Distribution box switch cabinet fault detection method and system

Abstract

The invention relates to the technical field of power equipment monitoring, and discloses a distribution box switch cabinet fault detection method and system, and the method comprises the steps: collecting the environment and operation state data of a distribution box switch cabinet; generating an environment feature matrix through dynamic weight calculation and weighted fusion; and calculating fault probability distribution based on the SVM fault classification model, judging a fault type, and generating an early warning information report. Compared with the prior art that the detection precision is low, the fault type judgment is indefinite, and real-time and reliable data transmission and fault monitoring cannot be realized especially under the low-bandwidth and high-interference network conditions, the method and the device have the advantages that the dynamic weight fusion, the preset network transmission processing and the SVM-based fault classification model are adopted, so that the real-time and reliable data transmission and fault monitoring can be realized; comprehensive analysis of multiple fault parameters is realized, so that the problem of misjudgment caused by dependence on a single parameter in a traditional method is avoided, and the accuracy and the real-time performance of fault detection are improved.

Description

Technical Field

[0001] The invention belongs to the technical field of power equipment monitoring, and in particular relates to a distribution box switch cabinet fault detection method and system. Background Art

[0002] At present, the distribution box switchgear is an important control and protection equipment in the power system. Its operating environment is complex and the fault forms are diverse. However, there are many shortcomings in the existing distribution box switchgear fault detection technology. For example, most existing detection methods are only based on single parameters such as temperature and current for monitoring, and cannot integrate multi-dimensional data for fault analysis; at the same time, under low-bandwidth and high-interference network conditions, traditional methods are difficult to achieve stable data transmission, resulting in data loss or transmission delay. In addition, the existing fault classification model often uses fixed weights and fails to adjust the model parameters according to the dynamic characteristics of the sensor, resulting in low detection accuracy. The existing technology cannot fully meet the needs of the distribution box switchgear in terms of multi-parameter comprehensive analysis, low-bandwidth reliable transmission, dynamic fault judgment, etc. Therefore, there is an urgent need for a distribution box switchgear fault detection method that can still achieve multi-dimensional data fusion, real-time transmission and high-precision fault classification under complex operating environments and network conditions to improve the operational reliability and safety of the power system. Summary of the invention

[0003] In view of the above-mentioned technical deficiencies, the purpose of the present invention is to propose a distribution box switch cabinet fault detection method, aiming to solve the technical problems in the prior art such as low detection accuracy and unclear fault type determination, especially under low bandwidth and high interference network conditions, and the inability to achieve real-time and reliable data transmission and fault monitoring.

[0004] In order to solve the above technical problems, the present invention adopts the following technical solutions: The present invention provides a method for detecting faults in a distribution box switch cabinet.

[0005] The distribution box switch cabinet fault detection method comprises:

[0006] Step S10: Collecting environment and equipment operation status data through various sensors installed in the switch cabinet of the distribution box, including:

[0007] Temperature sensor, used to collect the operating temperature T of the distribution box switch cabinet;

[0008] Humidity sensor, used to collect humidity H of the switch cabinet of the distribution box;

[0009] Gas concentration sensor, used to collect sulfur hexafluoride gas concentration G;

[0010] Current sensor and voltage sensor: used to collect current I and voltage V respectively;

[0011] The above data are normalized to generate a normalized multi-parameter data set D = {T norn ,

[0012] H norn , G norn , I norn , V norn};

[0013] Step S20: Generate an environmental feature matrix M by introducing dynamic weights and fusing multi-parameter data sets. The steps include:

[0014] Get the historical data of the sensor within the preset time period, and calculate the dynamic weight based on the standard deviation of the historical data of the sensor. The formula is:

[0015]

[0016] Among them, i is the i-th sensor, ω i is the standard deviation of the ith sensor, ω i is the dynamic weight of the i-th sensor;

[0017] Construct the environment feature matrix M after dynamic weighted fusion, M = {T norn ·ω T , H norn ·ω H ,

[0018] G norn ·ω G , I norn ·ω I , V norn ·ω V}, where ω T is the temperature dynamic weight, ω H is the humidity dynamic weight, ω G is the dynamic weight of gas concentration, ω I is the current dynamic weight, ω V is the voltage dynamic weight;

[0019] Step S30: Packing the environmental feature matrix M generated in step S20 and transmitting it to the remote server via a preset network in the form of a data packet;

[0020] Step S40: After receiving the data packet, the remote server uses the cyclic redundancy check code in the data packet to perform data verification:

[0021] CRC Packet =R(packet)mod G(X)

[0022] Among them, R(packet) is the binary polynomial representation of the data packet, G(X) is a predefined fixed polynomial, CRCpacket is the calculated value of the cyclic redundancy check code, and mod is the remainder operation;

[0023] If the calculated value of the cyclic redundancy check code CRC packet Equal to the cyclic redundancy check code CRC, the check passes and the data packet is marked as valid;

[0024] If the calculated value of the cyclic redundancy check code CRC packet Not equal to the cyclic redundancy check code CRC, the check fails, and the device is requested to resend;

[0025] Step S50: Input the feature matrix M in the data packet marked as valid into the SVM fault classification model and output the fault probability distribution P, where:

[0026] P={p1,p2,p3,p4,p5,p6}

[0027] Among them, p1 is the probability of over-temperature fault, p2 is the probability of abnormal humidity fault, p3 is the probability of current overload fault, p4 is the probability of abnormal voltage fault, p5 is the probability of gas leakage fault, and p6 is the probability of grounding fault;

[0028] Arrange p1, p2, p3, p4, p5, and p6 in ascending order, and set the largest value among them as p max ;

[0029] Preset the probability threshold p t , when p max >p t , then determine whether there is max The corresponding fault type, when p max ≤p t , it is judged as normal state;

[0030] Step S60: Generate a warning information report based on the determination result of the fault type, trigger an alarm and send the warning information report to the remote maintenance terminal.

[0031] Preferably, in step S30, the preset network includes a physical layer, a network layer, a transport layer and an application layer. The physical layer and the network layer of the preset network use the NB-IoT protocol to transmit data frames, the transport layer of the preset network uses the UDP protocol to transmit data frames, and the application layer of the preset network uses the CoAP protocol.

[0032] Preferably, in step S30, the data packet format is: Packet = [DeviceID, Timestamp, M, CRC], Packet is the data packet, DeviceID is the unique identifier of the distribution box switch cabinet, Timestamp is the data collection time, and CRC is the cyclic redundancy check code.

[0033] Preferably, in step S50, the SVM fault classification model is trained offline using historical fault data. During offline training, the input is a standardized feature vector and the output is a true fault probability.

[0034] Preferably, in step S50, the SVM fault classification model calculates the fault probability by the following formula:

[0035]

[0036] Where k is the kth type of fault, ω k is the classification hyperplane parameter, b k is the offset, f k (x) is the score of the k-th type of fault, T is the transpose of the matrix;

[0037] Use the Softmax function to convert the k-th class score into probability.

[0038] Preferably, in step S60, the Softmax function is used to convert the k-th category score into a probability, and the formula is:

[0039]

[0040] Among them, p k is the probability of the kth type of failure, exp is the exponential mapping, and j represents the index.

[0041] Preferably, in step S10, the normalization process adopts a maximum-minimum normalization method.

[0042] The present invention also provides a distribution box switch cabinet fault detection system comprising:

[0043] The data acquisition module is used to collect environmental and equipment operating status data through various sensors installed in the distribution box switch cabinet, including:

[0044] Temperature sensor, used to collect the operating temperature T of the distribution box switch cabinet;

[0045] Humidity sensor, used to collect humidity H of the switch cabinet of the distribution box;

[0046] Gas concentration sensor, used to collect sulfur hexafluoride gas concentration G;

[0047] Current sensor and voltage sensor: used to collect current I and voltage V respectively;

[0048] The above data are normalized to generate a normalized multi-parameter data set D = {T norn , H norn , G norn , I norn , V norn};

[0049] The environment feature construction module is used to generate the environment feature matrix M by introducing dynamic weights and fusing multi-parameter data sets. The steps include:

[0050] Get the historical data of the sensor within the preset time period, and calculate the dynamic weight based on the standard deviation of the historical data of the sensor. The formula is:

[0051]

[0052] Among them, i is the i-th sensor, ω i is the standard deviation of the ith sensor, ω i is the dynamic weight of the i-th sensor;

[0053] Construct the environment feature matrix M after dynamic weighted fusion, M = {T norn ·ω T , H norn ·ω H , G norn ·ω G , I norn ·ω I , V norn ·ω V}, where ω T is the temperature dynamic weight, ω H is the humidity dynamic weight, ω G is the dynamic weight of gas concentration, ω I is the current dynamic weight, ω V is the voltage dynamic weight;

[0054] A data transmission module, used to package the environmental feature matrix M generated in step S20 and transmit it to a remote server via a preset network in the form of a data packet;

[0055] The data verification module is used to verify the data using the cyclic redundancy check code in the data packet after the remote server receives the data packet:

[0056] CRC packet =R(packet)mod G(X)

[0057] Among them, R(packet) is the binary polynomial representation of the data packet, G(X) is a predefined fixed polynomial, CRC packet is the calculated value of the cyclic redundancy check code, and mod is the remainder operation;

[0058] If the calculated value of the cyclic redundancy check code CRC packet Equal to the cyclic redundancy check code CRC, the check passes and the data packet is marked as valid;

[0059] If the calculated value of the cyclic redundancy check code CRC packet Not equal to the cyclic redundancy check code CRC, the check fails, and the device is requested to resend;

[0060] The fault classification module is used to input the feature matrix M in the data packet marked as valid into the SVM fault classification model and output the fault probability distribution P, where

[0061] P={p1,p2,p3,p4,p5,p6}

[0062] Among them, p1 is the probability of over-temperature fault, p2 is the probability of abnormal humidity fault, p3 is the probability of current overload fault, p4 is the probability of abnormal voltage fault, p5 is the probability of gas leakage fault, and p6 is the probability of grounding fault;

[0063] Arrange p1, p2, p3, p4, p5, and p6 in ascending order, and set the largest value among them as p max ;

[0064] Preset the probability threshold p t , when p max > t , then determine whether there is max The corresponding fault type, when p max ≤p t , it is judged as normal state;

[0065] The warning generation module is used to generate a warning information report based on the determination result of the fault type, trigger an alarm and send the warning information report to the remote maintenance terminal.

[0066] The present invention also provides a distribution box switch cabinet fault detection device, including a memory, a processor, and a distribution box switch cabinet fault detection program stored in the memory and executable on the processor. When the distribution box switch cabinet fault detection program is executed by the processor, the distribution box switch cabinet fault detection method is implemented.

[0067] The present invention also provides a computer program product, including a distribution box switch cabinet fault detection program, and the distribution box switch cabinet fault detection program implements the distribution box switch cabinet fault detection method when executed by a processor.

[0068] The beneficial effect of the present invention is that compared with the technical problems of low detection accuracy and unclear fault type determination in the prior art, especially under low bandwidth and high interference network conditions, which make it impossible to achieve real-time and reliable data transmission and fault monitoring, the present application realizes comprehensive analysis of multiple fault parameters by adopting dynamic weight fusion, preset network transmission processing and SVM-based fault classification model, thereby avoiding the misjudgment problem caused by reliance on a single parameter in traditional methods and improving the accuracy and real-time performance of fault detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0070] Figure 1 The present invention is a flow chart of a first embodiment of a method for detecting faults in a distribution box switch cabinet.

[0071] Figure 2 The present invention is a schematic diagram of a device for detecting a fault in a distribution box switch cabinet. DETAILED DESCRIPTION

[0072] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of 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.

[0073] Embodiment 1: Figure 1 , which is a flow chart of the first embodiment of the distribution box switch cabinet fault detection method of the present invention, and proposes the first embodiment of the distribution box switch cabinet fault detection method of the present invention.

[0074] In a first embodiment, the distribution box switch cabinet fault detection method includes:

[0075] Step S10: Collecting environment and equipment operation status data through various sensors installed in the switch cabinet of the distribution box, including:

[0076] Temperature sensor, used to collect the operating temperature T of the distribution box switch cabinet;

[0077] Humidity sensor, used to collect humidity H of the switch cabinet of the distribution box;

[0078] Gas concentration sensor, used to collect sulfur hexafluoride gas concentration G;

[0079] Current sensor and voltage sensor: used to collect current I and voltage V respectively;

[0080] The above data are normalized to generate a normalized multi-parameter data set D = {T norn , H norn , G norn , I norn , V norn};

[0081] It should be noted that multi-dimensional data of the environment and equipment operating status, such as temperature, humidity, current, etc., are collected in real time, covering important parameters that affect the safe operation of distribution box switch cabinets; early detection of insulating gas leakage can avoid significant degradation of electrical insulation performance due to gas leakage; after data normalization, a unified standard data set is formed, ensuring the reasonable distribution of relative weights between multiple parameters in subsequent analysis; with the comprehensive data support of temperature, humidity, gas concentration, current, and voltage, fault hazards can be identified earlier and early warnings can be issued, thereby improving the operating reliability of equipment and the stability of the power system.

[0082] Step S20: Generate an environmental feature matrix M by introducing dynamic weights and fusing multi-parameter data sets. The steps include:

[0083] Get the historical data of the sensor within the preset time period, and calculate the dynamic weight based on the standard deviation of the historical data of the sensor. The formula is:

[0084]

[0085] Among them, i is the i-th sensor, ω i is the standard deviation of the ith sensor, ω i is the dynamic weight of the i-th sensor;

[0086] Construct the environment feature matrix M after dynamic weighted fusion, M = {T norn ·ω T , H norn ·ω H ,

[0087] G norn ·ω G , I norn ·ω I , V norn ·ω V}, where ω T is the temperature dynamic weight, ω H is the humidity dynamic weight, ω G is the dynamic weight of gas concentration, ωI is the current dynamic weight, ω V is the voltage dynamic weight;

[0088] It should be noted that different sensors are affected to different degrees by factors such as noise and interference in the operating environment. The introduction of dynamic weights can improve the credibility of sensor data; dynamic weights are updated in real time as sensor performance or environmental conditions change, enhancing the adaptability and robustness of the system.

[0089] It is understandable that after weighting the features, data with higher stability, such as sensors with smaller standard deviations, have higher weights in the analysis, ensuring more reliable results; changes in weights reflect the contribution of sensors to fault detection in different environments, thereby optimizing overall detection performance. Fixed weights are usually unable to adapt to measurement fluctuations caused by sensor aging, environmental changes, etc.; dynamic weights are automatically adjusted according to historical standard deviations, which can better reflect the actual situation of the current environment, thereby improving the accuracy of detection.

[0090] Step S30: Packing the environmental feature matrix M generated in step S20 and transmitting it to the remote server via a preset network in the form of a data packet;

[0091] It should be noted that the preset network includes a physical layer, a network layer, a transport layer, and an application layer. The physical layer and the network layer of the preset network use the NB-IoT protocol for data frame transmission, the transport layer of the preset network uses the UDP protocol for data frame transmission, and the application layer of the preset network uses the CoAP protocol.

[0092] It should be understood that the NB-IoT network can provide stable low-bandwidth communications in wide coverage and complex environments, which is suitable for real-time monitoring applications of distribution box switch cabinets; the cyclic redundancy check code can quickly determine whether the data packet is damaged or lost during transmission;

[0093] Step S40: After receiving the data packet, the remote server uses the cyclic redundancy check code in the data packet to perform data verification:

[0094] CRC packet =R(packet)mod G(X)

[0095] Among them, R(packet) is the binary polynomial representation of the data packet, G(X) is a predefined fixed polynomial, CRC packet is the calculated value of the cyclic redundancy check code, and mod is the remainder operation;

[0096] If the calculated value of the cyclic redundancy check code CRC packet Equal to the cyclic redundancy check code CRC, the check passes and the data packet is marked as valid;

[0097] If the calculated value of the cyclic redundancy check code CRC packet Not equal to the cyclic redundancy check code CRC, the check fails, and the device is requested to resend;

[0098] It should be noted that CRC check is a fast and efficient error detection method, which is particularly suitable for the integrity verification of data transmission in network communication. It has a small amount of calculation and is suitable for embedded systems or low-bandwidth network scenarios, such as NB-IoT networks.

[0099] It is understandable that data may be affected by noise, signal interference, etc. during transmission, resulting in partial bit errors. CRC check detects these errors to avoid misclassification of faults caused by data packet errors.

[0100] Step S50: Input the feature matrix M in the data packet marked as valid into the SVM fault classification model and output the fault probability distribution P, where:

[0101] P={p1,p2,p3,p4,p5,p6}

[0102] Among them, p1 is the probability of over-temperature fault, p2 is the probability of abnormal humidity fault, p3 is the probability of current overload fault, p4 is the probability of abnormal voltage fault, p5 is the probability of gas leakage fault, and p6 is the probability of grounding fault;

[0103] Arrange p1, p2, p3, p4, p5, and p6 in ascending order, and set the largest value among them as p max ;

[0104] Preset the probability threshold p t , when p max >p t , then determine whether there is max The corresponding fault type, when p max ≤p t , it is judged as normal state;

[0105] It should be understood that the SVM model can effectively process high-dimensional input data, such as the environmental feature matrix M, and has strong generalization ability; compared with the traditional rule judgment method, the SVM model can dynamically adjust the classification boundaries according to historical data to improve the classification accuracy.

[0106] It should be understood that the maximum probability method is to select p max Significantly reduces the uncertainty of the classification results. Even if the probabilities of multiple categories are close, they can still be classified according to p max Make the best choice and reduce the risk of misjudgment.

[0107] Step S60: Generate a warning information report based on the determination result of the fault type, trigger an alarm and send the warning information report to the remote maintenance terminal.

[0108] It should be noted that the early warning information report contains the status of the equipment and recommended maintenance actions, including:

[0109] Equipment identification, used to uniquely locate the faulty distribution box switch cabinet;

[0110] Fault type: the determined fault type, such as over-temperature fault, abnormal humidity fault, etc.;

[0111] Failure probability, the maximum probability associated with the failure type, reflecting the severity of the problem;

[0112] Trigger time: records the time when the fault event occurs, which is used for event tracing;

[0113] Recommended actions provide specific maintenance recommendations for the type of fault.

[0114] It is understandable that the system triggers an alarm based on the type and severity of the fault, and the alarm forms include on-site buzzer, warning light, and remote notification such as SMS, APP push and email.

[0115] In addition, the distribution box switch cabinet fault detection system provided by the present invention adopts a distribution box switch cabinet fault detection method in the above embodiment, which can solve the technical problem of distribution box switch cabinet fault detection. Compared with the prior art, the beneficial effect of the distribution box switch cabinet fault detection system provided by the present invention is the same as the beneficial effect of the distribution box switch cabinet fault detection method provided by the above embodiment, and the other technical features of the distribution box switch cabinet fault detection system are the same as the features disclosed in the above embodiment method, which will not be repeated here.

[0116] The present invention provides a distribution box switch cabinet fault detection device, please refer to Figure 2, a distribution box switch cabinet fault detection device includes: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute a distribution box switch cabinet fault detection method in the above-mentioned embodiment 1. A distribution box switch cabinet fault detection device in the embodiment of the present invention may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. A distribution box switch cabinet fault detection device is only an example and should not bring any limitations to the functions and scope of use of the embodiments of the present invention. A distribution box switch cabinet fault detection device may include a processing device 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM) 1004. In RAM 1004, various programs and data required for the operation of a distribution box switch cabinet fault detection device are also stored. The processing device 1001, ROM 1002, and RAM 1004 are connected to each other via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to the I / O interface 1006: input devices 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; storage devices 1003 including, for example, a magnetic tape, a hard disk, etc.; and communication devices 1009. The communication device 1009 can allow a distribution box switch cabinet fault detection device to communicate wirelessly or wired with other devices to exchange data. Although a distribution box switch cabinet fault detection device with various systems is shown in the figure, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems may be implemented or have alternatively.

[0117] The present invention also provides a computer program product, including a computer program, wherein when the computer program is executed by a processor, the steps of the above-mentioned method for detecting a fault in a distribution box switch cabinet are implemented. The computer program product provided by the present invention can solve the technical problem of detecting a fault in a distribution box switch cabinet. Compared with the prior art, the beneficial effects of the computer program product provided by the present invention are the same as the beneficial effects of the method for detecting a fault in a distribution box switch cabinet provided by the above-mentioned embodiment, and will not be described in detail here.

[0118] In particular, according to the embodiments disclosed in the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present invention include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through a communication device, or installed from the storage device 1003, or installed from the ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present invention are executed.

[0119] It should be understood that the various parts disclosed in the present invention can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0120] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for detecting faults in a distribution box switch cabinet, characterized in that: Methods include: Step S10: Collecting environment and equipment operation status data through various sensors installed in the switch cabinet of the distribution box, including: Temperature sensor, used to collect the operating temperature T of the distribution box switch cabinet; Humidity sensor, used to collect humidity H of the distribution box switch cabinet; Gas concentration sensor, used to collect sulfur hexafluoride gas concentration G; Current sensor and voltage sensor: used to collect current I and voltage V respectively; The above data are normalized to generate a normalized multi-parameter data set D = {T norn , H norn , G norn , I norn , V norn }; Step S20: Generate an environmental feature matrix M by introducing dynamic weights and fusing multi-parameter data sets. The steps include: Get the historical data of each sensor within the preset time period, and calculate the dynamic weight based on the standard deviation of the historical data of each sensor. The formula is: Among them, i is the i-th sensor, ω i is the standard deviation of the ith sensor, ω i is the dynamic weight of the i-th sensor; Construct the environment feature matrix M after dynamic weighted fusion, M = {T norn ·ω T , H norn ·ω H , G norn ·ω G , I norn ·ω I , V norn ·ω V }, where ω T is the temperature dynamic weight, ω H is the humidity dynamic weight, ω G is the dynamic weight of gas concentration, ω I is the current dynamic weight, ω V is the voltage dynamic weight; Step S30: Packing the environmental feature matrix M generated in step S20 and transmitting it to the remote server via a preset network in the form of a data packet; Step S40: After receiving the data packet, the remote server uses the cyclic redundancy check code in the data packet to perform data verification: CRC packet =R(packet)mod G(X) Among them, R(packet) is the binary polynomial representation of the data packet, G(X) is a predefined fixed polynomial, CRC packet is the calculated value of the cyclic redundancy check code, and mod is the remainder operation; If the calculated value of the cyclic redundancy check code CRC packet Equal to the cyclic redundancy check code CRC, the check passes and the data packet is marked as valid; If the calculated value of the cyclic redundancy check code CRC packet Not equal to the cyclic redundancy check code CRC, the check fails, and the device is requested to resend; Step S50: Input the feature matrix M in the data packet marked as valid into the SVM fault classification model and output the fault probability distribution P, where: P={p1,p2,p3,p4,p5,p6} Among them, p1 is the probability of over-temperature fault, p2 is the probability of abnormal humidity fault, p3 is the probability of current overload fault, p4 is the probability of abnormal voltage fault, p5 is the probability of gas leakage fault, and p6 is the probability of grounding fault; Arrange p1, p2, p3, p4, p5, and p6 in ascending order, and set the largest value among them as p max ; Preset the probability threshold p t , when p max >p t , then determine whether there is max The corresponding fault type, when p max ≤p t , it is judged as normal state; Step S60: Generate a warning information report based on the determination result of the fault type, trigger an alarm and send the warning information report to the remote maintenance terminal.

2. A method for detecting faults in a distribution box switch cabinet according to claim 1, characterized in that: In step S30, the preset network includes a physical layer, a network layer, a transport layer and an application layer. The physical layer and the network layer of the preset network use the NB-IoT protocol to transmit data frames, the transport layer of the preset network uses the UDP protocol to transmit data frames, and the application layer of the preset network uses the CoAP protocol.

3. A method for detecting faults in a distribution box switch cabinet according to claim 1, characterized in that: In step S30, the data packet format is: Packet = [DeviceID, Timestamp, M, CRC], Packet is the data packet, DeviceID is the unique identifier of the distribution box switch cabinet, Timestamp is the data collection time, and CRC is the cyclic redundancy check code.

4. A method for detecting faults in a distribution box switch cabinet according to claim 1, characterized in that: In step S50, the SVM fault classification model is trained offline using historical fault data. During offline training, the input is a standardized feature vector and the output is a true fault probability.

5. A method for detecting faults in a distribution box switch cabinet according to claim 1, characterized in that: In step S50, the SVM fault classification model calculates the fault probability by the following formula: Where k is the kth type of fault, ω k is the classification hyperplane parameter, b k is the offset, f k (x) is the score of the k-th type of fault, T is the transpose of the matrix; Use the Softmax function to convert the k-th class score into probability.

6. A method for detecting faults in a distribution box switch cabinet as claimed in claim 5, characterized in that: In step S60, the Softmax function is used to convert the k-th class score into probability, and the formula is: Among them, p k is the probability of the kth type of failure, exp is the exponential mapping, and j represents the index.

7. A method for detecting faults in a distribution box switch cabinet according to claim 1, characterized in that: In step S10, the normalization process adopts the maximum and minimum normalization method.

8. A distribution box switch cabinet fault detection system, characterized in that: The distribution box switch cabinet fault detection system comprises: The data acquisition module is used to collect environmental and equipment operating status data through various sensors installed in the distribution box switch cabinet, including: Temperature sensor, used to collect the operating temperature T of the distribution box switch cabinet; Humidity sensor, used to collect humidity H of the switch cabinet of the distribution box; Gas concentration sensor, used to collect sulfur hexafluoride gas concentration G; Current sensor and voltage sensor: used to collect current I and voltage V respectively; The above data are normalized to generate a normalized multi-parameter data set D = {T norn , H norn , G norn , I norn , V norn }; The environment feature construction module is used to generate the environment feature matrix M by introducing dynamic weights and fusing multi-parameter data sets. The steps include: Get the historical data of the sensor within the preset time period, and calculate the dynamic weight based on the standard deviation of the historical data of the sensor. The formula is: Among them, i is the i-th sensor, ω i is the standard deviation of the ith sensor, ω i is the dynamic weight of the i-th sensor; Construct the environment feature matrix M after dynamic weighted fusion, M = {T norn ·ω T , H norn ·ω H , G norn ·ω G , I norn ·ω I , V norn ·ω V }, where ω T is the temperature dynamic weight, ω H is the humidity dynamic weight, ω G is the dynamic weight of gas concentration, ω I is the current dynamic weight, ω V is the voltage dynamic weight; A data transmission module, used to package the environmental feature matrix M generated in step S20 and transmit it to a remote server via a preset network in the form of a data packet; The data verification module is used to verify the data using the cyclic redundancy check code in the data packet after the remote server receives the data packet: CRC packet =R(packet)mod G(X) Among them, R(packet) is the binary polynomial representation of the data packet, G(X) is a predefined fixed polynomial, CRC packet is the calculated value of the cyclic redundancy check code, and mod is the remainder operation; If the calculated value of the cyclic redundancy check code CRC packet Equal to the cyclic redundancy check code CRC, the check passes and the data packet is marked as valid; If the calculated value of the cyclic redundancy check code CRC packet Not equal to the cyclic redundancy check code CRC, the check fails, and the device is requested to resend; The fault classification module is used to input the feature matrix M in the data packet marked as valid into the SVM fault classification model and output the fault probability distribution P, where P={p1, p2, p3, p4, p5, p6} Among them, p1 is the probability of over-temperature fault, p2 is the probability of abnormal humidity fault, p3 is the probability of current overload fault, p4 is the probability of abnormal voltage fault, p5 is the probability of gas leakage fault, and p6 is the probability of grounding fault; Arrange p1, p2, p3, p4, p5, and p6 in ascending order, and set the largest value among them as P max ; Preset the probability threshold p t , when pma x >p t , then it is determined that there is a x The corresponding fault type, when pma x ≤p t , it is judged as normal state; The warning generation module is used to generate a warning information report based on the determination result of the fault type, trigger an alarm and send the warning information report to the remote maintenance terminal.

9. A distribution box switch cabinet fault detection device, characterized in that: The distribution box switch cabinet fault detection device includes: a memory, a processor, and a distribution box switch cabinet fault detection program stored in the memory and executable on the processor. When the distribution box switch cabinet fault detection program is executed by the processor, the distribution box switch cabinet fault detection method described in any one of claims 1 to 7 is implemented.

10. A computer program product, characterized in that The computer program product includes a distribution box switch cabinet fault detection program, and when the distribution box switch cabinet fault detection program is executed by a processor, it implements the distribution box switch cabinet fault detection method described in any one of Claims 1 to 7.

Images