed module With distributed modules The degree of feature correlation between . The correlation can be expressed as a numerical value, usually in the range of , the larger the value, the higher the correlation.

[0124] Here is an example matrix:

[0125] Among them, the self-correlation (diagonal elements) : Indicates that the correlation between each module and itself is 1, which is completely correlated.

[0126] The degree of association between distributed modules: Distributed module 1 (transformer module) and distributed module 2 (insulation performance module): The insulation performance of the transformer is crucial to its operation. The DC resistance of the transformer and transformation ratio May be affected by the insulation condition.

[0127] Insulation resistance It reflects the insulation condition between the transformer winding and the core, and directly affects the performance of the transformer.

[0128] Therefore, the correlation between the two is high.

[0129] Distributed module 2 (insulation performance module) and distributed module (Dielectric Properties Module): The insulation performance is closely related to the dielectric performance, and the dielectric loss factor Reflects the dielectric loss of the insulating material. Insulation resistance and Both are important parameters for measuring insulation status. Therefore, the two are highly correlated.

[0130] Distributed module 1 (transformer module) and distributed module 3 (dielectric performance module): .

[0131] The dielectric properties of the transformer affect its insulation state and thus the DC resistance and transformation ratio .

[0132] Although not as large as the direct insulation performance effect, there is still a strong correlation.

[0133] Distributed module 1 and distributed module 4 (high voltage switch module): The operating state of the transformer is somewhat related to the action of the high-voltage switch. For example, in the case of overload or fault, the high-voltage switch may act. However, the correlation is relatively low.

[0134] Distributed module 4 and distributed module 5 (lightning arrester module): .

[0135] High voltage switches and lightning arresters are both related to overvoltage protection.

[0136] In case of lightning strike or overvoltage, the arrester will operate and the high-voltage switch may need to cut off the circuit.

[0137] There is a certain correlation between the two.

[0138] Distributed Module 1 and Attack Module 5: . The correlation between transformers and lightning arresters is low.

[0139] Other low correlations: Distributed module 2 and distributed module 4: , Distributed module 3 and distributed module 4: , Distributed module 2 and distributed module 5: , Distributed module 3 and distributed module 5: , These distributed modules are less well connected because the direct relationship between their functions and the correction characteristic parameters is not strong.

[0140] Optionally, based on physical relationships and professional knowledge, the physical connections and working principles between the distributed modules are analyzed to determine the correlation between the correction characteristic parameters. For example, insulation performance and dielectric performance both involve the state of the insulation material and have a high degree of correlation.

[0141] Optionally, the correlation coefficient of the correction characteristic parameters between the distributed modules is calculated using historical operation data. If the correction characteristic parameters of two distributed modules often change synchronously in history, it means that the correlation is high.

[0142] In one embodiment, the head node is based on the probability of failure Determine whether there is a fault, including the following steps: Step 1: Determine the prior probability .

[0143] Determine the distributed modules based on historical data or equipment reliability analysis Prior probability of failure , and calculate the prior probability of normal operation .

[0144] For example, suppose that according to historical statistics, the distributed module The prior probability of failure is , then the prior probability of normal operation is: .

[0145] Step 2: Calculate the weights of the correction feature parameters .

[0146] Weight The calculation formula is: , : Correction characteristic parameters The weight of reflects the importance of this feature to fault judgment.

[0147] : Distributed modules and calibration characteristic parameters Distributed modules The feature correlation degree ranges from .

[0148] : Correction characteristic parameters The importance coefficient of , determined according to the degree of influence of the feature on fault judgment.

[0149] : The weighted sum of all correction feature parameters, used for normalization so that the sum of the weights is 1.

[0150] Step 3: Calculate the conditional probability density function of the correction feature parameters.

[0151] The calibration characteristic parameters are assumed to follow a normal distribution under both fault and normal conditions.

[0152] In case of fault conditions:

[0153] Under normal conditions: in: : Correction characteristic parameters Probability density function (PDF) under fault and normal conditions.

[0154] : The actual measured calibration characteristic parameter value.

[0155] : Correction characteristic parameters Mean and standard deviation under fault conditions.

[0156] : Correction characteristic parameters Mean and standard deviation under normal conditions.

[0157] : Natural exponential function, that is .

[0158] Step 4: Calculate the weighted log-likelihood function, including calculating: Log-likelihood function under fault conditions: .

[0159] Log-likelihood function under normal conditions: .

[0160] in: : Observed under fault conditions The log-likelihood value of .

[0161] : Observed under normal conditions The log-likelihood value of .

[0162] : Correction characteristic parameters The weight of .

[0163] and : PDF value of the correction characteristic parameter (calculated in step 3).

[0164] Step 5: Calculate the likelihood ratio and apply Bayes’ formula to calculate the posterior probability , specifically including: Step 5.1: Calculate the likelihood ratio : , in: : Likelihood ratio, which represents the ratio of the likelihood of observing data under fault conditions to that under normal conditions.

[0165] The likelihood ratio compares the likelihood of observing data under two conditions and reflects which hypothesis the data supports more. The difference in log likelihood is converted to exponential form to obtain the likelihood ratio.

[0166] Step 5.2: Calculate the posterior probability .

[0167] , in: : Given observation data After that, the distributed module The a posteriori probability of failure occurring.

[0168] : Likelihood ratio (calculated).

[0169] : Distributed modules The prior probability of failure occurring.

[0170] : Distributed modules The prior probability of normal operation.

[0171] Among them, the likelihood function Indicated in the parameter Under, the observed data possibility.

[0172] In this embodiment, the likelihood functions are: : Assuming a fault occurs (parameters are ), the observed data possibility.

[0173] : Assuming that the fault does not occur (parameters are ), the observed data possibility.

[0174] Likelihood Ratio Meaning: , Likelihood Ratio The likelihood of the observed data under the two assumptions is compared.

[0175] If Λ>1, it means that the data more supports the hypothesis that the fault occurred.

[0176] If Λ<1, it means that the data is more supportive of the assumption of normal operation.

[0177] Using the Bayesian formula, the likelihood ratio and prior probability are combined to calculate the posterior probability.

[0178] The posterior probability reflects the likelihood that a distributed module fails after observing the data.

[0179] Step 6: Fault determination.

[0180] Setting the posterior probability threshold , determined according to the specific application scenario and risk preference. The general value is between 0.5 and 0.8.

[0181] Criteria: like , then determine the distributed module A failure occurred.

[0182] like Then determine the distributed module Normal operation.

[0183] For example, the weights of the correction feature parameters are , there are the following processes: The following modules and their key characteristic parameters are determined: Distributed module 1 (transformer module): Characteristic parameter 1: DC resistance (denoted as ) Correction characteristic parameter 2: Transformation ratio (denoted as ) Distributed module 2 (insulation performance module): Correction characteristic parameter 3: Insulation resistance (denoted as ) Distributed Module 3 (Dielectric Properties Module): Correction characteristic parameter 4: dielectric loss factor (denoted as ) Distributed module 4 (high voltage switch module): Correction characteristic parameter 5: Operation time (denoted as ) Correction characteristic parameter 6: Speed (denoted as ) Correction characteristic parameter 7: Acceleration (denoted as ) Distributed module 5 (lightning arrester module): Calibration characteristic parameter 8: Reference voltage (denoted as ) Assume that based on professional knowledge and experience, each correction feature parameter is assigned the following correction feature parameter importance coefficient :

[0184] Determine the distributed module to which the characteristic parameters belong , the module number to which each characteristic parameter belongs is as follows:

[0185] Then calculate the weight ,include: For target modules , assume that distributed module 1 (transformer module) is being evaluated for failure.

[0186] Calculate the correlation of each feature parameter multiplied by the importance coefficient .

[0187] According to the given feature correlation matrix and importance coefficient ,calculate: Correction characteristic parameters (Distributed module 1, itself) , .

[0188] Correction characteristic parameters (Distributed module 1, itself) , .

[0189] Correction characteristic parameters (Distributed Module 2) , .

[0190] Correction characteristic parameters (Distributed Module 3) .

[0191] Correction characteristic parameters (Distributed Module 4) , .

[0192] Correction characteristic parameters (Distributed Module 4) , .

[0193] Correction characteristic parameters (Distributed Module 4) , .

[0194] Correction characteristic parameters (Distributed Module 5) , .

[0195] Denominator: Weighted sum of all calibration characteristic parameters .

[0196] Calculate the weight of each calibration feature parameter .

[0197] Weight .

[0198] Correction characteristic parameters Weight .

[0199] Correction characteristic parameters Weight .

[0200] Correction characteristic parameters Weight .

[0201] Correction characteristic parameters Weight .

[0202] Correction characteristic parameters Weight .

[0203] Correction characteristic parameters Weight .

[0204] Correction characteristic parameters Weight .

[0205] Each module of the present invention independently performs data collection and preprocessing, calculates key characteristic parameters in real time, reduces the centralized data processing time, and speeds up the detection speed. Each module immediately determines whether there is a possibility of a fault based on its own corrected characteristic parameters, without waiting for global information, and responds quickly. When a possible fault is detected, the module actively initiates abnormal signaling and communicates only with associated modules, reducing communication delays and data transmission volume, and quickly acquiring related data. Using the Bayesian formula, combined with the corrected characteristic parameters of itself and the associated modules, the fault probability is quickly calculated to achieve efficient fault judgment.

[0206] The distributed architecture of the present invention avoids the bottleneck of the centralized system, and each module is processed in parallel, which improves the detection efficiency. Through the feature association model, only the data of the relevant modules is focused on, which avoids the interference of redundant information and improves the accuracy of fault judgment.

[0207] like Figure 2 As shown, the present invention discloses a distributed collaborative modular rapid detection device for a box-type substation, which adopts the distributed collaborative modular rapid detection method for a box-type substation. The device: A possibility judgment module is used to judge whether there is a possibility of a fault based on the collected operation data; A signaling initiation module, which is used to initiate abnormal signaling when there is a possibility of failure; A signaling determination module, which is used to confirm whether to fill in its own correction characteristic parameters at the corresponding position in the abnormal signaling; The fault determination module is used to summarize all the corrected characteristic parameters, calculate the a posteriori fault probability, and determine that a fault exists when the a posteriori fault probability exceeds a set threshold.

[0208] The present invention also discloses a box-type substation rapid detection system, the system comprises a communication module, a memory and a processor, the memory stores a computer program, and the processor is used to execute the computer program to implement the above method.

[0209] It should be noted that the computer-readable medium mentioned above in the present disclosure may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. Computer readable signal media may also be any computer readable medium other than computer readable storage media, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0210] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.

[0211] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0212] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0213] The units involved in the embodiments described in the present disclosure may be implemented by software or hardware, wherein the name of a unit does not, in some cases, limit the unit itself.

[0214] The above introduces the preferred embodiments of the present invention, which is intended to make the spirit of the present invention clearer and easier to understand, but is not intended to limit the present invention. All modifications, substitutions, and improvements made within the spirit and principles of the present invention should be included in the scope of protection outlined by the claims attached to the present invention.

Claims

1. A distributed collaborative modular rapid detection method for a box-type substation, the method comprising the following steps: Each distributed module determines whether it has the possibility of failure based on the collected operating data; A distributed module with a possibility of failure initiates abnormal signaling; After receiving the abnormal signaling, other distributed modules confirm whether to fill in their own correction characteristic parameters in the corresponding position in the abnormal signaling, and the abnormal signaling is transmitted in sequence among all other distributed modules; All corrected characteristic parameters are summarized, and the distributed module that initiates the abnormal signaling calculates the a posteriori fault probability, and determines that a fault exists when the a posteriori fault probability exceeds a set threshold.

2. A distributed collaborative modular rapid detection method for a box-type substation according to claim 1, characterized in that: Constructing a distributed detection network for each of the distributed modules, including deploying intelligent detection nodes and building self-organizing network connections; Each distributed module obtains the operating data collected after time synchronization, and determines whether there is a possibility of failure of itself based on the operating data.

3. A distributed collaborative modular rapid detection method for a box-type substation according to claim 1, characterized in that: Each distributed module independently collects its own operating data, and after preprocessing, calculates and corrects the key characteristic parameters to obtain the corrected characteristic parameters; Each distributed module determines whether it has a possibility of failure based on the correction characteristic parameters; A distributed module with the possibility of failure sets itself as the head node, initiates abnormal signaling, and requests other distributed modules to provide correction characteristic parameters.

4. A distributed collaborative modular rapid detection method for a box-type substation as claimed in claim 3, characterized in that: After receiving the abnormal signaling, the other distributed modules confirm the association between themselves and the head node; Filling the correction characteristic parameters of the distributed modules whose correlation degree exceeds the correlation threshold into the corresponding positions in the abnormal signaling; The abnormal signaling is transmitted in sequence among other distributed modules whose association degree with the head node exceeds the association threshold, and finally returned to the head node by the tail node.

5. A distributed collaborative modular rapid detection method for a box-type substation as claimed in claim 3, characterized in that: The other distributed modules are based on the internal feature association matrix Confirm the association between itself and the head node; Other distributed modules search for feature association matrices based on their own internally stored feature association models. The correlation between itself and the head node .

6. A distributed collaborative modular rapid detection method for a box-type substation as claimed in claim 3, characterized in that: The head node summarizes all the correction feature parameters. Then, the Bayesian formula is used to combine the prior failure probability and the observed data to calculate the posterior failure probability ; The head node whose posterior failure probability exceeds the set threshold is judged to be faulty and an alarm is triggered.

7. A distributed collaborative modular rapid detection method for a box-type substation as claimed in claim 6, characterized in that: The head node is based on the a posteriori failure probability Determine whether there is a fault, including the following steps: Determine prior probability ; Calculate weights for calibration feature parameters ; Calculate the conditional probability density function of the correction characteristic parameter; Calculate the weighted log-likelihood function; Calculate likelihood ratios and apply Bayesian formula to calculate posterior failure probability ; Setting the posterior probability threshold , like , then determine the distributed module If a fault occurs; Then determine the distributed module Normal operation.

8. A distributed collaborative modular rapid detection method for a box-type substation as claimed in claim 6, characterized in that: in, Weight The calculation formula is: ， : Correction characteristic parameters The weight of : Distributed modules and calibration characteristic parameters Distributed modules The characteristic correlation of : Correction characteristic parameters The importance coefficient of The conditional probability density function of the correction characteristic parameter is calculated, including determining: In case of fault conditions: ; Under normal conditions: ; in: : Correction characteristic parameters The conditional probability density function PDF under fault and normal conditions respectively; : actual measured characteristic parameter value; : Correction characteristic parameters mean and standard deviation under fault conditions; : Correction characteristic parameters mean and standard deviation under normal conditions; : Natural exponential function, that is ; Among them, calculating the weighted log-likelihood function includes determining: Log-likelihood function under fault conditions: ; Log-likelihood function under normal conditions: ; in: : Observed under fault conditions The log-likelihood value of ; : Observed under normal conditions The log-likelihood value of ; : Correction characteristic parameters The weight of Among them, the likelihood ratio is calculated and the posterior failure probability is calculated by applying the Bayesian formula , specifically including: Calculate likelihood ratio : ， in: : Likelihood ratio, which represents the ratio of the likelihood of observing the data under fault conditions to the likelihood of observing the data under normal conditions.

9. A distributed collaborative modular rapid detection method for a box-type substation as claimed in claim 8, characterized in that: Calculate the a posteriori failure probability : ， in: : Given observation data After that, the distributed module the a posteriori probability of failure; : Distributed modules The prior probability of failure; : Distributed modules The prior probability of normal operation.

10. A distributed collaborative modular rapid detection device for a box-type substation, characterized in that: According to any one of claims 1 to 9, the distributed collaborative modular rapid detection method for the box-type substation is adopted, wherein the device: A possibility judgment module is used to judge whether there is a possibility of a fault based on the collected operation data; A signaling initiation module, which is used to initiate abnormal signaling when there is a possibility of failure; A signaling determination module, which is used to confirm whether to fill in its own correction characteristic parameters at the corresponding position in the abnormal signaling; The fault determination module is used to summarize all the corrected characteristic parameters, calculate the a posteriori fault probability, and determine that a fault exists when the a posteriori fault probability exceeds a set threshold.

Images