Method, system and device for identifying working status of in-station equipment based on information fusion

Abstract

The present application provides a method, system and device for identifying the working status of in-station equipment based on information fusion, including: obtaining basic image information of in-station equipment in the target area where the offshore booster station is located; preprocessing the basic image information to obtain first basic image information; performing equipment information fusion and decision fusion based on the first basic image information to obtain a fusion feature image of the in-station equipment; constructing an in-station equipment status detection network based on the fusion feature image of the in-station equipment to identify the operating status of the in-station equipment. The present application can effectively improve inspection efficiency, reduce inspection costs, and effectively identify the working status of in-station equipment. It not only provides an accurate assessment of the equipment status, but also has the flexibility to adapt to different equipment types and environments.

Description

Technical Field

[0001] The present invention belongs to the technical field of marine engineering monitoring, and relates to a method for identifying the working status of in-station equipment, and in particular to a method, system and device for identifying the working status of in-station equipment based on information fusion. Background Art

[0002] In modern industrial and commercial environments, the monitoring and identification of the operating status of equipment in offshore booster stations is crucial to the effectiveness of maintenance and operation. These station equipment can be main transformers, GIS, medium-voltage switchgear and other electrical equipment. Accurately understanding the working status of these equipment can replace manual inspection methods, improve inspection efficiency and reduce inspection costs.

[0003] However, current methods for identifying the working status of in-station equipment have a series of challenges and limitations. Traditional methods mainly rely on a single data source or sensor, which may lead to insufficient information or misleading results. In addition, different types of equipment may require different identification methods, making it difficult for existing technologies to adapt to the diverse in-station equipment environment.

[0004] Therefore, how to provide an innovative method and process to solve these challenges through information fusion to achieve more accurate and reliable identification of the working status of in-station equipment is an urgent problem to be solved. Summary of the invention

[0005] In view of the shortcomings of the prior art mentioned above, the purpose of the present application is to provide a method, system and device for identifying the working status of in-station equipment based on information fusion, which is used to solve the problem that in the operation and maintenance process of in-station equipment in the prior art, due to excessive reliance on a single data source or sensor, insufficient information or misleading results are caused, and different types of equipment require different identification methods, making it difficult to adapt to the diverse in-station equipment environment.

[0006] To achieve the above-mentioned purpose and other related purposes, in a first aspect, the present application provides a method for identifying the working status of in-station equipment based on information fusion, comprising the following steps: obtaining basic image information of in-station equipment in the target area where the offshore booster station is located; preprocessing the basic image information to obtain first basic image information; performing equipment information fusion and decision fusion based on the first basic image information to obtain a fused feature image of the in-station equipment; and constructing an in-station equipment status detection network based on the in-station equipment fused feature image to identify the operating status of the in-station equipment.

[0007] In an implementation of the first aspect, obtaining basic image information of in-station equipment in a target area where an offshore substation is located includes the following steps: obtaining video information of in-station equipment from a camera in the offshore substation; obtaining basic image information of a number of in-station equipment based on the in-station equipment video information; wherein the basic image information of the in-station equipment includes: any one or more combinations of main transformer operation information, GIS operation information, medium-voltage switchgear equipment information, circuit breaker opening and closing information, microcomputer detuning device operation information, voltage transformer operation information, and line protection measurement and control device operation information.

[0008] In an implementation of the first aspect, preprocessing the basic image information to obtain first basic image information includes the following steps: performing image screening, image enhancement, and image normalization processing on the basic image information to obtain first basic image information.

[0009] In an implementation of the first aspect, device information fusion and decision fusion are performed based on the first basic image information to obtain a fused feature image of the in-station device, including the following steps: information fusion is performed on the first basic image information with the in-station device data source and sensor information to obtain a first basic image feature vector; during the information fusion process, decision fusion is performed on the first basic image feature vector and the in-station device status based on a decision fusion rule to obtain a fused feature image of the in-station device; wherein the in-station device data source includes: any one or more combinations of images, sensor data, sound, temperature, and humidity.

[0010] In an implementation of the first aspect, information fusion of the first basic image information with the in-station device data source and sensor information includes the following steps: performing feature extraction based on the first basic image information, and extracting corresponding features from the in-station device data source and sensor data to obtain first basic image features; performing feature fusion based on the first basic image features to obtain a first basic image feature vector.

[0011] In an implementation of the first aspect, the equipment status in the station includes: any one or more combinations of the circuit breaker opening and closing status, the energy storage device operating status, the microcomputer detuning device operating status, the voltage transformer operating status, the line protection measurement and control device operating status, the charging protection measurement and control device operating status, the gas pressure device operating status, and the foreign object detection status; the decision fusion rule includes: the circuit breaker opening and closing status, setting its normal working status according to the actual use situation, and setting the circuit breaker closing to the normal state when the test is interrupted, and no alarm; otherwise, it is an abnormal operation and an alarm is processed; the energy storage device operating status, when the energy storage device indicator light is red, the energy storage The device works normally and no alarm is issued; otherwise, it is an abnormal operation and an alarm is processed; the operating status of the microcomputer detuning device includes: resonance, device failure, grounding, operation, and overvoltage; the operating status of the voltage transformer includes: operation, alarm, and maintenance; the operating status of the line protection measurement and control device includes: operation, alarm, tripping, closing, tripping, closing, charging, and PT disconnection; the operating status of the gas pressure device, when the gas pressure drops to the low alarm pressure threshold, a yellow alarm is issued; when the gas pressure drops to the tripping pressure value, a red alarm is issued; the foreign object detection status, when no foreign object is detected, the environment is normal; when a foreign object is detected, a red alarm is issued.

[0012] In an implementation of the first aspect, constructing an in-station equipment status detection network based on the in-station equipment fusion feature image to identify the operating status of the in-station equipment includes the following steps: inputting the in-station equipment fusion feature image into the target detection network to obtain the in-station equipment operating status characteristics; analyzing the in-station equipment operating status feature distribution based on the in-station equipment operating status characteristics to identify the operating status of the in-station equipment.

[0013] In an implementation of the first aspect, analyzing the distribution of in-station equipment operating status characteristics based on the in-station equipment operating status characteristics to identify the operating status of the in-station equipment includes the following steps: inputting the in-station equipment fusion feature image into a target detection network for feature fusion to obtain high-level semantic information; classifying and regressing the high-level semantic information through a multi-scale decoupling detection head to detect the target; determining channel information and in-station equipment location information by splicing; matching the channel information with the in-station equipment location information to identify the operating status of the in-station equipment.

[0014] In an implementation of the first aspect, a splicing method is used to determine the channel information and the location information of the equipment in the station, including the following steps: decomposing the global pooling operation into a one-to-one dimensional encoding operation; encoding each channel along the horizontal coordinate and the vertical coordinate respectively to obtain a coding feature; the coding feature includes: two spatial direction aggregation features are aggregated to obtain a direction perception feature map, thereby extracting the target location information.

[0015] In the second aspect, the present application provides an in-station equipment working status identification system based on information fusion, including: an acquisition module, used to acquire basic image information of in-station equipment in the target area where the offshore booster station is located; a data processing module, used to pre-process the basic image information to obtain first basic image information; an information fusion module, used to perform equipment information fusion and decision fusion based on the first basic image information to obtain a fused feature image of the in-station equipment; a network construction and detection module, used to construct an in-station equipment status detection network based on the in-station equipment fused feature image to identify the operating status of the in-station equipment.

[0016] In the last aspect, the present application provides an in-station device working state identification device based on information fusion, comprising: a processor and a memory. The memory is used to store a computer program; the processor is connected to the memory and is used to execute the computer program stored in the memory, so that the in-station device working state identification device based on information fusion executes the in-station device working state identification method based on information fusion.

[0017] As described above, the method, system and device for identifying the working status of in-station equipment based on information fusion of the present invention have the following beneficial effects:

[0018] (1) This application provides a method for identifying the working status of in-station equipment based on information fusion, which can improve inspection efficiency and reduce inspection costs.

[0019] (2) This application can effectively identify the working status of the equipment in the station, not only providing an accurate assessment of the equipment status, but also having the flexibility to adapt to different equipment types and environments. Through the application of information fusion technology, this application has a broad application prospect and can be applied in various fields, including but not limited to the energy field, the power industry, and the inspection and maintenance industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 Shown is a flow chart of an embodiment of the method for identifying the working status of in-station equipment based on information fusion described in the present invention.

[0021] Figure 2Shown is a schematic diagram of the signal transmission flow of an embodiment of the method for identifying the working status of in-station equipment based on information fusion according to the present invention.

[0022] Figure 3 Shown is a flow chart of S11 in the method for identifying the working status of in-station equipment based on information fusion described in the present invention.

[0023] Figure 4 Shown is a flow chart of S13 in the method for identifying the working status of in-station equipment based on information fusion described in the present invention.

[0024] Figure 5 Shown is a flow chart of S14 in the method for identifying the working status of in-station equipment based on information fusion described in the present invention.

[0025] Figure 6 Shown is a structural schematic diagram of a target detection algorithm based on the Yolox-s algorithm in an embodiment of the method for identifying the working status of in-station equipment based on information fusion described in the present invention.

[0026] Figure 7 Shown is a schematic diagram of the principle structure of an embodiment of the in-station equipment working status identification system based on information fusion described in the present invention.

[0027] Figure 8 Shown is a schematic diagram of the principle structure of an apparatus for identifying the working status of in-station equipment based on information fusion in one embodiment of the present invention.

[0028] Description of Reference Numerals

[0029] 71 Get Module

[0030] 72 Data processing module

[0031] 73 Information Fusion Module

[0032] 74 Network construction and detection module

[0033] 81 Processor

[0034] 82 Memory DETAILED DESCRIPTION

[0035] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.

[0036] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and thus the drawings only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0037] The following will describe in detail the method for identifying the working status of in-station equipment based on information fusion provided in the embodiments of the present application in conjunction with the drawings in the embodiments of the present application.

[0038] See also Figure 1 and Figure 2 , respectively showing a flow chart of an in-station device working state identification method based on information fusion in one embodiment of the present invention and a signal transmission flow chart of an in-station device working state identification method based on information fusion in one embodiment of the present invention. Figure 1 and Figure 2 As shown, this embodiment provides a method for identifying the working status of in-station equipment based on information fusion.

[0039] The method for identifying the working status of in-station equipment based on information fusion specifically comprises the following steps:

[0040] S11, obtain basic image information of the equipment in the target area where the offshore booster station is located. Figure 3 , which is a flow chart of S11 in the method for identifying the working status of in-station equipment based on information fusion according to the present invention. Figure 3 As shown, the S11 comprises the following steps:

[0041] S111, obtain video information of equipment in the station through the camera in the offshore booster station.

[0042] In this embodiment, a camera is installed in the target area of ​​the offshore booster station, and a network camera is used to capture video data of all equipment in the offshore booster station.

[0043] Specifically, in the target area, use cameras or other image acquisition devices to monitor the equipment in the offshore booster station in real time to ensure that the camera can comprehensively, accurately and clearly record the operating status and surrounding environment of the equipment in the station. Then, debug the camera to ensure that it can stably transmit video information. Through the camera, continuously obtain video information of the equipment in the station. This can be done automatically or manually to ensure that the video information of the equipment can be captured in all directions and from multiple angles. The collected video information is stored on a high-performance storage device to ensure the integrity of the video and the convenience of subsequent processing.

[0044] S112: Obtain basic image information of several in-station devices based on the in-station device video information.

[0045] In this embodiment, the basic image information of the equipment in the station includes but is not limited to: main transformer operation information, GIS operation information, medium voltage switch cabinet equipment information, circuit breaker opening and closing information, microcomputer detuning device operation information, voltage transformer operation information, line protection measurement and control device operation information, etc.

[0046] Specifically, based on the acquired video information, image processing software or algorithms can be used to capture frames of the stored video at certain time intervals to generate a series of static images; that is, basic image information of the equipment in the station can be extracted through image processing and recognition technology. This includes the appearance, features, color, etc. of the equipment. This process can capture the details and instantaneous status of the equipment, providing a rich data source for subsequent analysis.

[0047] It should be noted that the image collection and processing system should be regularly maintained and updated to ensure system stability and data accuracy. At the same time, the system functions and performance should be continuously optimized according to actual needs and technological development.

[0048] S12: pre-process the basic image information to obtain first basic image information.

[0049] In this embodiment, the collected basic image information is preprocessed, including steps such as image screening, image enhancement, and image normalization, to obtain processed basic image information, thereby obtaining first basic image information.

[0050] Specifically, to screen images, we need to screen these basic image information according to some specific standards and user needs, and retain those images that meet the requirements. For example, we can screen according to factors such as image resolution, color, and content. Then, the images left after screening are enhanced by contrast enhancement, brightness adjustment, color balance, etc., so that the visual effect of the basic image information is significantly enhanced, so that it can be better analyzed and processed. Then, the pixel values ​​of the image after image enhancement are standardized so that their values ​​are all within a certain range, such as [0, 1] [0, 255], which can ensure that the subsequent processing steps are more stable and reliable.

[0051] It should be noted that the technology used in the preprocessing operation process can be implemented through corresponding processing methods and is not limited to the process of the embodiment described in this application.

[0052] S13, performing device information fusion and decision fusion based on the first basic image information to obtain a fusion feature image of the devices in the station. Figure 4, which is a flow chart of S13 in the method for identifying the working status of in-station equipment based on information fusion according to the present invention. Figure 4 As shown, the S13 comprises the following steps:

[0053] S131, fusing the first basic image information with the in-station device data source and sensor information to obtain a first basic image feature vector, wherein the in-station device data source includes but is not limited to: image, sensor data, sound, temperature, humidity, etc.

[0054] Information fusion is the process of integrating information from multiple in-station devices, data sources or sensors to produce more comprehensive and accurate identification results of the working status of in-station equipment.

[0055] In this application, we use a variety of different types of data sources, including but not limited to images, sensor data, sound, temperature, humidity, etc. These data sources capture multiple aspects of the working status of the equipment in the station and provide a more comprehensive information basis. Ensure data collection and time synchronization between various data sources to ensure data consistency and accuracy.

[0056] In this embodiment, feature extraction is performed based on the first basic image information, and corresponding features are extracted from the in-station device data source and sensor data to obtain first basic image features; feature fusion is performed based on the first basic image features to obtain a first basic image feature vector.

[0057] Specifically, first, feature extraction is performed on the first basic image information obtained from the video of the device in the station. This usually involves image processing technology, such as edge detection, corner detection, texture analysis, etc., which is used to identify and extract important features in the image. At the same time, corresponding features are extracted from the data source and sensor data of the device in the station. These features may include the working parameters, operating status, environmental parameters, etc. of the device, through which the performance and status of the device can be further understood.

[0058] Then, the features extracted from the image are combined with the features extracted from the device data source and sensor data to form the first basic image feature vector. The purpose of this step is to combine the visual information in the image with the data information of the device to provide a more comprehensive feature description.

[0059] Next, the first basic image feature vector obtained by fusion is further processed and analyzed, such as dimension reduction, normalization, classification or clustering, so as to better understand and utilize these features.

[0060] Finally, according to the processing and analysis, a first basic image feature vector is obtained.

[0061] From the above, it can be seen that in the process of information fusion, the present application adopts the decision fusion method.

[0062] It should be noted that in the actual application process, according to the output results and user feedback, the algorithms and methods of feature extraction, fusion and processing are continuously optimized to improve the accuracy and efficiency of the entire system. At the same time, it can be seen that the above implementation process is a general framework, and the specific steps and details may be adjusted according to the actual application scenarios and needs. In addition, the technologies involving image processing and data analysis need to select appropriate methods and technologies for implementation according to specific circumstances.

[0063] S132, in the information fusion process, performing decision fusion on the first basic image feature vector and the in-station device status based on a decision fusion rule to obtain an in-station device fusion feature image.

[0064] In this embodiment, since the information fusion in step S131 adopts decision fusion, specific decision rules need to be formulated for different data sources.

[0065] Specifically, the equipment status in the station includes but is not limited to: circuit breaker opening and closing status, energy storage device operation status, microcomputer detuning device operation status, voltage transformer operation status, line protection measurement and control device operation status, charging protection measurement and control device operation status, gas pressure device operation status, foreign object detection status and other equipment status. The decision fusion rules for these equipment statuses are as follows:

[0066] (1) Decision-making fusion rules for circuit breaker opening and closing states:

[0067] Set its normal working state according to the actual usage. If the test is interrupted, set the circuit breaker to close as the normal state. When the red light is on, it is in normal working state and no alarm is given. Otherwise, when the green light is on, it is considered as abnormal working and an alarm is given.

[0068] (2) Decision fusion rules for the operating status of energy storage devices:

[0069] When the indicator light of the energy storage device is red, the energy storage device is working normally and no alarm occurs; otherwise, when the indicator light is off, it is working abnormally and a red alarm occurs.

[0070] (3) Decision-making fusion rules for the operation status of the microcomputer detuning device:

[0071] There are five working states on the control panel of the microcomputer detuning device, namely: resonance, device failure, grounding, operation, and overvoltage.

[0072] (4) Decision fusion rules for voltage transformer operating status:

[0073] There are five working states on the control panel: operation, alarm and maintenance.

[0074] (5) Decision-making fusion rules for the operating status of line protection measurement and control devices:

[0075] There are eight working states on the control panel of the line protection measurement and control device, namely: operation, alarm, tripping, closing, tripping position, closing position, charging, and PT line break; the actual operating status of the device can be obtained through its indicator light display.

[0076] (6) Decision-making fusion rules for the operating status of the charging protection measurement and control device:

[0077] There are eight working states on the control panel of the charging protection measurement and control device, namely: operation, alarm, tripping, closing, tripping position, closing position, charging, and PT line break; the actual operating status of the device can be obtained through its indicator light display.

[0078] (7) Decision fusion rules for the operating status of gas pressure devices:

[0079] When the gas pressure drops to the low alarm pressure threshold, a yellow alarm is issued; when the gas pressure drops to the trip pressure value, a red alarm is issued.

[0080] Specifically, the SF6 gas pressure gauge is used as an example for explanation. For the SF6 gas pressure gauge, the normal range of gas pressure is set based on the pointer angle. When the gas pressure is at the standard value, there is no feedback information; when the gas pressure drops to the specified low alarm pressure value, a yellow alarm signal is issued; when the gas pressure drops to the trip (lockout) pressure value, a red alarm signal is issued. For the ammeter, the current value is obtained through the pointer angle and the normal range of current is set. If the current value is abnormal, an alarm is issued. For the oil level gauge, the oil volume percentage is obtained through the pointer angle and the normal range of oil volume is set. If the oil volume exceeds the normal range, a red alarm is issued.

[0081] (8) Decision fusion rules for foreign body detection status:

[0082] For foreign object detection, when no foreign object is detected, the environment is normal; when a foreign object is detected, a red alarm is issued.

[0083] S14, constructing an in-station equipment status detection network based on the in-station equipment fusion feature image to identify the operating status of the in-station equipment. Figure 5 and Figure 6 , showing a flow chart of S14 in the method for identifying the working state of an in-station device based on information fusion according to the present invention and a structural chart of a target detection algorithm based on the Yolox-s algorithm in an embodiment of the method for identifying the working state of an in-station device based on information fusion according to the present invention. Figure 5 and Figure 6 As shown, the S14 comprises the following steps:

[0084] S141, inputting the fused feature image of the in-station equipment into the target detection network to obtain the operating status features of the in-station equipment.

[0085] In this embodiment, the Yolox algorithm is preferably used for processing. After the input image size is scaled to 640×640, it is sent to Backbone to extract features; then the 8-fold, 16-fold, and 32-fold sampling information is fused in the feature fusion network.

[0086] Specifically, first, the size of the fusion feature image of the in-station equipment is adjusted to 640×640 pixels. The purpose of this step is to unify the size of the image for subsequent feature extraction and network processing. Then the resized image is input into a pre-trained Backbone network, such as ResNet, VGG, etc. The main function of the Backbone network is to extract basic features from the image. Through a series of operations such as convolution and pooling, the network extracts useful feature representations from the image. In the Backbone network, high-level features are extracted from the input image through a series of convolution layers and pooling layers. These features contain information such as edges, textures, shapes, etc. in the image, which can reflect the basic structure and content of the image. Further, in the feature fusion network, features sampled at different times are fused. Specifically, features sampled at 8 times, 16 times, and 32 times are merged to obtain feature representations at different scales. This multi-scale feature fusion can capture the information of the image at different scales and enhance the representation ability of the features.

[0087] For example, in the Yolox algorithm, in the feature utilization part, YoloX extracts multiple feature layers for target detection, and extracts three feature layers in total. That is, when the input is (640, 640, 3), it first passes through the feature extraction structure Focus (320, 320, 12), CBA (Classification base of Association, fusion convolution and batch normalization) / DWConv (Depthwise Separable Convolution, depth separable convolution) (160, 160, 128) and CSPLayer (160, 160, 128), performs slicing operation Slice, and concatenates operations Concat, extracting a total of three feature layers. The three feature layers are located in different positions of the main part darknet (using the darknet structure in deep learning), respectively in the middle layer, the lower middle layer, and the bottom layer. The shapes of the three feature layers are Dark3 = (20, 20, 1024), Dark4 = (40, 40, 512), and Dark5 = (80, 80, 256). Next, after obtaining three valid feature layers, we use these three valid feature layers to construct the FPN (Feature Pyramid Networks) layer, and perform downsampling Upsamble, concatenation operation Concat and other processes on the three valid feature layers.

[0088] Then, a suitable feature fusion method is selected, such as simple splicing, weighted fusion, or using a specific fusion structure such as convolution, pooling, etc. According to the application requirements of the offshore booster station, this implementation adopts the splicing method.

[0089] S142: Analyze the distribution of the operating status characteristics of the in-station equipment based on the operating status characteristics of the in-station equipment to identify the operating status of the in-station equipment.

[0090] This embodiment preferably adopts the Yolox-s algorithm.

[0091] Yolox-s is the lightest algorithm with a very concise network structure, which only contains 6 Backbone layers and 3 Decouple Head layers. The Backbone uses CSPDarknet53, and the Decouple Head uses two structures, SPP (Spatial Pyramid Pooling) and PAFPN, the head structures of YOLOv3. SPP is a spatial pyramid pooling structure, which is used to perform the maximum pooling of feature maps of different scales in order to extract more comprehensive feature information; PAFPN is a pyramid-shaped feature aggregation network, which is used to aggregate feature maps of different scales for detection.

[0092] The Yolox-s algorithm used in this application can realize intelligent detection and accurate identification of in-station equipment, and can be used to improve the accuracy and efficiency of equipment status monitoring. The Yolox-s algorithm is optimized in terms of network depth to ensure excellent performance under limited resources.

[0093] In this embodiment, the fused feature image of the in-station equipment is input into the target detection network for feature fusion to obtain high-level semantic information; the high-level semantic information is classified and regressed through a multi-scale decoupling detection head to achieve detection of the target; the channel information and the in-station equipment location information are determined by splicing; the channel information is matched with the in-station equipment location information to identify the operating status of the in-station equipment.

[0094] Specifically, the fused image obtained in step S141 is input into the Yolox-s algorithm, and the basic features are fully utilized to fuse features of different degrees. A residual branch is introduced to inject feature maps of different levels, and the output from the backbone network, the GSN output of the same level, and the output after upsampling of the next level GSN are fused to generate semantic information with basic features. At the same time, the splicing method is used in this embodiment to replace the original addition, and more feature information, especially channel information, is retained. Among them, the location information is crucial for the positioning of in-station equipment. In order to help the feature extraction network accurately extract the target location information, the global pooling operation is decomposed into a pair of one-dimensional encoding operations, and each channel is encoded along the horizontal and vertical coordinates respectively. After aggregating the encoded features, a pair of direction-aware feature maps are generated, which is conducive to the backbone network to extract more target location information.

[0095] For example, three effective feature layers are obtained in step S141, and then the three effective feature layers are used to construct the FPN layer. The specific construction method is as follows:

[0096] The Dark3 = (20, 20, 1024) feature layer is subjected to a 1×1 convolution adjustment channel to obtain P3. P3 is upsampled by UmSampling2d and combined with the Dark4 = (40, 40, 512) feature layer. Then, CSPLayer and CBA are used to extract features and obtain P3_upsample downsampling. At this time, the feature layer obtained is (40, 40, 512).

[0097] The feature layer P3_upsample = (40,40,512) is subjected to a 1×1 convolution to adjust the channel to obtain P4. P4 is upsampled by UmSampling2d and combined with the feature layer Dark5 = (80,80,256). Then, CSPLayer and CBA / DWConv are used to extract feature P5_out. At this time, the feature layer obtained is (80,80,256).

[0098] The feature layer of P5_out = (80, 80, 256) uses CSPLayer and CBA / DWConv to extract features, stacks with P4, and then uses CSPLayer to extract features of P4_out. At this time, the feature layer obtained is (40, 40, 512).

[0099] The feature layer of P4_out = (40, 40, 512) uses CBA / DWConv for feature extraction, stacked with P3, and then uses CSPLayer to extract P3_out. At this time, the feature layer obtained is (20, 20, 1024).

[0100] The above three enhanced features are obtained. The shapes of these three enhanced features are (20,20,1024), (40,40,512), and (80,80,256). The feature layers of these three shapes are respectively passed into the Decoupled Head, and CBA and two CBA / DWConv are used to extract features to obtain the prediction results:

[0101] Reg prediction result, the number of channels of convolution is 4, and the final result is (20,20,4). The 4 can be divided into two 2s. The first 2 is the offset of the center point of the prediction box compared to the feature point, and the second 2 is the parameter of the width and height of the prediction box compared to the logarithmic exponent.

[0102] Obj prediction result, the number of channels of the convolution is 1, and the final result is (20,20,1), which represents the probability that each feature point prediction box contains an object;

[0103] Cls prediction result, the number of channels of the convolution is num_classes, and the final result is (20,20,num_classes), which represents the probability that each feature point corresponds to a certain type of object. The prediction value in the last dimension num_classes represents the probability of belonging to each class.

[0104] In the above prediction results, the Reg prediction result is used to determine the regression parameters of each feature point. After the regression parameters are adjusted, the prediction box can be obtained; the Obj prediction result is used to determine whether each feature point contains an object; the Cls prediction result is used to determine the type of object contained in each feature point.

[0105] In the above encoding operation, for a given input feature: X = [x1, x2, ... x c ]∈R C×H×W , the output formula of the cth channel with height H and width W is:

[0106]

[0107] in, and They are respectively features aggregated in two spatial directions; C represents the total number of channels; H represents the height of the image; W represents the width of the image; h represents the h-th row of pixels in the image, 1≤h≤H; w represents the w-th column of pixels in the image, 1≤w≤W; (h,i) represents the i-th pixel in the h-th row; (j,w) represents the j-th pixel in the w-th column.

[0108] Then, the generated pair of direction-aware feature maps are concatenated, denoted as f, and features are extracted using 1×1 convolution. The specific formula is as follows:

[0109] f=δ(F1([z h ,z w ]))

[0110] Among them, [z h ,z w ] represents the connection operation along the spatial dimension; F1 is a 1×1 regular convolution operation; δ is a nonlinear activation function; f∈R Cr×(H+W) , r is the scaling factor.

[0111] Next, f is divided into two eigenvectors along the spatial dimension, namely: f h and f w Among them, f h ∈R Cr×H , f w ∈R Cr×W . Use 1×1 convolution on f respectively. h and fw Transform so that f h and f w The number of channels of the input X is the same as that of the input X, and the attention weights in two directions are obtained and recorded as: h and g w , the formula is as follows:

[0112] g h =σ(F1(f h ))

[0113] g w =σ(F1(f w ))

[0114] Among them, is represented by the Sigmoid activation function.

[0115] Finally, the obtained weight is multiplied by the input feature to get the output result. The formula is as follows:

[0116]

[0117] Among them, y c (i,j) represents the output in text form, which contains the working status of the equipment in a certain station; x c (i,j) represents the eigenvalue in the input eigenvector; Represents the attention weight of the input in the h direction; Represents the attention weight of the input in the w direction.

[0118] Therefore, the final output content is the working status of a device within a certain station, and the output is mainly in the form of text, but not limited to text.

[0119] The method for identifying the working status of in-station equipment based on information fusion provided by this application can improve inspection efficiency and reduce inspection costs. This application can effectively identify the working status of in-station equipment, not only providing an accurate assessment of the equipment status, but also having the flexibility to adapt to different equipment types and environments. Through the application of information fusion technology, this application has broad application prospects and can be applied in various fields, including but not limited to the energy field, the power industry, and the inspection and maintenance industry.

[0120] The protection scope of the method for identifying the working status of in-station equipment based on information fusion described in the embodiment of the present application is not limited to the execution order of the steps listed in the present embodiment. All solutions implemented by adding, reducing or replacing steps in the prior art based on the principles of the present application are included in the protection scope of the present application.

[0121] This embodiment further provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the following is achieved: Figure 1The method for identifying the working status of in-station equipment based on information fusion.

[0122] At any possible level of technical detail combination, the present application may be a system, method and / or computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present application.

[0123] Computer readable storage medium can be a tangible device that can hold and store instructions used by an instruction execution device. Computer readable storage medium can be, for example, (but not limited to) an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination thereof. More specific examples (non-exhaustive list) of computer readable storage medium include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, for example, a punch card or a convex structure in a groove on which instructions are stored, and any suitable combination thereof. The computer readable storage medium used here is not interpreted as a transient signal itself, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagated by a waveguide or other transmission medium (for example, a light pulse by an optical fiber cable), or an electrical signal transmitted by a wire.

[0124] The computer-readable program described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device through a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives a computer-readable program instruction from the network, and forwards the computer-readable program instruction for storage in the computer-readable storage medium in each computing / processing device. The computer program instruction for performing the operation of the present application can be an assembly instruction, an instruction set architecture (ISA) instruction, a machine instruction, a machine-related instruction, a microcode, a firmware instruction, a state setting data, an integrated circuit configuration data, or a source code or object code written in any combination of one or more programming languages, wherein the programming language includes an object-oriented programming language, such as Smalltalk, C++, etc., and a procedural programming language, such as "C" language or similar programming language. Computer readable program instructions can be executed completely on a user's computer, partially on a user's computer, as an independent software package, partially on a user's computer, partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (e.g., using an Internet service provider to connect through the Internet). In certain embodiments, by using the state information of computer readable program instructions to personalize electronic circuits, such as programmable logic circuits, field programmable gate arrays (FPGAs) or programmable logic arrays (PLAs), the electronic circuits can execute computer readable program instructions, thereby realizing various aspects of the present application.

[0125] The embodiment of the present application also provides an in-station equipment working status identification system based on information fusion, and the in-station equipment working status identification system based on information fusion can implement the in-station equipment working status identification method based on information fusion described in the present application, but the implementation device of the in-station equipment working status identification method based on information fusion described in the present application includes but is not limited to the structure of the in-station equipment working status identification system based on information fusion listed in the present embodiment. All structural deformations and replacements of the prior art made according to the principles of the present application are included in the protection scope of the present application.

[0126] The following will describe in detail the in-station equipment working status identification system based on information fusion provided by this embodiment with reference to the drawings.

[0127] This embodiment provides an in-station equipment working status identification system based on information fusion.

[0128] See also Figure 7 , which is a schematic diagram showing the principle structure of the in-station equipment working status identification system based on information fusion in one embodiment of the present invention. Figure 7 As shown, the in-station equipment working status identification system based on information fusion includes: an acquisition module 71, a data processing module 72, an information fusion module 73, and a network construction and detection module 74.

[0129] The acquisition module 71 is used to acquire basic image information of the equipment in the target area where the offshore booster station is located.

[0130] The camera in the offshore substation obtains video information of the equipment in the station.

[0131] In this embodiment, a camera is installed in the target area of ​​the offshore booster station, and a network camera is used to capture video data of all equipment in the offshore booster station.

[0132] Specifically, in the target area, use cameras or other image acquisition devices to monitor the equipment in the offshore booster station in real time to ensure that the camera can comprehensively, accurately and clearly record the operating status and surrounding environment of the equipment in the station. Then, debug the camera to ensure that it can stably transmit video information. Through the camera, continuously obtain video information of the equipment in the station. This can be done automatically or manually to ensure that the video information of the equipment can be captured in all directions and from multiple angles. The collected video information is stored on a high-performance storage device to ensure the integrity of the video and the convenience of subsequent processing.

[0133] Based on the video information of the in-station devices, basic image information of several in-station devices is obtained.

[0134] In this embodiment, the basic image information of the equipment in the station includes but is not limited to: main transformer operation information, GIS operation information, medium voltage switch cabinet equipment information, circuit breaker opening and closing information, microcomputer detuning device operation information, voltage transformer operation information, line protection measurement and control device operation information, etc.

[0135] Specifically, based on the acquired video information, image processing software or algorithms can be used to capture frames of the stored video at certain time intervals to generate a series of static images; that is, basic image information of the equipment in the station can be extracted through image processing and recognition technology. This includes the appearance, features, color, etc. of the equipment. This process can capture the details and instantaneous status of the equipment, providing a rich data source for subsequent analysis.

[0136] The data processing module 72 is used to pre-process the basic image information to obtain first basic image information.

[0137] In this embodiment, the collected basic image information is preprocessed, including steps such as image screening, image enhancement, and image normalization, to obtain processed basic image information, thereby obtaining first basic image information.

[0138] Specifically, to screen images, we need to screen these basic image information according to some specific standards and user needs, and retain those images that meet the requirements. For example, we can screen according to factors such as image resolution, color, and content. Then, the images left after screening are enhanced by contrast enhancement, brightness adjustment, color balance, etc., so that the visual effect of the basic image information is significantly enhanced, so that it can be better analyzed and processed. Then, the pixel values ​​of the image after image enhancement are standardized so that their values ​​are all within a certain range, such as [0, 1] [0, 255], which can ensure that the subsequent processing steps are more stable and reliable.

[0139] It should be noted that the technology used in the preprocessing operation process can be implemented through corresponding processing methods and is not limited to the process of the embodiment described in this application.

[0140] The information fusion module 73 is used to perform device information fusion and decision fusion based on the first basic image information to obtain a fusion feature image of the in-station equipment.

[0141] The first basic image information is fused with the in-station device data source and sensor information to obtain a first basic image feature vector, wherein the in-station device data source includes but is not limited to: image, sensor data, sound, temperature, humidity, etc.

[0142] Information fusion is the process of integrating information from multiple in-station devices, data sources or sensors to produce more comprehensive and accurate identification results of the working status of in-station equipment.

[0143] In this application, we use a variety of different types of data sources, including but not limited to images, sensor data, sound, temperature, humidity, etc. These data sources capture multiple aspects of the working status of the equipment in the station and provide a more comprehensive information basis. Ensure data collection and time synchronization between various data sources to ensure data consistency and accuracy.

[0144] In this embodiment, feature extraction is performed based on the first basic image information, and corresponding features are extracted from the in-station device data source and sensor data to obtain first basic image features; feature fusion is performed based on the first basic image features to obtain a first basic image feature vector.

[0145] Specifically, first, feature extraction is performed on the first basic image information obtained from the video of the device in the station. This usually involves image processing technology, such as edge detection, corner detection, texture analysis, etc., which is used to identify and extract important features in the image. At the same time, corresponding features are extracted from the data source and sensor data of the device in the station. These features may include the working parameters, operating status, environmental parameters, etc. of the device, through which the performance and status of the device can be further understood.

[0146] Then, the features extracted from the image are combined with the features extracted from the device data source and sensor data to form the first basic image feature vector. The purpose of this step is to combine the visual information in the image with the data information of the device to provide a more comprehensive feature description.

[0147] Next, the first basic image feature vector obtained by fusion is further processed and analyzed, such as dimension reduction, normalization, classification or clustering, so as to better understand and utilize these features.

[0148] Finally, according to the processing and analysis, a first basic image feature vector is obtained.

[0149] From the above, it can be seen that in the process of information fusion, the present application adopts the decision fusion method.

[0150] In the information fusion process, the first basic image feature vector and the in-station device state are subjected to decision fusion based on a decision fusion rule to obtain an in-station device fusion feature image.

[0151] In this embodiment, since information fusion adopts decision fusion, specific decision rules need to be formulated for different data sources.

[0152] Specifically, the equipment status in the station includes but is not limited to: circuit breaker opening and closing status, energy storage device operation status, microcomputer detuning device operation status, voltage transformer operation status, line protection measurement and control device operation status, charging protection measurement and control device operation status, gas pressure device operation status, foreign object detection status and other equipment status. The decision fusion rules for these equipment statuses are as follows:

[0153] (1) Decision-making fusion rules for circuit breaker opening and closing states:

[0154] Set its normal working state according to the actual usage. If the test is interrupted, set the circuit breaker to close as the normal state. When the red light is on, it is in normal working state and no alarm is given. Otherwise, when the green light is on, it is considered as abnormal working and an alarm is given.

[0155] (2) Decision fusion rules for the operating status of energy storage devices:

[0156] When the indicator light of the energy storage device is red, the energy storage device is working normally and no alarm occurs; otherwise, when the indicator light is off, it is working abnormally and a red alarm occurs.

[0157] (3) Decision-making fusion rules for the operation status of the microcomputer detuning device:

[0158] There are five working states on the control panel of the microcomputer detuning device, namely: resonance, device failure, grounding, operation, and overvoltage.

[0159] (4) Decision fusion rules for voltage transformer operating status:

[0160] There are five working states on the control panel: operation, alarm and maintenance.

[0161] (5) Decision-making fusion rules for the operating status of line protection measurement and control devices:

[0162] There are eight working states on the control panel of the line protection measurement and control device, namely: operation, alarm, tripping, closing, tripping position, closing position, charging, and PT line break; the actual operating status of the device can be obtained through its indicator light display.

[0163] (6) Decision-making fusion rules for the operating status of the charging protection measurement and control device:

[0164] There are eight working states on the control panel of the charging protection measurement and control device, namely: operation, alarm, tripping, closing, tripping position, closing position, charging, and PT line break; the actual operating status of the device can be obtained through its indicator light display.

[0165] (7) Decision fusion rules for the operating status of gas pressure devices:

[0166] When the gas pressure drops to the low alarm pressure threshold, a yellow alarm is issued; when the gas pressure drops to the trip pressure value, a red alarm is issued.

[0167] Specifically, the SF6 gas pressure gauge is used as an example for explanation. For the SF6 gas pressure gauge, the normal range of gas pressure is set based on the pointer angle. When the gas pressure is at the standard value, there is no feedback information; when the gas pressure drops to the specified low alarm pressure value, a yellow alarm signal is issued; when the gas pressure drops to the trip (lockout) pressure value, a red alarm signal is issued. For the ammeter, the current value is obtained through the pointer angle and the normal range of current is set. If the current value is abnormal, an alarm is issued. For the oil level gauge, the oil volume percentage is obtained through the pointer angle and the normal range of oil volume is set. If the oil volume exceeds the normal range, a red alarm is issued.

[0168] (9) Decision fusion rules for foreign body detection status:

[0169] For foreign object detection, when no foreign object is detected, the environment is normal; when a foreign object is detected, a red alarm is issued.

[0170] The network construction and detection module 74 is used to construct an in-station device status detection network based on the in-station device fusion feature image to identify the operating status of the in-station device.

[0171] The fused feature image of the in-station equipment is input into the target detection network to obtain the operating status features of the in-station equipment.

[0172] In this embodiment, the Yolox algorithm is preferably used for processing. After the input image size is scaled to 640×640, it is sent to Backbone to extract features; then the 8-fold, 16-fold, and 32-fold sampling information is fused in the feature fusion network.

[0173] Specifically, first, the size of the fusion feature image of the in-station equipment is adjusted to 640×640 pixels. The purpose of this step is to unify the size of the image for subsequent feature extraction and network processing. Then the resized image is input into a pre-trained Backbone network, such as ResNet, VGG, etc. The main function of the Backbone network is to extract basic features from the image. Through a series of operations such as convolution and pooling, the network extracts useful feature representations from the image. In the Backbone network, high-level features are extracted from the input image through a series of convolution layers and pooling layers. These features contain information such as edges, textures, shapes, etc. in the image, which can reflect the basic structure and content of the image. Further, in the feature fusion network, features sampled at different times are fused. Specifically, features sampled at 8 times, 16 times, and 32 times are merged to obtain feature representations at different scales. This multi-scale feature fusion can capture the information of the image at different scales and enhance the representation ability of the features.

[0174] Then, a suitable feature fusion method is selected, such as simple splicing, weighted fusion, or using a specific fusion structure such as convolution, pooling, etc. According to the application requirements of the offshore booster station, this implementation adopts the splicing method.

[0175] The distribution of the operating status characteristics of the in-station equipment is analyzed based on the operating status characteristics of the in-station equipment to identify the operating status of the in-station equipment. In this embodiment, the Yolox-s algorithm is preferably used.

[0176] In this embodiment, the fused feature image of the in-station equipment is input into the target detection network for feature fusion to obtain high-level semantic information; the high-level semantic information is classified and regressed through a multi-scale decoupling detection head to achieve detection of the target; the channel information and the in-station equipment location information are determined by splicing; the channel information is matched with the in-station equipment location information to identify the operating status of the in-station equipment.

[0177] Specifically, the obtained fused image is input into the Yolox-s algorithm, which makes full use of the basic features, fuses features of different degrees, introduces a residual branch, injects feature maps of different levels, fuses the output from the backbone network, the output of the GSN at the same level, and the output after upsampling of the next level GSN, and generates semantic information with basic features. At the same time, the splicing method is used in this embodiment to replace the original addition, retaining more feature information, especially channel information. Among them, the location information is crucial for the positioning of in-station equipment. In order to help the feature extraction network accurately extract the target location information, the global pooling operation is decomposed into a pair of one-dimensional encoding operations, and each channel is encoded along the horizontal and vertical coordinates respectively. After aggregating the encoded features, a pair of direction-aware feature maps are generated, which is conducive to the backbone network to extract more target location information. Among them, in the encoding operation, the generated pair of direction-aware feature maps are spliced ​​and features are extracted using 1×1 convolution; then, the spliced ​​direction-aware features are divided into two feature vectors along the spatial dimension, and the two feature vectors are transformed using 1×1 convolution respectively, so that the number of channels of the two feature vectors is the same as the number of channels of the input features, and the attention weights in the two directions are obtained; finally, the obtained weights are multiplied by the input features to obtain the output result; the final output content is the working status of a certain in-station equipment, and the output form is mainly text, but not limited to text.

[0178] The in-station equipment working status identification model based on information fusion is built. The in-station equipment working status identification system based on information fusion can monitor the wave height in real time efficiently and accurately, reduce the workload of staff, and the monitoring data provides a basis for subsequent analysis of the laws of wave movement.

[0179] It should be noted that it should be understood that the division of the various modules of the above system is only a division of logical functions. In actual implementation, they can be fully or partially integrated into one physical entity, or they can be physically separated. And these modules can all be implemented in the form of software called by processing elements; they can also be all implemented in the form of hardware; some modules can also be implemented in the form of software called by processing elements, and some modules can be implemented in the form of hardware. For example, the x module can be a separately established processing element, or it can be integrated in a certain chip of the above system for implementation. In addition, it can also be stored in the memory of the above system in the form of program code, and called and executed by a certain processing element of the above system. The implementation of other modules is similar. In addition, these modules can be fully or partially integrated together, or they can be implemented independently. The processing element described here can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each module above can be completed by an integrated logic circuit of hardware in the processor element or instructions in the form of software.

[0180] The above modules may be one or more integrated circuits configured to implement the above methods, such as one or more application specific integrated circuits (ASIC), or one or more digital signal processors (DSP), or one or more field programmable gate arrays (FPGA). For another example, when a module is implemented in the form of a processing element scheduling program code, the processing element may be a general-purpose processor, such as a central processing unit (CPU) or other processor that can call program code. For another example, these modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).

[0181] See also Figure 8 , which is a schematic diagram showing the principle structure of the device for identifying the working status of in-station equipment based on information fusion in one embodiment of the present invention. Figure 8 As shown, this embodiment provides an in-station equipment working status identification device based on information fusion, and the in-station equipment working status identification device based on information fusion includes: a processor 81 and a memory 82; the memory 82 is used to store computer programs; the processor 81 is connected to the memory 82, and is used to execute the computer program stored in the memory 82, so that the in-station equipment working status identification device based on information fusion performs the various steps of the in-station equipment working status identification method based on information fusion as described above.

[0182] Preferably, the memory may include a random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0183] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0184] In summary, the method, system and device for identifying the working status of in-station equipment based on information fusion provided by the present application have the following beneficial effects:

[0185] The method for identifying the working status of in-station equipment based on information fusion provided by this application can improve inspection efficiency and reduce inspection costs. This application can effectively identify the working status of in-station equipment, not only providing an accurate assessment of the equipment status, but also having the flexibility to adapt to different equipment types and environments. Through the application of information fusion technology, this application has a broad application prospect and can be applied in various fields, including but not limited to the energy field, the power industry, and the inspection and maintenance industry. The in-station equipment working status identification model based on information fusion and the target detection network structure are simple, and have the characteristics of fast detection speed, high detection accuracy, and flexible deployment on the mobile terminal. It can efficiently and accurately identify the operating status of equipment in the offshore booster station in real time, reduce the workload of staff, and improve the effectiveness of the maintenance and operation of the operating status monitoring and identification of equipment in the offshore booster station.

[0186] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. A method for identifying the working status of in-station equipment based on information fusion, characterized in that: The following steps are involved: Obtain basic image information of the equipment in the target area where the offshore booster station is located; Preprocessing the basic image information to obtain first basic image information; Based on the first basic image information and the in-station device data source information, device information fusion and decision fusion are performed to obtain an in-station device fusion feature image; including: fusing the first basic image information with the in-station device data source information to obtain a first basic image feature vector; extracting features based on the first basic image information, and extracting corresponding features from the in-station device data source information to obtain first basic image features; fusing features based on the first basic image features to obtain a first basic image feature vector; in the information fusion process, performing decision fusion on the obtained first basic image feature vector and the in-station device state based on a decision fusion rule to obtain an in-station device fusion feature image; wherein the in-station device data source information includes: any one or more combinations of images, sensor data, sound, temperature, and humidity; Based on the fused feature image of the in-station equipment, an in-station equipment status detection network is constructed to identify the operating status of the in-station equipment; the in-station equipment fused feature image is input into the target detection network to obtain the operating status characteristics of the in-station equipment; based on the in-station equipment operating status characteristics, the in-station equipment operating status feature distribution is analyzed to identify the operating status of the in-station equipment; including: inputting the in-station equipment fused feature image into the target detection network for feature fusion to obtain high-level semantic information; classifying and regressing the high-level semantic information through a multi-scale decoupling detection head to achieve target detection; using a splicing method to determine the channel information and the in-station equipment location information; matching the channel information with the in-station equipment location information to identify the operating status of the in-station equipment.

2. The method for identifying the working status of in-station equipment based on information fusion according to claim 1, characterized in that: Obtaining basic image information of the equipment in the target area of ​​the offshore booster station includes the following steps: The camera in the offshore booster station obtains video information of the equipment in the station; Acquire basic image information of several devices in the station based on the video information of the devices in the station; Among them, the basic image information of the equipment in the station includes: any one or more combinations of main transformer operation information, GIS operation information, medium-voltage switchgear equipment information, circuit breaker opening and closing information, microcomputer detuning device operation information, voltage transformer operation information, and line protection and measurement and control device operation information.

3. The method for identifying the working status of in-station equipment based on information fusion according to claim 1, characterized in that: Preprocessing the basic image information to obtain first basic image information includes the following steps: The basic image information is subjected to image screening, image enhancement, and image normalization processing to obtain first basic image information.

4. The method for identifying the working status of in-station equipment based on information fusion according to claim 1, characterized in that: The station equipment status includes: any one or more combinations of the circuit breaker opening and closing status, the energy storage device operating status, the microcomputer detuning device operating status, the voltage transformer operating status, the line protection measurement and control device operating status, the charging protection measurement and control device operating status, the gas pressure device operating status, and the foreign object detection status; The decision fusion rules include: The circuit breaker is opened and closed, and its normal working state is set according to the actual use situation. If the test is interrupted, the circuit breaker is set to the normal state without alarm; otherwise, it is abnormal and an alarm is processed; The energy storage device is in operation state. When the indicator light of the energy storage device is red, the energy storage device is working normally and no alarm is given. Otherwise, the energy storage device is working abnormally and an alarm is given. The operating states of the microcomputer detuning device include: resonance, device failure, grounding, operation, and overvoltage; The voltage transformer operating status includes: operation, alarm, and maintenance; The operating status of the line protection measurement and control device includes: operation, alarm, tripping, closing, tripping, closing, charging, and PT disconnection; When the gas pressure device is in operation, a yellow alarm is issued when the gas pressure drops to the low alarm pressure threshold; a red alarm is issued when the gas pressure drops to the trip pressure value; In the foreign object detection state, when no foreign object is detected, the environment is normal; when a foreign object is detected, a red alarm is issued.

5. The method for identifying the working status of in-station equipment based on information fusion according to claim 1, characterized in that: Using the splicing method, determining the channel information and the location information of the equipment in the station includes the following steps: Decompose the global pooling operation into a one-to-one dimensional encoding operation; Each channel is encoded along the horizontal and vertical coordinates to obtain encoding features; the encoding features include: two spatial direction aggregation features After aggregating the coding features, a direction-aware feature map is obtained, thereby extracting target position information.

6. A system for identifying the working status of in-station equipment based on information fusion, characterized in that: include: An acquisition module is used to acquire basic image information of the equipment in the target area where the offshore booster station is located; A data processing module, used for preprocessing the basic image information to obtain first basic image information; An information fusion module is used to perform device information fusion and decision fusion based on the first basic image information and the in-station device data source information to obtain an in-station device fusion feature image; including: performing information fusion on the first basic image information and the in-station device data source information to obtain a first basic image feature vector; performing feature extraction based on the first basic image information and extracting corresponding features from the in-station device data source information to obtain first basic image features; performing feature fusion based on the first basic image features to obtain a first basic image feature vector; in the information fusion process, performing decision fusion on the obtained first basic image feature vector and the in-station device state based on a decision fusion rule to obtain an in-station device fusion feature image; wherein the in-station device data source information includes: any one or more combinations of images, sensor data, sound, temperature, and humidity; The network construction and detection module is used to construct an in-station equipment status detection network based on the in-station equipment fusion feature image to identify the operating status of the in-station equipment; input the in-station equipment fusion feature image into the target detection network to obtain the in-station equipment operating status characteristics; analyze the in-station equipment operating status feature distribution based on the in-station equipment operating status characteristics to identify the operating status of the in-station equipment; including: inputting the in-station equipment fusion feature image into the target detection network for feature fusion to obtain high-level semantic information; classifying and regressing the high-level semantic information through a multi-scale decoupling detection head to achieve target detection; using a splicing method to determine the channel information and the in-station equipment location information; matching the channel information with the in-station equipment location information to identify the operating status of the in-station equipment.

7. A device for identifying the working status of equipment in a station based on information fusion, characterized in that: include: Processor and memory; The memory is used to store computer programs; The processor is connected to the memory and is used to execute the computer program stored in the memory so that the in-station equipment working status identification device based on information fusion executes the in-station equipment working status identification method based on information fusion as described in any one of claims 1 to 5.

Images