Substation fault detection method, device, computer equipment and storage medium

By deploying multiple target sensors in the substation, using wireless network and fault detection models to automatically judge substation faults, the problem of low efficiency in the existing technology is solved, and the rapid and accurate detection and positioning of substation faults is achieved.

CN113869444BActive Publication Date: 2024-11-08ANNING BUREAU OF ULTRA HIGH VOLTAGE TRANSMISSION

Patent Information

Application Number
CN202111176423.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-10-09
Publication Date
2024-11-08
Estimated Expiration
2041-10-09

AI Technical Summary

Technical Problem

In the prior art, substation fault detection relies on manual inspection, which is inefficient and time-consuming, making it difficult to meet the needs of diversity and complexity of electrical equipment in large-scale power grids.

Method used

By deploying multiple target sensors in the substation, using tree or multi-hop network topology for wireless monitoring, obtaining target status data and historical fault status data, building a fault monitoring matrix and historical fault matrix, calculating the mutual information difference, and determining whether the substation is There is a fault and enter data into the fault detection model to determine the fault location and type.

Benefits of technology

It realizes the automation, real-time judgment and rapid positioning of substation faults, improves fault detection efficiency, reduces the demand for manual inspection, and adapts to the monitoring needs of complex equipment in large-scale power grids.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN113869444B_ABST
    Figure CN113869444B_ABST
Patent Text Reader

Abstract

The present application relates to a substation fault detection method, device, computer equipment and storage medium. The method comprises: obtaining target state data obtained by a target sensor monitoring a substation within a target time period; obtaining fault state data corresponding to a target sensor and a historical time period, wherein the fault state data is obtained by the target sensor monitoring a substation fault; performing a similarity comparison between the target state data and the fault state data to obtain a comparison result, and determining whether the substation has a fault based on the comparison result. The use of this method can improve the efficiency of substation fault detection.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of power systems, and in particular to a substation fault detection method, device, computer equipment and storage medium. Background Art

[0002] In the power system, ensuring the safety and stability of substations is an important part of achieving the normal operation of the power system. Therefore, fault detection of substations is of great significance.

[0003] In the related art, manual inspection is usually performed, and the inspectors judge whether there is a fault in the substation based on the appearance of the substation equipment or information such as indicating instruments.

[0004] However, there are many types of electrical equipment in substations. If substation fault detection is achieved only by manpower, it will consume a lot of manpower and material resources, and the efficiency of substation fault detection is low. Summary of the invention

[0005] Based on this, it is necessary to provide a substation fault detection method, device, computer equipment and storage medium that can improve the efficiency of substation fault detection in response to the above technical problems.

[0006] In a first aspect, a substation fault detection method is provided, the method comprising:

[0007] Obtain target status data obtained by the target sensor monitoring the substation within the target time period;

[0008] Obtaining fault status data of the target sensor corresponding to the historical time period, where the fault status data is obtained by monitoring the target sensor when a substation fails;

[0009] The target state data and the fault state data are compared for similarity to obtain a comparison result, and whether the substation has a fault is determined based on the comparison result.

[0010] In one embodiment, the target state data and the fault state data are compared for similarity to obtain a comparison result, including:

[0011] Construct a fault monitoring matrix based on the target status data;

[0012] According to the fault status data, a historical fault matrix is ​​constructed;

[0013] The mutual information between the fault monitoring matrix and the historical fault matrix is ​​obtained, and a comparison result is obtained according to the mutual information.

[0014] In one embodiment, the target state data includes the location information of the target sensor and the target perception data corresponding to at least one sampling time in the target time period. According to the target state data, a fault monitoring matrix is ​​constructed, including:

[0015] For each sampling moment, the row vector of the matrix is ​​determined according to the position information of the target sensor, and the column vector of the matrix is ​​determined according to the target perception data corresponding to the target sensor and the sampling moment;

[0016] Based on the matrix row vectors and the matrix column vectors, a target fault monitoring matrix corresponding to the sampling time is constructed, and the target fault monitoring matrix corresponding to each sampling time is used as the fault monitoring matrix.

[0017] In one embodiment, the historical fault matrix includes a target historical fault matrix corresponding to at least one fault moment in the historical time period, obtaining mutual information between the fault monitoring matrix and the historical fault matrix, and obtaining a comparison result based on the mutual information, including:

[0018] For each sampling moment, determine the target fault moment corresponding to the sampling moment, and calculate the target mutual information between the target fault monitoring matrix corresponding to the sampling moment and the target historical fault matrix corresponding to the target fault moment;

[0019] Calculate the mutual information difference of the target mutual information corresponding to two adjacent sampling moments in each sampling moment;

[0020] Compare each target mutual information with a first threshold to obtain a first comparison result, and compare each mutual information difference with a second threshold to obtain a second comparison result;

[0021] Each first comparison result and each second comparison result are taken as a comparison result.

[0022] In one of the embodiments, the target sensor is wirelessly deployed in a substation; there are multiple target sensors, and the multiple target sensors are distributed in the substation in a tree network topology structure or a multi-hop network topology structure.

[0023] In one embodiment, the method further comprises:

[0024] If there is a fault in the substation, the target state data is input into the fault detection model to obtain the fault location and fault type of the substation;

[0025] The fault location and fault type of the substation are matched with the fault information in the fault knowledge base to obtain the fault solution corresponding to the fault location and fault type.

[0026] In one embodiment, the training process of the fault detection model includes:

[0027] Acquire training data, where the training data includes location information of a target sensor, sensor perception data corresponding to multiple historical fault moments, fault location information corresponding to each historical fault moment, and fault type corresponding to each historical fault moment;

[0028] The initial fault detection model is iteratively trained using the training data to obtain a fault detection model.

[0029] In a second aspect, a substation fault detection device is provided, the device comprising:

[0030] An acquisition module is used to acquire target state data obtained by a target sensor monitoring a substation within a target time period; acquire fault state data corresponding to the target sensor and a historical time period, wherein the fault state data is obtained by the target sensor monitoring when the substation fails;

[0031] The judgment module is used to compare the target state data with the fault state data for similarity to obtain a comparison result, and determine whether the substation has a fault according to the comparison result.

[0032] In one embodiment, the determination module is specifically used for:

[0033] Construct a fault monitoring matrix based on the target status data;

[0034] According to the fault status data, a historical fault matrix is ​​constructed;

[0035] The mutual information between the fault monitoring matrix and the historical fault matrix is ​​obtained, and a comparison result is obtained according to the mutual information.

[0036] In one embodiment, the target state data includes location information of the target sensor and target perception data corresponding to at least one sampling time in the target time period, and the judgment module is further used to:

[0037] For each sampling moment, the row vector of the matrix is ​​determined according to the position information of the target sensor, and the column vector of the matrix is ​​determined according to the target perception data corresponding to the target sensor and the sampling moment;

[0038] Based on the matrix row vectors and the matrix column vectors, a target fault monitoring matrix corresponding to the sampling time is constructed, and the target fault monitoring matrix corresponding to each sampling time is used as the fault monitoring matrix.

[0039] In one embodiment, the historical fault matrix includes a target historical fault matrix corresponding to at least one fault moment in the historical time period, and the judgment module is further used to:

[0040] For each sampling moment, determine the target fault moment corresponding to the sampling moment, and calculate the target mutual information between the target fault monitoring matrix corresponding to the sampling moment and the target historical fault matrix corresponding to the target fault moment;

[0041] Calculate the mutual information difference of the target mutual information corresponding to two adjacent sampling moments in each sampling moment;

[0042] Compare each target mutual information with a first threshold to obtain a first comparison result, and compare each mutual information difference with a second threshold to obtain a second comparison result;

[0043] Each first comparison result and each second comparison result are taken as a comparison result.

[0044] In one of the embodiments, the target sensor is wirelessly deployed in a substation; there are multiple target sensors, and the multiple target sensors are distributed in the substation in a tree network topology structure or a multi-hop network topology structure.

[0045] In one embodiment, the device further comprises:

[0046] An output module is used to input the target state data into the fault detection model when there is a fault in the substation, so as to obtain the fault location and fault type of the substation;

[0047] The solution module is used to match the fault location and fault type of the substation with the fault information in the fault knowledge base to obtain a fault solution corresponding to the fault location and fault type.

[0048] In one embodiment, the training process of the fault detection model includes:

[0049] Acquire training data, where the training data includes location information of a target sensor, sensor perception data corresponding to multiple historical fault moments, fault location information corresponding to each historical fault moment, and fault type corresponding to each historical fault moment;

[0050] The initial fault detection model is iteratively trained using the training data to obtain a fault detection model.

[0051] In a third aspect, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the substation fault detection method as described in the first aspect is implemented.

[0052] In a fourth aspect, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the substation fault detection method as described in the first aspect is implemented.

[0053] The above-mentioned substation fault detection method, device, computer equipment and storage medium obtain the target state data obtained by the target sensor monitoring the substation within the target time period and the fault state data corresponding to the target sensor and the historical time period, compare the target state data and the fault state data for similarity to obtain a comparison result, and determine whether the substation has a fault based on the comparison result. Since the fault state data is obtained by the target sensor monitoring when the substation fails, the fault state data can characterize the state characteristics of the target sensor when the substation fails. In this way, the target state data and the fault state data are compared for similarity. If the target state data and the fault state data have a large similarity, it characterizes that the target state data has the state characteristics of the target sensor when the substation fails, so it can be determined that the substation has a fault. If the target state data and the fault state data have a small similarity, it characterizes that the target state data does not have the state characteristics of the target sensor when the substation fails, so it can be determined that the substation does not have a fault, thereby realizing automatic judgment of substation faults and real-time fault judgment, and improving the efficiency of substation fault detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 A schematic diagram of a process flow of a substation fault detection method in one embodiment;

[0055] Figure 2 is a flow chart of step 103 in one embodiment;

[0056] Figure 3 is a flow chart of step 201 in one embodiment;

[0057] Figure 4 is a flow chart of step 203 in one embodiment;

[0058] Figure 5 is a schematic diagram of topological distribution of target sensors in one embodiment;

[0059] Figure 6 A schematic diagram of a process flow of a substation fault detection method in one embodiment;

[0060] Figure 7 A schematic diagram of a process flow of a fault detection model training process in one embodiment;

[0061] Figure 8 A schematic diagram of a process flow of a substation fault detection method in one embodiment;

[0062] Fig. 9 is a structural block diagram of a substation fault detection device in one embodiment;

[0063] Fig.10is a structural block diagram of a substation fault detection device in one embodiment;

[0064] Fig.11 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0065] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0066] In the power system, ensuring the safety and stability of substations is an important part of achieving the normal operation of the power system. Therefore, fault detection of substations is of great significance.

[0067] In the related art, manual inspection is usually performed, and the inspectors judge whether there is a fault in the substation based on the appearance of the substation equipment or information such as indicating instruments.

[0068] However, as the scale of the power grid increases, there are more and more electrical equipment in the substation, and the number of manual inspection items is also increasing. The inspection items are diverse, such as detecting whether the reactor and resistor of the ice melting filter are burned, detecting whether the temperature and humidity in the operation box are too high, detecting whether the main pump is abnormal, detecting whether there are small animals in the substation, etc. If substation fault detection is achieved only by manpower, the efficiency of substation fault detection is low and can no longer meet the ever-changing needs.

[0069] In view of this, an embodiment of the present application provides a substation fault detection method to automatically determine whether a substation is faulty.

[0070] It should be noted that the substation fault detection method provided in the embodiment of the present application may be executed by a substation fault detection device, and the substation fault detection method may be implemented as part or all of a terminal through software, hardware, or a combination of software and hardware.

[0071] In the following method embodiments, the execution subject is taken as an example to illustrate, wherein the terminal can be a personal computer, a laptop computer, a media player, a smart TV, a smart phone, a tablet computer, and a portable wearable device, etc. It can be understood that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server.

[0072] Please refer to Figure 1 , which shows a flow chart of a substation fault detection method provided by an embodiment of the present application. Figure 1 As shown, the substation fault detection method may include the following steps:

[0073] Step 101, obtaining target state data obtained by a target sensor monitoring a substation within a target time period.

[0074] The target sensor can be deployed in the substation. After the target sensor is powered on, it senses the monitored object and environment in the substation to obtain the target status data.

[0075] Exemplarily, the target state data may include deployment location information of the target sensor, target perception data, and location information of the monitoring object of the target sensor, etc. The monitoring object may be equipment in the substation and small animals in the substation, etc.

[0076] The following is an introduction to the contents of target perception data.

[0077] In the embodiment of the present application, the target sensor includes at least one of a humidity sensor, a temperature sensor, a vibration sensor, a displacement sensor, an infrared sensor, and a sound sensor. Among them, the humidity sensor is used to monitor whether the humidity of the mechanism box exceeds the standard; the temperature sensor is used to monitor whether the humidity of the mechanism box exceeds the standard; the vibration sensor is used to monitor whether the vibration acceleration of the valve cooling main pump exceeds the standard; the displacement sensor is used to monitor the displacement of the expansion joint of the GIS or GIL equipment; the infrared sensor is used to detect whether there are small animals in the substation; the sound sensor is used to detect whether the sound of the transformer operation is abnormal.

[0078] In this way, the target sensing data may include humidity data inside the mechanism box sensed by a humidity sensor, temperature data inside the mechanism box sensed by a temperature sensor; vibration acceleration data of the valve cooling main pump sensed by a vibration sensor, displacement data of the expansion joint of the GIS or GIL equipment sensed by a displacement sensor, thermal image data of the surface temperature of small animals sensed by an infrared sensor, and at least one of the transformer operation sound signals sensed by a sound sensor.

[0079] Optionally, the terminal is provided with a sampling period, for example, the sampling period can be set to 20 minutes. Optionally, the target time period is a time period with the current time as the end point in terms of timing. The length of the target time period is proportional to the sampling period. According to the sampling period, the sampling moment corresponding to the target time is determined, and the target perception data includes the target perception data obtained by the target sensor monitoring the substation at each sampling moment.

[0080] Optionally, when the sampling time is reached, the terminal sends a request instruction for target sensing data to the target sensor, and after receiving the request, the target sensing data uploads the target sensing data obtained by monitoring the substation to the terminal. The terminal receives the target sensing data for substation fault detection.

[0081] Step 102, obtaining fault status data of the target sensor corresponding to the historical time period, where the fault status data is obtained by monitoring the target sensor when a substation fails.

[0082] The historical time period is any time period that is before the target time period in terms of time sequence. The number of the historical time periods can be one or more. The length of the historical time period can be the same as or different from the length of the target time period.

[0083] When the terminal detects that a fault exists in the substation, it determines the moment as the fault moment, and saves the state data monitored by the target sensor corresponding to the fault moment as fault state data in the corresponding storage path.

[0084] Among them, the number of samples of fault status data is related to the number of target sensors. The fault status data may be fault status data at multiple fault moments corresponding to the same historical time period, or it may be fault status data at a certain fault moment in different time periods. When the number of target sensors is multiple, the fault status data at least includes fault status data when the monitoring object of each target sensor is abnormal. Taking the number of target sensors as 2 as an example, the monitoring objects of the target sensors are device A and device B respectively, then the fault status data includes the fault status data of the target sensor when only device A is faulty, the fault status data of the target sensor when only device B is faulty, and the fault status data of devices A and B.

[0085] Optionally, the terminal divides the fault status data into multiple object types according to the target sensor corresponding to the detected abnormality. When saving the fault status data, the terminal correspondingly saves the object type, historical time period and fault time corresponding to the fault status data.

[0086] Step 103, performing a similarity comparison between the target state data and the fault state data to obtain a comparison result, and determining whether there is a fault in the substation according to the comparison result.

[0087] Optionally, the terminal selects a fault status data under each object type. For each sampling moment, the target status data corresponding to the sampling moment is compared with each fault status data for similarity. Optionally, a similarity algorithm is used to calculate the similarity value between the target status data corresponding to the sampling moment and each fault status data to obtain multiple similarity values. The terminal can preset a threshold, compare each similarity value with the threshold, and then sum or weighted sum multiple similarities to obtain a final similarity value. If the final similarity value exceeds the threshold, it means that the substation has a fault, otherwise the substation has not failed.

[0088] Optionally, the terminal determines the data at the end of the target time period as the target state data, and determines the data at the end of the historical time period as the historical fault data. The similarity value between the target state data and the fault state data is calculated using a similarity algorithm to obtain a similarity value. The terminal can preset a threshold value and compare the similarity value with the threshold value. If the similarity value exceeds the threshold value, it means that the substation has a fault, otherwise the substation has no fault.

[0089] Optionally, when the length of the historical time period is the same as the length of the target time period, the terminal matches each sampling moment of the target time period with each fault moment in the historical time period, calculates the similarity value between the target state data corresponding to each sampling moment and the fault state data corresponding to the corresponding fault moment, obtains multiple similarity values, and then sums or weighted sums the multiple similarities to obtain the final similarity value. The terminal can preset a threshold value, compare the final similarity value with the threshold value, if the final similarity value exceeds the threshold value, it means that the substation has a fault, otherwise the substation has no fault.

[0090] This embodiment obtains the target state data obtained by the target sensor monitoring the substation within the target time period and the fault state data corresponding to the target sensor and the historical time period, compares the target state data and the fault state data for similarity to obtain a comparison result, and determines whether the substation has a fault based on the comparison result. Since the fault state data is obtained by the target sensor monitoring when the substation fails, the fault state data can characterize the state characteristics of the target sensor when the substation fails. In this way, the target state data and the fault state data are compared for similarity. If the target state data and the fault state data have a large similarity, it characterizes that the target state data has the state characteristics of the target sensor when the substation fails, so it can be determined that the substation has a fault. If the target state data and the fault state data have a small similarity, it characterizes that the target state data does not have the state characteristics of the target sensor when the substation fails, so it can be determined that the substation does not have a fault. The automatic judgment of substation faults and the real-time nature of fault judgment are realized, thereby improving the efficiency of substation fault detection.

[0091] In the implementation of this application, see Figure 2 ,based on Figure 1 The embodiment shown in the figure, this embodiment involves comparing the target state data and the fault state data for similarity in step 103, obtaining a comparison result, and determining whether the substation has a fault according to the comparison result, including steps 201, 202 and 203:

[0092] Step 201: construct a fault monitoring matrix according to target state data.

[0093] The target state data includes the location information of the target sensor and the target perception data; the location information is the deployment location information of the target sensor or the location information of the monitoring object of the target sensor.

[0094] The deployment location information of the target sensor and the location information of the monitoring object of the target sensor are in one-to-one correspondence. Optionally, the terminal stores a mapping relationship table of the deployment location information of the target sensor and the location information of the monitoring object of the target sensor.

[0095] In an optional implementation, the terminal determines the row vector of the matrix according to the location information of the target sensor, and determines the column vector of the matrix according to the target sensing data of the target sensor, and constructs a fault monitoring matrix X, where X=(x i,j ) M×N , x i,j represents the target sensing data of a target sensor at a certain location information, M is the number of location information, and N is the number of target sensors. In another optional implementation, the terminal determines the row vector of the matrix according to the target sensing data of the target sensor, and determines the column vector of the matrix according to the location information of the target sensor, and constructs a fault monitoring matrix X, where X=(x i,j ) N×M , x i,j It represents the target perception data of a certain target sensor at a certain position information, M is the number of position information, and N is the number of target sensors.

[0096] Step 202: construct a historical fault matrix based on the fault status data.

[0097] Among them, the fault state data includes the location information of the target sensor and the target historical perception data monitored by the target sensor under the fault state of the substation.

[0098] Optionally, when determining the row vector of the matrix according to the location information of the target sensor and determining the column vector of the matrix according to the target perception data of the target sensor to construct the fault monitoring matrix X, the terminal determines the row vector of the matrix according to the location information of the target sensor and determines the column vector of the matrix according to the target historical perception data of the target sensor to construct the historical fault matrix Y, where Y=(y i,j ) M×N ,y i,j It represents the historical target perception data of a target sensor at a certain location information, M is the number of location information, and N is the number of target sensors.

[0099] Optionally, when determining the row vector of the matrix according to the target perception data of the target sensor and determining the column vector of the matrix according to the location information of the target sensor to construct the fault monitoring matrix X, the terminal determines the row vector of the matrix according to the target historical perception data of the target sensor and determines the column vector of the matrix according to the location information of the target sensor to construct the historical fault matrix Y, where Y=(y i,j ) N×M ,y i,j It represents the historical target perception data of a target sensor at a certain location information, M is the number of location information, and N is the number of target sensors.

[0100] Step 203: Obtain mutual information between the fault monitoring matrix and the historical fault matrix, and obtain a comparison result according to the mutual information.

[0101] Optionally, calculate the mutual information MI(X,Y) between the fault monitoring matrix X and the historical fault matrix Y:

[0102]

[0103] Among them, P X (i) = ∑ j P X,Y (i,j),P Y (j) = ∑ i P X,Y (i, j) are the marginal probability distributions of the fault monitoring matrix X and the historical fault matrix Y, respectively. X,Y (i, j) is the joint probability distribution of the fault monitoring matrix X and the historical fault matrix Y.

[0104] The embodiment of the present application constructs a fault monitoring matrix according to the target state data and a historical fault matrix according to the fault state data, obtains the mutual information of the fault monitoring matrix and the historical fault matrix, and obtains the comparison result according to the mutual information, thereby realizing a rapid judgment on whether the current monitoring data has an abnormality and reducing the difficulty of substation fault detection.

[0105] In the present application example, the target state data includes the location information of the target sensor and the target perception data corresponding to at least one sampling time in the target time period, and the fault state data includes the target historical perception data corresponding to at least one fault time in the historical time period. Figure 3 ,based on Figure 2 In the embodiment shown, the implementation process of the above steps 201 and 202 includes steps 301, 302, 303 and 304:

[0106] Step 301, for each sampling moment, determine the row vector of the first matrix according to the position information of the target sensor, and determine the column vector of the first matrix according to the target perception data corresponding to the target sensor and the sampling moment.

[0107] The location information is deployment location information of the target sensor or location information of a monitoring object of the target sensor.

[0108] Optionally, the terminal is provided with a sampling period, for example, the sampling period may be set to 20 minutes. The sampling time is determined according to the sampling period.

[0109] Optionally, the terminal obtains multiple location points based on the location information of the target sensor, and sets numbers for the multiple location points in a certain order, and the number i of each location point is 1, 2, ...N, where N represents the number of location points; the terminal sets a number for each target sensor based on the number of target sensors, and the number j of each target sensor is 1, 2, ...M, where M represents the number of location points.

[0110] The terminal uses the position point number i as the matrix row vector, and uses the target perception data of the target sensor corresponding to the number j as the matrix column vector.

[0111] Step 302: construct a target fault monitoring matrix corresponding to the sampling time based on the first matrix row vector and the first matrix column vector, and use the target fault monitoring matrix corresponding to each sampling time as the fault monitoring matrix.

[0112] The target fault monitoring matrix is ​​used to characterize the target status monitored by the corresponding target sensor at each location point.

[0113] Optionally, for each sampling moment, the terminal obtains the target state data of the target sensor corresponding to each position point at the sampling moment, takes the corresponding position point number i as the matrix row vector, takes the target perception data of the target sensor corresponding to number j as the matrix column vector, and constructs the target fault monitoring matrix X as follows:

[0114]

[0115] Among them, x i,j Represents the target perception data of sensor j at the i-th position.

[0116] Step 303, for each fault moment, determine the row vector of the second matrix according to the position information of the target sensor, and determine the column vector of the second matrix according to the target historical perception data corresponding to the target sensor and the fault moment.

[0117] The target fault time refers to the time when the target historical perception data of the corresponding target sensor is collected when the substation fault is detected during the non-target time period.

[0118] Optionally, a data table of target fault moments and target historical sensing data is stored in the terminal. A target fault moment is determined, the corresponding target historical sensing data is obtained, and then a target historical fault matrix is ​​constructed according to the location information of the target sensor and the target historical sensing data.

[0119] Optionally, based on the location information of the target sensor, multiple location points are obtained, and numbers are set for the multiple location points in a certain order, and the number i of each location point is 1, 2, ...N, where N represents the number of location points; based on the number of target sensors, numbers are set for each target sensor, and the number j of each target sensor is 1, 2, ...M, where M represents the number of location points.

[0120] The position point number i is used as the matrix row vector, and the target historical perception data of the target sensor corresponding to the number j is used as the matrix column vector.

[0121] Step 304 , construct a target historical fault matrix corresponding to the fault moment based on the row vectors of the second matrix and the column vectors of the second matrix, and use the target historical fault matrix corresponding to each fault moment as the historical fault matrix.

[0122] The target historical fault matrix is ​​used to characterize the fault status detected by the corresponding target sensor at each location point.

[0123] Optionally, for each fault moment, the target historical perception data collected by the target sensor corresponding to each position point at the fault moment is obtained, the corresponding position point number i is used as the matrix row vector, and the target historical perception data of the target sensor corresponding to number j is used as the matrix column vector, and the target historical fault matrix Y is constructed as follows:

[0124]

[0125] Among them, y i,j Represents the historical target perception data of sensor j at the i-th location.

[0126] The embodiment of the present application constructs a fault monitoring matrix for each sampling moment based on the position information of the target sensor and the target perception data corresponding to the target sensor and the sampling moment, and constructs a historical fault matrix for each fault moment based on the position information of the target sensor and the corresponding target historical perception data of the target sensor. Since the perception data and position information of the target sensor are represented together in the same matrix for fault judgment, the calculation method for fault judgment is simple, which facilitates subsequent fault location.

[0127] In the embodiments of this application, please refer to Figure 4 ,based on Figure 3 In the embodiment shown, the mutual information between the fault monitoring matrix and the historical fault matrix is ​​obtained in the above step 203, and a comparison result is obtained according to the mutual information, including steps 401, 402, 403 and 404:

[0128] Step 401 : for each sampling moment, determine the target fault moment corresponding to the sampling moment, and calculate the target mutual information between the target fault monitoring matrix corresponding to the sampling moment and the target historical fault matrix corresponding to the target fault moment.

[0129] Optionally, the sampling moment corresponds to one or more target fault moments.

[0130] Optionally, for each sampling moment, when there are multiple target fault moments corresponding to the sampling moment, the mutual information between the target fault monitoring matrix corresponding to the sampling moment and the target historical fault matrix corresponding to each target fault moment is calculated respectively to obtain multiple mutual information, and then the multiple target mutual information are summed or weightedly summed to obtain the final mutual information, which is the target mutual information.

[0131] Optionally, the target mutual information of the target fault monitoring matrix corresponding to the sampling time and the target historical fault matrix corresponding to the target fault time is calculated to obtain a mutual information sequence S = [MI1, MI2, ..., MI n ], where MI i Represents the target mutual information corresponding to the i-th sampling time point.

[0132] Step 402, calculating the mutual information difference of the target mutual information corresponding to two adjacent sampling moments in each sampling moment.

[0133] Optionally, the calculation formula of the mutual information difference is λ=MI i -MI i-1 Or λ=|MI i -MI i-1 |.

[0134] Step 403 : Compare each target mutual information with a first threshold to obtain a first comparison result, and compare each mutual information difference with a second threshold to obtain a second comparison result.

[0135] Optionally, the target mutual information corresponding to the latest sampling time point is compared with the first threshold. Then, the target mutual information corresponding to the previous sampling time point of the latest sampling time point is obtained, and the mutual information difference between the target mutual information corresponding to the latest sampling time point and the target mutual information corresponding to the previous sampling time point is calculated, and the mutual information difference is compared with the second threshold.

[0136] Step 404: taking each first comparison result and each second comparison result as a comparison result.

[0137] Optionally, when the target mutual information is greater than a first threshold and the mutual information difference is greater than a second threshold, the substation has a fault.

[0138] Optionally, when a substation failure is detected, an alarm signal is sent to a terminal device of a substation worker or maintenance personnel.

[0139] This embodiment obtains a first comparison result by comparing each target mutual information with a first threshold, and obtains a second comparison result by comparing each mutual information difference with a second threshold. Each first comparison result and each second comparison result are used as comparison results to determine whether there is a fault in the substation, thereby improving the accuracy of rapid fault judgment.

[0140] In the present application embodiment, based on Figure 1 Based on the embodiment shown, the target sensor is wirelessly deployed in the substation; there are multiple target sensors, and the multiple target sensors are distributed in the substation in a tree network topology structure or a multi-hop network topology structure.

[0141] Optionally, a wireless sensor network is constructed based on target sensors, power wireless private network base stations and local area network servers. Figure 5 As shown, the topology of the wireless sensor network is a tree network topology or a multi-hop network topology. In another optional implementation, a wireless sensor network is constructed based on target sensors, power wireless private network base stations and cloud servers.

[0142] Optionally, a wireless sensor network is constructed using wireless ad hoc networking technology.

[0143] Optionally, the target sensor transmits the collected target state data to the power wireless private network base station through the LIE-A wireless communication interface, and the power wireless private network base station uploads the received target perception data to the local area network server or the cloud server through the Internet. The terminal collects the target perception data from the local area network server or the cloud server. The terminal performs real-time abnormal state monitoring based on the collected target perception data to determine whether the substation has a fault.

[0144] This embodiment wirelessly deploys the target sensor in the substation, thereby overcoming the problem of inconvenience in target sensor deployment caused by connecting the target sensor in a wired communication manner, and achieving flexible deployment of the target sensor.

[0145] In the present application, please refer to Figure 6 ,based on Figure 1In the embodiment shown, the substation fault detection method further includes step 501 and step 502:

[0146] Step 501: If there is a fault in the substation, the target state data is input into the fault detection model to obtain the fault location and fault type of the substation.

[0147] Among them, the fault detection model is a convolutional neural network model.

[0148] Optionally, the target state data is input into the fault detection model, the abnormal data in the target perception data is located, and the corresponding target sensor is determined based on the abnormal data, and the position information of the monitoring object of the target sensor is determined based on the target sensor, that is, the fault location is obtained. The fault type is determined based on the type of the target perception data of the target sensor, such as temperature, humidity, sound, displacement, vibration acceleration, and thermal image data.

[0149] Optionally, the target sensor and the location information of the monitored object of the target sensor are in one-to-one correspondence. Optionally, the terminal stores a mapping relationship table of the target sensor and the location information of the monitored object of the target sensor. By means of a table lookup method, the location information of the monitored object of the target sensor is determined according to the target sensor.

[0150] Step 502: Match the fault location and fault type of the substation with the fault information in the fault knowledge base to obtain a fault solution corresponding to the fault location and fault type.

[0151] A large amount of fault information is pre-stored in the fault knowledge base, and each fault information includes fault description information, corresponding solutions and solution effect score values, wherein the fault information includes the fault location and fault type of a certain fault, etc. The solution effect score value reflects the quality of the solution effect, and the higher the score value, the better the solution effect.

[0152] Optionally, the degree of match between the fault location and fault type of the substation and the fault description information in the fault knowledge base is calculated, and the fault solution corresponding to the fault description information with a high degree of match is determined as the fault solution corresponding to the fault location and fault type.

[0153] Optionally, the fault location and fault type of the substation are matched with the fault description information in the fault knowledge base to obtain multiple fault solutions corresponding to the fault description information, and the final solution is determined based on the score value of the solution effect.

[0154] This embodiment inputs the target state data into the fault detection model when there is a fault in the substation, obtains the fault location and fault type of the substation, improves the accuracy of fault location, and takes into account the real-time and accuracy of fault diagnosis. In addition, since the fault location and fault type of the substation are matched with the fault information in the fault knowledge base, the fault location and fault type corresponding to the fault solution are obtained, which improves the efficiency of fault resolution.

[0155] In the present application, please refer to Figure 7 ,based on Figure 6 In the embodiment shown, the training process of the fault detection model includes:

[0156] Step 601, obtaining training data.

[0157] The training data includes location information of the target sensor, sensor perception data corresponding to multiple historical fault moments, fault location information corresponding to each historical fault moment, and fault type corresponding to each historical fault moment.

[0158] Optionally, when the terminal detects a fault in the substation, it will save the sensor sensing data, fault location information and fault type monitored by the corresponding target sensor at that moment as fault status data to the corresponding storage path. The storage path may be located in the terminal or in the server.

[0159] Optionally, the fault detection model can be trained by the terminal itself; optionally, in order to save computing resources of the terminal, the fault detection model can also be trained by the server and sent to the terminal after the training is completed.

[0160] Optionally, the terminal or server may call the stored fault status data when training the fault detection model.

[0161] Optionally, in order to expand the number of data samples of the fault state data, the number of fault state data samples can be expanded by using sample expansion methods such as random oversampling algorithm, SMOTE algorithm, ADASYN algorithm, generative adversarial network method, etc.

[0162] Step 602: Iteratively train the initial fault detection model using training data to obtain a fault detection model.

[0163] Optionally, the location information of the target sensor and the sensor perception data corresponding to multiple historical fault moments are used as the input of the initial fault detection model, and the fault location information corresponding to each historical fault moment and the fault type corresponding to each historical fault moment are used as the initial fault detection model, and the initial fault detection model is iteratively trained using the back propagation algorithm.

[0164] Optionally, the terminal can also divide the training data set into a training set and a test set, obtain a fault detection model by training the training set, and verify the model effect of the fault detection model by the test set. If the fault detection model passes the verification, the training is determined to be completed and the final fault detection model is obtained.

[0165] This embodiment obtains training data and uses the training data to iteratively train the initial fault detection model to obtain a fault detection model, thereby improving the reliability of the fault detection model.

[0166] In the present application embodiment, Figure 8 As shown, a substation fault detection method is provided, the method comprising the following steps:

[0167] Step 701: wirelessly deploy multiple target sensors in a substation. The multiple target sensors are distributed in the substation in a tree network topology structure or a multi-hop network topology structure.

[0168] Step 702, obtaining location information of a target sensor and target sensing data obtained by monitoring the substation at at least one sampling time within a target time period.

[0169] Step 703, for each sampling moment, determine the matrix row vector according to the position information of the target sensor, and determine the matrix column vector according to the target sensor and the target perception data corresponding to the sampling moment.

[0170] Step 704: construct a target fault monitoring matrix corresponding to the sampling time based on the row vector and the column vector, and use the target fault monitoring matrix corresponding to each sampling time as the fault monitoring matrix.

[0171] Step 705: Acquire fault status data of the target sensor corresponding to the historical time period, where the fault status data is obtained by monitoring the target sensor when a substation fails.

[0172] Among them, the fault status data includes target historical perception data corresponding to at least one fault moment in a historical time period.

[0173] Step 706: construct a historical fault matrix based on the fault status data.

[0174] Among them, the historical fault matrix includes a target historical fault matrix corresponding to each fault moment. For each fault moment, the second matrix row vector is determined according to the position information of the target sensor, and the second matrix column vector is determined according to the target historical perception data corresponding to the target sensor and the fault moment; based on the second matrix row vector and the second matrix column vector, the target historical fault matrix corresponding to the fault moment is constructed, and the target historical fault matrix corresponding to each fault moment is used as the historical fault matrix.

[0175] Step 707 , for each sampling moment, determine the target fault moment corresponding to the sampling moment, and calculate the target mutual information between the target fault monitoring matrix corresponding to the sampling moment and the target historical fault matrix corresponding to the target fault moment.

[0176] Step 708, calculating the mutual information difference of the target mutual information corresponding to two adjacent sampling moments in each sampling moment.

[0177] Step 709 , compare each target mutual information with the first threshold to obtain a first comparison result, and compare each mutual information difference with the second threshold to obtain a second comparison result.

[0178] Step 710: taking each first comparison result and each second comparison result as a comparison result, and determining whether there is a fault in the substation according to the comparison result.

[0179] Step 711: If there is a fault in the substation, the target state data is input into the fault detection model to obtain the fault location and fault type of the substation.

[0180] Among them, the fault detection model is a convolutional neural network model.

[0181] Among them, the training process of the fault detection model includes: obtaining training data, the training data includes the location information of the target sensor, the sensor perception data corresponding to multiple historical fault moments, the fault location information corresponding to each historical fault moment, and the fault type corresponding to each historical fault moment; using the training data to iteratively train the initial fault detection model to obtain the fault detection model.

[0182] Step 712: Match the fault location and fault type of the substation with the fault information in the fault knowledge base to obtain a fault solution corresponding to the fault location and fault type.

[0183] This embodiment realizes flexible deployment of target sensors by wirelessly deploying multiple target sensors in the substation; by calculating the target mutual information of the target fault monitoring matrix corresponding to the sampling moment and the target historical fault matrix corresponding to the target fault moment and the mutual information difference of the target mutual information corresponding to two adjacent sampling moments in each sampling moment, and comparing each target mutual information with the first threshold to obtain the first comparison result and comparing each mutual information difference with the second threshold to obtain the second comparison result, it is determined whether the substation has a fault according to the first comparison result and the second comparison result, thereby reducing the difficulty of fault judgment and improving the speed of fault judgment. In addition, since the target state data is input into the fault detection model if there is a fault in the substation, the fault location and fault type of the substation are obtained, the accuracy of fault location is improved, while taking into account the real-time and accuracy of fault diagnosis.

[0184] It should be understood that although Figure 1-4 as well as Figure 6-8 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1-4 as well as Figure 6-8 At least part of the steps may include multiple steps or multiple stages. These steps or stages are not necessarily performed at the same time, but can be performed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed in turn or alternately with other steps or at least part of the steps or stages in other steps.

[0185] In the embodiments of the present application, Fig. 9 As shown, a substation fault detection device is provided, including: an acquisition module and a judgment module, wherein:

[0186] An acquisition module is used to acquire target state data obtained by a target sensor monitoring a substation within a target time period; acquire fault state data corresponding to the target sensor and a historical time period, wherein the fault state data is obtained by the target sensor monitoring when the substation fails;

[0187] The judgment module is used to compare the target state data with the fault state data for similarity to obtain a comparison result, and determine whether the substation has a fault according to the comparison result.

[0188] In one embodiment, the determination module is specifically used for:

[0189] Construct a fault monitoring matrix based on the target status data;

[0190] According to the fault status data, a historical fault matrix is ​​constructed;

[0191] The mutual information between the fault monitoring matrix and the historical fault matrix is ​​obtained, and a comparison result is obtained according to the mutual information.

[0192] In one embodiment, the target state data includes location information of the target sensor and target perception data corresponding to at least one sampling time within the target time period, and the judgment module is further used to:

[0193] For each sampling moment, the row vector of the matrix is ​​determined according to the position information of the target sensor, and the column vector of the matrix is ​​determined according to the target perception data corresponding to the target sensor and the sampling moment;

[0194] Based on the matrix row vectors and the matrix column vectors, a target fault monitoring matrix corresponding to the sampling time is constructed, and the target fault monitoring matrix corresponding to each sampling time is used as the fault monitoring matrix.

[0195] In one embodiment, the historical fault matrix includes a target historical fault matrix corresponding to at least one fault moment in the historical time period, and the judgment module is further used to:

[0196] For each sampling moment, determine the target fault moment corresponding to the sampling moment, and calculate the target mutual information between the target fault monitoring matrix corresponding to the sampling moment and the target historical fault matrix corresponding to the target fault moment;

[0197] Calculate the mutual information difference of the target mutual information corresponding to two adjacent sampling moments in each sampling moment;

[0198] Compare each target mutual information with a first threshold to obtain a first comparison result, and compare each mutual information difference with a second threshold to obtain a second comparison result;

[0199] Each first comparison result and each second comparison result are taken as a comparison result.

[0200] In one embodiment, the target sensor is wirelessly deployed in the substation; there are multiple target sensors, and the multiple target sensors are distributed in the substation in a tree network topology structure or a multi-hop network topology structure.

[0201] In one embodiment, Fig.10 As shown, the device also includes an output module and a solution module, specifically:

[0202] An output module is used to input the target state data into the fault detection model when there is a fault in the substation, so as to obtain the fault location and fault type of the substation;

[0203] The solution module is used to match the fault location and fault type of the substation with the fault information in the fault knowledge base to obtain a fault solution corresponding to the fault location and fault type.

[0204] In one embodiment, the training process of the fault detection model includes:

[0205] Acquire training data, where the training data includes location information of a target sensor, sensor perception data corresponding to multiple historical fault moments, fault location information corresponding to each historical fault moment, and fault type corresponding to each historical fault moment;

[0206] The initial fault detection model is iteratively trained using the training data to obtain a fault detection model.

[0207] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Fig.11 As shown. The computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a substation fault detection method is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covered on the display screen, or a key, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.

[0208] Those skilled in the art will understand that Fig.11 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0209] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0210] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0211] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0212] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0213] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A substation fault detection method, characterized in that: The method comprises: Obtain target status data obtained by the target sensor monitoring the substation within the target time period; Acquire fault status data of the target sensor corresponding to a historical time period, wherein the fault status data is obtained by monitoring the target sensor when the substation fails; Comparing the target state data and the fault state data for similarity to obtain a comparison result, and determining whether the substation has a fault according to the comparison result; The similarity comparison between the target state data and the fault state data to obtain a comparison result includes: constructing a fault monitoring matrix according to the target state data; Constructing a historical fault matrix according to the fault status data; Mutual information between the fault monitoring matrix and the historical fault matrix is ​​obtained, and the comparison result is obtained according to the mutual information.

2. The method according to claim 1, characterized in that The target state data includes the location information of the target sensor and the target perception data corresponding to at least one sampling time within the target time period; The step of constructing a fault monitoring matrix according to the target state data includes: For each of the sampling moments, determining a matrix row vector according to the position information of the target sensor, and determining a matrix column vector according to the target perception data corresponding to the target sensor and the sampling moment; Based on the matrix row vectors and the matrix column vectors, a target fault monitoring matrix corresponding to the sampling moments is constructed, and the target fault monitoring matrix corresponding to each of the sampling moments is used as the fault monitoring matrix.

3. The method according to claim 2, characterized in that The historical fault matrix includes a target historical fault matrix corresponding to at least one fault moment in the historical time period; the acquiring mutual information between the fault monitoring matrix and the historical fault matrix, and acquiring the comparison result according to the mutual information, includes: For each of the sampling moments, determining a target fault moment corresponding to the sampling moment, and calculating target mutual information between the target fault monitoring matrix corresponding to the sampling moment and the target historical fault matrix corresponding to the target fault moment; Calculate the mutual information difference of the target mutual information corresponding to two adjacent sampling moments in each of the sampling moments; Compare each target mutual information with a first threshold to obtain a first comparison result, and compare each mutual information difference with a second threshold to obtain a second comparison result; Each of the first comparison results and each of the second comparison results are used as the comparison result.

4. The method according to claim 1, characterized in that: The target sensor is wirelessly deployed in the substation; there are multiple target sensors, and the multiple target sensors are distributed in the substation in a tree network topology structure or a multi-hop network topology structure.

5. The method according to claim 1, characterized in that The method further comprises: If there is a fault in the substation, the target state data is input into a fault detection model to obtain the fault location and fault type of the substation; The fault location and the fault type of the substation are matched with the fault information in a fault knowledge base to obtain a fault solution corresponding to the fault location and the fault type.

6. The method according to claim 5, characterized in that The training process of the fault detection model includes: Acquire training data, the training data including location information of the target sensor, sensor perception data corresponding to multiple historical fault moments, fault location information corresponding to each of the historical fault moments, and fault type corresponding to each of the historical fault moments; The initial fault detection model is iteratively trained using the training data to obtain the fault detection model.

7. A substation fault detection device, characterized in that: The device comprises: An acquisition module is used to acquire target state data obtained by a target sensor monitoring a substation within a target time period; acquire fault state data corresponding to the target sensor and a historical time period, wherein the fault state data is obtained by the target sensor monitoring when the substation fails; A judgment module, used for comparing the target state data with the fault state data for similarity, obtaining a comparison result, and determining whether the substation has a fault according to the comparison result; The judgment module is specifically used for: constructing a fault monitoring matrix according to the target state data; Constructing a historical fault matrix according to the fault status data; Mutual information between the fault monitoring matrix and the historical fault matrix is ​​obtained, and the comparison result is obtained according to the mutual information.

8. The device according to claim 7, characterized in that The target state data includes the location information of the target sensor and the target perception data corresponding to at least one sampling time in the target time period; the judgment module is also used for: For each of the sampling moments, determining a matrix row vector according to the position information of the target sensor, and determining a matrix column vector according to the target perception data corresponding to the target sensor and the sampling moment; Based on the matrix row vectors and the matrix column vectors, a target fault monitoring matrix corresponding to the sampling moments is constructed, and the target fault monitoring matrix corresponding to each of the sampling moments is used as the fault monitoring matrix.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Citation Information

Patent Citations

  • Massive monitoring data based substation equipment fault diagnosis method

    CN104317778A

  • Semi-supervised clustering method and semi-supervised clustering system based on nonnegative matrix factorization

    CN105023026A

  • System for monitoring and evaluating safety of operating personnel in hazardous environment

    CN105184433A

  • Fault classification diagnosis method based on non-Gaussian similarity matching

    CN106444665A

  • Transformer substation fault diagnosis method based on internet of things and server

    CN107634988A

Cited By

  • Predicting electrical component failure

    US12803278B2