A circuit breaker electromagnet condition assessment method and system

Abstract

The application discloses a circuit breaker electromagnet state evaluation method and system, and relates to the technical field of circuit breaker electromagnet states. Action data of a circuit breaker electromagnet in a closing and opening state is acquired, and image conversion is performed on the action data to obtain an electromagnet action image. The electromagnet action image is input into a pre-trained electromagnet state evaluation model. The electromagnet state evaluation model comprises a backbone network and a classification network. Feature extraction is performed on the electromagnet action image by the backbone network to obtain a state feature map. State evaluation is performed on the state feature map by the classification network to obtain a state evaluation result corresponding to the circuit breaker electromagnet. The traditional evaluation method needs to stop power supply of a circuit breaker for operation, and then a mechanical characteristic tester is used to test the circuit breaker electromagnet. However, the sticking condition of the electromagnet changes with an increase in the number of operations, thereby reducing the reliability of the circuit breaker electromagnet state evaluation.

Description

Technical Field

[0001] The present invention relates to the technical field of circuit breaker electromagnet states, and in particular to a circuit breaker electromagnet state evaluation method and system. Background Art

[0002] Circuit breakers are crucial components of power systems, and their safe and reliable operation directly impacts the safety and stability of the system. The vast majority of accidents caused by circuit breaker failures are caused by stuck electromagnets. Therefore, preventative assessments of the electromagnet status of circuit breakers are crucial. This not only enables timely detection and elimination of faults, but also effectively reduces unnecessary workload for operators, playing a crucial role in the safety and stability of the power grid.

[0003] Currently, the traditional assessment method requires operating the circuit breaker to cut off the power supply and then testing the circuit breaker electromagnet using a mechanical properties tester. However, the jamming of the electromagnet will change with the number of operations, reducing the reliability of the circuit breaker electromagnet status assessment. Summary of the Invention

[0004] The present invention provides a method and system for evaluating the status of a circuit breaker electromagnet, which solves the technical problem that the traditional evaluation method requires the circuit breaker to be powered off and then the circuit breaker electromagnet is tested using a mechanical property tester, but the electromagnet becomes stuck and changes with the number of operations, which reduces the reliability of the circuit breaker electromagnet status evaluation.

[0005] A first aspect of the present invention provides a method for evaluating the state of a circuit breaker electromagnet, comprising:

[0006] Acquiring action data of the circuit breaker electromagnet in the open and closed states, and performing image conversion on the action data to obtain an electromagnet action image;

[0007] Inputting the electromagnet motion image into a pre-trained electromagnet state assessment model, wherein the electromagnet state assessment model includes a backbone network and a classification network;

[0008] Extracting features from the electromagnet action image through the backbone network to obtain a state feature map;

[0009] The classification network is used to perform state evaluation on the state characteristic diagram to obtain a state evaluation result corresponding to the circuit breaker electromagnet.

[0010] Optionally, the action data includes closing electromagnet action data and opening electromagnet action data, and the step of performing image conversion on the action data to obtain the electromagnet action image includes:

[0011] constructing a closing electromagnet curve using the closing electromagnet action data, and extracting closing current information during a closing current valley period in the closing electromagnet curve;

[0012] constructing a tripping electromagnet curve using the tripping electromagnet action data, and extracting the tripping current information during the tripping current valley period in the tripping electromagnet curve;

[0013] The closing current information and the opening current information are fused to generate an electromagnet action image.

[0014] Optionally, the backbone network includes a convolution module, a maximum pooling layer, a first residual module, a second residual module, and an attention module connected in sequence, and the step of extracting features of the electromagnet action image through the backbone network to obtain a state feature map includes:

[0015] The convolution module is used to extract features from the electromagnet motion image to obtain a first motion feature map; wherein the convolution module includes a 3×3 convolution layer, a pooling layer, and a Relu activation layer connected in sequence;

[0016] Downsampling the first action feature map using a maximum pooling layer to obtain a second action feature map;

[0017] Performing feature extraction on the second action feature map through a first residual module to obtain a third action feature map; wherein the first residual module includes a 3×1 convolution layer, a first activation block, a 1×3 convolution layer, a first activation block, and a feature fusion layer;

[0018] Performing feature extraction on the third action feature map through a second residual module to obtain a fourth action feature map; wherein the second residual module includes a first extraction branch, a second extraction branch and a feature fusion layer;

[0019] The attention module is used to extract features from the fourth action feature map to obtain a state feature map.

[0020] Optionally, the classification network includes a second residual module, a third residual module, a global pooling layer, and an activation layer connected in sequence. The step of using the classification network to perform state evaluation on the state feature graph to obtain a state evaluation result corresponding to the circuit breaker electromagnet includes:

[0021] Performing feature extraction on the state feature map through a second residual module to obtain a first state feature map;

[0022] Performing feature extraction on the first state feature map through a third residual module to obtain a second state feature map; wherein the third residual module includes a third extraction branch and a feature fusion layer;

[0023] Performing a pooling operation on the second state feature map through a global pooling layer to obtain a third state feature map;

[0024] An activation layer is used to perform a state evaluation on the third state characteristic diagram to obtain a state evaluation result corresponding to the circuit breaker electromagnet.

[0025] Optionally, the step of performing feature extraction on the second action feature map by using a first residual module to obtain a third action feature map includes:

[0026] Performing feature extraction on the second action feature map through a 3×1 convolutional layer to obtain a first feature map;

[0027] Performing nonlinear mapping on the first feature map using a first activation block to obtain a second feature map; wherein the first activation block includes a batch normalization layer and a DY-ReLU activation layer connected in sequence;

[0028] Perform feature extraction on the second feature map through a 1×3 convolutional layer to obtain a third feature map;

[0029] Performing nonlinear mapping on the third feature map using the first activation block to obtain a fourth feature map;

[0030] A feature fusion layer is used to perform feature fusion on the second action feature map and the fourth feature map to obtain a third action feature map.

[0031] Optionally, the step of performing feature extraction on the third action feature map through a second residual module to obtain a fourth action feature map includes:

[0032] Performing feature extraction on the third action feature map through the first extraction branch to obtain a fifth feature map; wherein the first extraction branch includes a 3×3 convolutional layer, a batch normalization layer, a Relu activation layer, a 3×3 convolutional layer, a batch normalization layer, and a Relu activation layer connected in sequence;

[0033] Performing feature extraction on the third action feature map through the second extraction branch to obtain a sixth feature map; wherein the second extraction branch includes a 3×3 convolutional layer, a batch normalization layer, and a Relu activation layer connected in sequence;

[0034] A feature fusion layer is used to perform feature fusion on the fifth feature map and the sixth feature map to obtain a fourth action feature map.

[0035] Optionally, the step of extracting features from the first state feature map using a third residual module to obtain a second state feature map includes:

[0036] Performing feature extraction on the first state feature map through the third extraction branch to obtain a seventh feature map; wherein the third extraction branch includes a 3×3 convolutional layer, a batch normalization layer, a Relu activation layer, a 3×3 convolutional layer, a batch normalization layer, and a Relu activation layer connected in sequence;

[0037] The first state feature map and the seventh feature map are subjected to feature fusion measurement layer fusion to obtain a second state feature map.

[0038] A second aspect of the present invention provides a circuit breaker electromagnet state assessment system, comprising:

[0039] An acquisition module is used to obtain the action data of the circuit breaker electromagnet in the open and close states, and perform image conversion on the action data to obtain an electromagnet action image;

[0040] an execution module, configured to input the electromagnet motion image into a pre-trained electromagnet state assessment model, wherein the electromagnet state assessment model includes a backbone network and a classification network;

[0041] An extraction module, configured to extract features from the electromagnet motion image through the backbone network to obtain a state feature map;

[0042] An evaluation module is used to use the classification network to perform a state evaluation on the state characteristic diagram to obtain a state evaluation result corresponding to the circuit breaker electromagnet.

[0043] A third aspect of the present invention provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the circuit breaker electromagnet state assessment method as described in any one of the above items.

[0044] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed, implements the circuit breaker electromagnet state evaluation method as described in any one of the above items.

[0045] It can be seen from the above technical solutions that the present invention has the following advantages:

[0046] The present invention obtains the action data of the circuit breaker electromagnet in the open and closed states and performs image conversion on the action data to obtain an electromagnet action image. The electromagnet action image is then evaluated using a pre-trained electromagnet state assessment model, thereby enabling real-time assessment of the state of the circuit breaker electromagnet. This overcomes the drawback of traditional assessment methods, which require powering off the circuit breaker and then testing the circuit breaker electromagnet using a mechanical properties tester, making it impossible to accurately assess the state of the circuit breaker electromagnet. Compared with traditional assessment methods, the present invention does not require powering off the circuit breaker and can still assess the state of the circuit breaker electromagnet, thus ensuring the accuracy of the circuit breaker electromagnet state assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0048] Figure 1 A flowchart of a method for evaluating the state of a circuit breaker electromagnet provided in the first embodiment of the present invention;

[0049] Figure 2 A flowchart of a method for evaluating the state of a circuit breaker electromagnet provided in the second embodiment of the present invention;

[0050] Figure 3 A schematic diagram of a closing electromagnet curve provided in the second embodiment of the present invention;

[0051] Figure 4 A schematic diagram of a tripping electromagnet curve provided in the second embodiment of the present invention;

[0052] Figure 5 A schematic diagram of the electromagnet state assessment model structure provided in the second embodiment of the present invention;

[0053] Figure 6 This is a structural block diagram of a circuit breaker electromagnet status assessment system provided in the third embodiment of the present invention;

[0054] Figure 7 This is a structural block diagram of a computer device provided in Example 4 of the present invention. DETAILED DESCRIPTION

[0055] An embodiment of the present invention provides a method and system for evaluating the status of a circuit breaker electromagnet, which is used to solve the technical problem that the traditional evaluation method requires operating the circuit breaker to shut down the power supply and then testing the circuit breaker electromagnet using a mechanical properties tester. However, the jamming of the electromagnet will change with the increase in the number of operations, which reduces the reliability of the circuit breaker electromagnet status evaluation.

[0056] In order to make the purpose, features, and advantages of the present invention more obvious and easy to understand, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described below are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0057] See also Figure 1 , Figure 1 This is a flowchart of the steps of a circuit breaker electromagnet status assessment method provided in Example 1 of the present invention.

[0058] The present invention provides a method for evaluating the state of a circuit breaker electromagnet, comprising:

[0059] Step 101: Acquire the action data of the circuit breaker electromagnet in the open and closed states, and perform image conversion on the action data to obtain an electromagnet action image;

[0060] Action data refers to the electromagnet action data when the circuit breaker is closed and opened.

[0061] In an embodiment of the present invention, electromagnet action data when the circuit breaker is opened and closed is obtained, and the electromagnet action data is converted into an image to obtain a card playing electromagnet action image.

[0062] Step 102: Input the electromagnet motion image into a pre-trained electromagnet state assessment model, wherein the electromagnet state assessment model includes a backbone network and a classification network;

[0063] In an embodiment of the present invention, the electromagnet motion image is used as input to a pre-trained electromagnet state assessment model, wherein the electromagnet state assessment model includes a backbone network and a classification network.

[0064] It's worth noting that the specific training process for the electromagnet state assessment model is as follows: 1. A damping spring is installed on the moving iron core of the opening and closing coil of a training circuit breaker (to simulate the resistance of the electromagnet's moving iron core to motion) to obtain a target circuit breaker. 2. Mechanical characteristic tests are performed on the target circuit breaker using a mechanical characteristic tester (measuring the opening and closing tripping time, opening current valley time, closing current valley time, opening time, and closing time) to obtain training motion data. 3. The training motion data is preprocessed to generate a motion feature set. 4. The motion feature set is used as input for training the preset initial electromagnet state assessment model, outputting motion assessment data. 5. The training loss function value of the motion feature set is calculated based on the motion assessment data. 6. When the training loss function value is greater than or equal to the preset standard loss function value, the network parameters of the initial electromagnet state assessment model are adjusted using a grid search method or a random search method until the training loss function value is less than the preset standard loss function value. 7. When the training loss function value is less than the preset standard loss function value, the electromagnet state assessment model is generated.

[0065] Step 103: extract features from the electromagnet action image through the backbone network to obtain a state feature map;

[0066] In an embodiment of the present invention, feature extraction is performed on the electromagnet action image through a backbone network to obtain a state feature map, wherein the backbone network includes a connected convolution module, a maximum pooling layer, a first residual module, a second residual module and an attention module.

[0067] Step 104: Use a classification network to perform state evaluation on the state characteristic diagram to obtain a state evaluation result corresponding to the circuit breaker electromagnet.

[0068] In an embodiment of the present invention, a classification network is used to perform a state assessment on the state characteristic diagram to obtain a circuit breaker electromagnet identification diagram. Based on the identification of the circuit breaker electromagnet identification diagram, a state assessment result corresponding to the circuit breaker electromagnet is obtained (the state assessment result is classified according to the degree: normal, caution, abnormal, and severe).

[0069] In an embodiment of the present invention, the present invention obtains the action data of the circuit breaker electromagnet in the open and closed states and performs image conversion on the action data to obtain an electromagnet action image, and then uses a pre-trained electromagnet state assessment model to perform state assessment on the electromagnet action image, thereby achieving real-time assessment of the state of the circuit breaker electromagnet. This overcomes the defect of the traditional assessment method that requires the circuit breaker to be powered off and then the circuit breaker electromagnet to be tested using a mechanical property tester, which cannot accurately assess the state of the circuit breaker electromagnet. Compared with the traditional assessment method, the present invention does not require the circuit breaker to be powered off and can also assess the state of the circuit breaker electromagnet, thereby ensuring the accuracy of the circuit breaker electromagnet state assessment.

[0070] See also Figure 2 , Figure 2 This is a flowchart of the steps of a circuit breaker electromagnet status assessment method provided in the second embodiment of the present invention.

[0071] The present invention provides a method for evaluating the state of a circuit breaker electromagnet, comprising:

[0072] Step 201: Acquire the action data of the circuit breaker electromagnet in the open and closed states, wherein the action data includes the closing electromagnet action data and the opening electromagnet action data;

[0073] In the embodiment of the present invention, closing electromagnet action data and opening electromagnet action data of the circuit breaker electromagnet in the open and closed states are acquired.

[0074] Step 202: construct a closing electromagnet curve using the closing electromagnet action data, and extract closing current information during the closing current valley period in the closing electromagnet curve;

[0075] In the embodiment of the present invention, see Figure 3 As shown, the closing electromagnet action data is used to construct the closing electromagnet curve, and the closing current curve of the closing current valley period T1 in the closing electromagnet curve is extracted.

[0076] Step 203: constructing a tripping electromagnet curve using the tripping electromagnet action data, and extracting the tripping current information during the tripping current valley period in the tripping electromagnet curve;

[0077] In the embodiment of the present invention, see Figure 4 As shown, the opening electromagnet action data is used to construct the opening electromagnet curve, and the opening current curve of the opening current valley period T1 in the opening electromagnet curve is extracted.

[0078] Step 204: Fusing the closing current information and the opening current information to generate an electromagnet action image.

[0079] In the embodiment of the present invention, the closing current curve and the opening current curve are spliced ​​together to obtain an RGB image of the electromagnet action.

[0080] Step 205: Input the electromagnet motion image into a pre-trained electromagnet state evaluation model, wherein the electromagnet state evaluation model includes a backbone network and a classification network;

[0081] In the embodiment of the present invention, see Figure 5 As shown, the electromagnet action image is used as the input of a pre-trained electromagnet state evaluation model, wherein the electromagnet state evaluation model includes a backbone network and a classification network.

[0082] Step 206: extract features from the electromagnet motion image through the backbone network to obtain a state feature map;

[0083] Furthermore, the backbone network includes a convolutional module, a maximum pooling layer, a first residual module, a second residual module, and an attention module connected in sequence. Step 206 includes the following sub-steps:

[0084] S11. Extract features from the electromagnet motion image through a convolution module to obtain a first motion feature map; wherein the convolution module includes a 3×3 convolution layer, a pooling layer, and a Relu activation layer connected in sequence;

[0085] In this embodiment of the present invention, a convolution module is used to extract image features of an electromagnet motion image, and pooling and nonlinear mapping are performed on the image features to obtain a first motion feature map. The convolution module includes a 3×3 convolution layer, a pooling layer, and a Relu activation layer connected in sequence.

[0086] S12, downsampling the first action feature map using a maximum pooling layer to obtain a second action feature map;

[0087] In the embodiment of the present invention, the maximum pooling layer is used to downsample the first action feature map (thereby reducing the size and number of parameters of the first action feature map) to obtain the second action feature map.

[0088] S13. Perform feature extraction on the second action feature map through a first residual module to obtain a third action feature map; wherein the first residual module includes a 3×1 convolution layer, a first activation block, a 1×3 convolution layer, a first activation block, and a feature fusion layer;

[0089] Furthermore, S13 includes the following sub-steps:

[0090] S131, extracting features from the second action feature map through a 3×1 convolutional layer to obtain a first feature map;

[0091] In the embodiment of the present invention, the image features of the second action feature map are extracted through a 3×1 convolutional layer to obtain the first feature map.

[0092] S132. Perform nonlinear mapping on the first feature map using a first activation block to obtain a second feature map; wherein the first activation block includes a batch normalization layer and a DY-Relu activation layer connected in sequence;

[0093] In an embodiment of the present invention, a first activation block is used to perform nonlinear mapping on the first feature map to obtain a second feature map, wherein the first activation block includes a batch normalization layer and a DY-Relu activation layer connected in sequence.

[0094] S133, extracting features from the second feature map through a 1×3 convolutional layer to obtain a third feature map;

[0095] In the embodiment of the present invention, the image features of the second feature map are extracted through a 1×3 convolutional layer to obtain a third feature map.

[0096] S134. Perform nonlinear mapping on the third feature map using the first activation block to obtain a fourth feature map.

[0097] S135. Use a feature fusion layer to perform feature fusion on the second action feature map and the fourth feature map to obtain a third action feature map.

[0098] In an embodiment of the present invention, the first activation block is used to perform nonlinear mapping on the third feature map to obtain a fourth feature map, and the second action feature map and the fourth feature map are feature spliced ​​through a feature fusion layer to obtain a third action feature map.

[0099] It is worth mentioning that by setting up 3×1 convolutional layers and 1×3 convolutional layers, the computational efficiency of the model is improved while ensuring the evaluation accuracy.

[0100] S14. Extract features from the third action feature map using a second residual module to obtain a fourth action feature map; wherein the second residual module includes a first extraction branch, a second extraction branch, and a feature fusion layer;

[0101] Furthermore, S14 includes the following sub-steps:

[0102] S141. Perform feature extraction on the third action feature map through the first extraction branch to obtain a fifth feature map; wherein the first extraction branch includes a 3×3 convolutional layer, a batch normalization layer, a Relu activation layer, a 3×3 convolutional layer, a batch normalization layer, and a Relu activation layer connected in sequence;

[0103] In this embodiment of the present invention, a fifth feature map is obtained by performing continuous image feature extraction and nonlinear mapping operations on the third action feature map through a first extraction branch. The first extraction branch includes a 3×3 convolutional layer, a batch normalization layer, a Relu activation layer, a 3×3 convolutional layer, a batch normalization layer, and a Relu activation layer connected in sequence.

[0104] S142. Perform feature extraction on the third action feature map through the second extraction branch to obtain a sixth feature map; wherein the second extraction branch includes a 3×3 convolutional layer, a batch normalization layer, and a Relu activation layer connected in sequence;

[0105] In an embodiment of the present invention, the second extraction branch performs image feature extraction and nonlinear mapping operations on the third action feature map to obtain a sixth feature map. The second extraction branch includes a 3×3 convolutional layer, a batch normalization layer, and a ReLU activation layer connected in sequence.

[0106] S143. Use a feature fusion layer to perform feature fusion on the fifth feature map and the sixth feature map to obtain a fourth action feature map.

[0107] In the embodiment of the present invention, the image features of the fifth feature map and the sixth feature map are fused through a feature fusion layer to obtain a fourth action feature map.

[0108] S15. Use the attention module to extract features from the fourth action feature map to obtain a state feature map.

[0109] In the embodiment of the present invention, the attention module is used to extract the important image features of the fourth action feature map to obtain the state feature map

[0110] Step 207: Use a classification network to perform state evaluation on the state characteristic graph to obtain a state evaluation result corresponding to the circuit breaker electromagnet.

[0111] Furthermore, the classification network includes a second residual module, a third residual module, a global pooling layer, and an activation layer connected in sequence, and step 207 includes the following sub-steps:

[0112] S21, extracting features from the state feature map through a second residual module to obtain a first state feature map;

[0113] In an embodiment of the present invention, a second residual module is used to extract image features of the state feature map to obtain a first state feature map.

[0114] S22. Perform feature extraction on the first state feature map through a third residual module to obtain a second state feature map; wherein the third residual module includes a third extraction branch and a feature fusion layer;

[0115] Furthermore, S22 includes the following sub-steps:

[0116] S221. Perform feature extraction on the first state feature map through a third extraction branch to obtain a seventh feature map; wherein the third extraction branch includes a 3×3 convolutional layer, a batch normalization layer, a Relu activation layer, a 3×3 convolutional layer, a batch normalization layer, and a Relu activation layer connected in sequence;

[0117] In this embodiment of the present invention, a third extraction branch performs continuous image feature extraction and nonlinear mapping on the first state feature map to obtain a seventh feature map. The third extraction branch includes a 3×3 convolutional layer, a batch normalization layer, a Relu activation layer, a 3×3 convolutional layer, a batch normalization layer, and a Relu activation layer connected in sequence.

[0118] S222: Use a feature fusion layer to perform feature fusion on the first state feature map and the seventh feature map to obtain a second state feature map.

[0119] In the embodiment of the present invention, a feature fusion layer is used to fuse the image features of the first state feature map and the seventh feature map to obtain the second state feature map.

[0120] S23. Perform a pooling operation on the second state feature map through a global pooling layer to obtain a third state feature map;

[0121] In an embodiment of the present invention, a pooling operation is performed on the entire area of ​​the second state feature map through a global pooling layer to obtain a third state feature map.

[0122] S24. Performing a state evaluation on the third state characteristic diagram using the activation layer to obtain a state evaluation result corresponding to the circuit breaker electromagnet.

[0123] In the embodiment of the present invention, the PLTanh activation layer or the Soflmax activation layer is used to perform state evaluation on the third state characteristic diagram to obtain a state evaluation result corresponding to the circuit breaker electromagnet.

[0124] In an embodiment of the present invention, the present invention obtains the action data of the circuit breaker electromagnet in the open and closed states and performs image conversion on the action data to obtain an electromagnet action image, and then uses a pre-trained electromagnet state assessment model to perform state assessment on the electromagnet action image, thereby achieving real-time assessment of the state of the circuit breaker electromagnet. This overcomes the defect of the traditional assessment method that requires the circuit breaker to be powered off and then the circuit breaker electromagnet to be tested using a mechanical property tester, which cannot accurately assess the state of the circuit breaker electromagnet. Compared with the traditional assessment method, the present invention does not require the circuit breaker to be powered off and can also assess the state of the circuit breaker electromagnet, thereby ensuring the accuracy of the circuit breaker electromagnet state assessment.

[0125] See also Figure 6 , Figure 6 This is a structural block diagram of a circuit breaker electromagnet status assessment system provided in Example 3 of the present invention.

[0126] The present invention provides a circuit breaker electromagnet state assessment system, comprising:

[0127] The acquisition module 301 is used to obtain the action data of the circuit breaker electromagnet in the open and close states, and perform image conversion on the action data to obtain an electromagnet action image;

[0128] An execution module 302 is configured to input the electromagnet motion image into a pre-trained electromagnet state evaluation model, wherein the electromagnet state evaluation model includes a backbone network and a classification network;

[0129] Extraction module 303, used to extract features from the electromagnet motion image through the backbone network to obtain a state feature map;

[0130] The evaluation module 304 is configured to perform a state evaluation on the state characteristic diagram using a classification network to obtain a state evaluation result corresponding to the circuit breaker electromagnet.

[0131] Furthermore, the action data includes closing electromagnet action data and opening electromagnet action data, and the acquisition module 301 includes:

[0132] A first extraction submodule is configured to construct a closing electromagnet curve using the closing electromagnet action data, and extract closing current information during a closing current valley period in the closing electromagnet curve;

[0133] The second extraction submodule is used to construct a tripping electromagnet curve using the tripping electromagnet action data, and extract the tripping current information during the tripping current valley period in the tripping electromagnet curve;

[0134] The image conversion submodule is used to fuse the closing current information and the opening current information to generate an electromagnet action image.

[0135] Furthermore, the backbone network includes a convolution module, a maximum pooling layer, a first residual module, a second residual module and an attention module connected in sequence, and the extraction module 303 includes:

[0136] A first feature extraction submodule is configured to extract features from the electromagnet motion image using a convolution module to obtain a first motion feature map; wherein the convolution module includes a 3×3 convolution layer, a pooling layer, and a Relu activation layer connected in sequence;

[0137] A second feature extraction submodule is used to downsample the first action feature map using a maximum pooling layer to obtain a second action feature map;

[0138] a third feature extraction submodule, configured to extract features from the second action feature map using the first residual module to obtain a third action feature map; wherein the first residual module includes a 3×1 convolution layer, a first activation block, a 1×3 convolution layer, a first activation block, and a feature fusion layer;

[0139] a fourth feature extraction submodule, configured to extract features from the third action feature map using the second residual module to obtain a fourth action feature map; wherein the second residual module includes a first extraction branch, a second extraction branch, and a feature fusion layer;

[0140] The fifth feature extraction submodule is used to use the attention module to extract features from the fourth action feature map to obtain a state feature map.

[0141] Furthermore, the classification network includes a second residual module, a third residual module, a global pooling layer, and an activation layer connected in sequence, and an evaluation module 304, including:

[0142] A first evaluation submodule is configured to extract features from the state feature map through a second residual module to obtain a first state feature map;

[0143] a second evaluation submodule, configured to extract features from the first state feature map through a third residual module to obtain a second state feature map; wherein the third residual module includes a third extraction branch and a feature fusion layer;

[0144] A third evaluation submodule is used to perform a pooling operation on the second state feature map through a global pooling layer to obtain a third state feature map;

[0145] The fourth evaluation submodule is configured to perform a state evaluation on the third state characteristic diagram using the activation layer to obtain a state evaluation result corresponding to the circuit breaker electromagnet.

[0146] Furthermore, the third feature extraction submodule includes:

[0147] A first extraction unit is configured to extract features from the second action feature map through a 3×1 convolutional layer to obtain a first feature map;

[0148] A second extraction unit is configured to perform nonlinear mapping on the first feature map using the first activation block to obtain a second feature map; wherein the first activation block includes a batch normalization layer and a DY-ReLU activation layer connected in sequence;

[0149] A third extraction unit is used to extract features from the second feature map through a 1×3 convolutional layer to obtain a third feature map;

[0150] a fourth extraction unit, configured to perform nonlinear mapping on the third feature map using the first activation block to obtain a fourth feature map;

[0151] The first fusion unit is used to perform feature fusion on the second action feature map and the fourth feature map using a feature fusion layer to obtain a third action feature map.

[0152] Furthermore, the fourth feature extraction submodule includes:

[0153] a fifth extraction unit, configured to perform feature extraction on the third action feature map through the first extraction branch to obtain a fifth feature map; wherein the first extraction branch includes a 3×3 convolutional layer, a batch normalization layer, a Relu activation layer, a 3×3 convolutional layer, a batch normalization layer, and a Relu activation layer connected in sequence;

[0154] a sixth extraction unit, configured to extract features from the third action feature map through the second extraction branch to obtain a sixth feature map; wherein the second extraction branch includes a 3×3 convolutional layer, a batch normalization layer, and a Relu activation layer connected in sequence;

[0155] The second fusion unit is used to perform feature fusion on the fifth feature map and the sixth feature map using a feature fusion layer to obtain a fourth action feature map.

[0156] Furthermore, the second evaluation submodule includes:

[0157] a first evaluation unit, configured to extract features from the first state feature map through a third extraction branch to obtain a seventh feature map; wherein the third extraction branch includes a 3×3 convolutional layer, a batch normalization layer, a Relu activation layer, a 3×3 convolutional layer, a batch normalization layer, and a Relu activation layer connected in sequence;

[0158] The second evaluation unit is used to perform feature fusion on the first state feature map and the seventh feature map by using a feature fusion measurement layer to obtain a second state feature map.

[0159] See also Figure 7 , Figure 7 This is a structural block diagram of a computer device provided in Example 4 of the present invention.

[0160] An electronic device according to an embodiment of the present invention includes: a memory 401 and a processor 402. The memory 401 stores a computer program. When the computer program is executed by the processor 402, the processor 402 executes the circuit breaker electromagnet state assessment method according to any of the above embodiments.

[0161] Memory 401 may be an electronic memory such as flash memory, EEPROM (Electrically Erasable Programmable Read-Only Memory), EPROM, hard disk, or ROM. Memory 401 has storage space 403 for program code 413 for executing any of the method steps described above. For example, storage space 403 for program code may include individual program codes 413 for implementing various steps in the method described above. These program codes may be read from or written to one or more computer program products. These computer program products include program code carriers such as hard disks, compact disks (CDs), memory cards, or floppy disks. The program codes may be compressed, for example, in a suitable format. When executed by a processing device, these codes cause the processing device to execute the various steps in the method described above. These program codes may be read from or written to one or more computer program products. These computer program products include program code carriers such as hard disks, compact disks (CDs), memory cards, or floppy disks. The program codes may be compressed, for example, in a suitable format. When these codes are executed by a computing and processing device, they cause the computing and processing device to execute the steps of the circuit breaker electromagnet state evaluation method described above.

[0162] The fifth embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the circuit breaker electromagnet state evaluation method according to any of the above embodiments is implemented.

[0163] Embodiment 6 of the present invention further provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer executes the circuit breaker electromagnet state assessment method as described in any of the above embodiments.

[0164] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0165] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interface, device or unit, which can be electrical, mechanical or other forms.

[0166] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0167] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0168] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the method of the present invention. The aforementioned storage medium includes various media that can store program code, such as USB flash drives, mobile hard drives, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0169] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for evaluating the state of a circuit breaker electromagnet, characterized in that: include: Acquiring action data of the circuit breaker electromagnet in the open and closed states, and performing image conversion on the action data to obtain an electromagnet action image; Inputting the electromagnet motion image into a pre-trained electromagnet state assessment model, wherein the electromagnet state assessment model includes a backbone network and a classification network; Extracting features from the electromagnet action image through the backbone network to obtain a state feature map; Using the classification network to perform state evaluation on the state characteristic diagram to obtain a state evaluation result corresponding to the circuit breaker electromagnet; The backbone network includes a convolution module, a maximum pooling layer, a first residual module, a second residual module, and an attention module connected in sequence. The step of extracting features from the electromagnet action image through the backbone network to obtain a state feature map includes: The convolution module is used to extract features from the electromagnet motion image to obtain a first motion feature map; wherein the convolution module includes a 3×3 convolution layer, a pooling layer, and a Relu activation layer connected in sequence; Downsampling the first action feature map using a maximum pooling layer to obtain a second action feature map; Performing feature extraction on the second action feature map through a first residual module to obtain a third action feature map; Performing feature extraction on the third action feature map through a second residual module to obtain a fourth action feature map; The attention module is used to extract features from the fourth action feature map to obtain a state feature map.

2. The circuit breaker electromagnet status evaluation method according to claim 1, characterized in that: The action data includes closing electromagnet action data and opening electromagnet action data, and the step of performing image conversion on the action data to obtain the electromagnet action image includes: constructing a closing electromagnet curve using the closing electromagnet action data, and extracting closing current information during a closing current valley period in the closing electromagnet curve; constructing a tripping electromagnet curve using the tripping electromagnet action data, and extracting the tripping current information during the tripping current valley period in the tripping electromagnet curve; The closing current information and the opening current information are fused to generate an electromagnet action image.

3. The circuit breaker electromagnet status evaluation method according to claim 1, characterized in that: The classification network includes a second residual module, a third residual module, a global pooling layer, and an activation layer connected in sequence. The step of using the classification network to perform state evaluation on the state feature graph to obtain a state evaluation result corresponding to the circuit breaker electromagnet includes: Performing feature extraction on the state feature map through a second residual module to obtain a first state feature map; Performing feature extraction on the first state feature map through a third residual module to obtain a second state feature map; wherein the third residual module includes a third extraction branch and a feature fusion layer; Performing a pooling operation on the second state feature map through a global pooling layer to obtain a third state feature map; An activation layer is used to perform a state evaluation on the third state characteristic diagram to obtain a state evaluation result corresponding to the circuit breaker electromagnet.

4. The circuit breaker electromagnet status evaluation method according to claim 1, characterized in that: The first residual module includes a 3×1 convolution layer, a first activation block, a 1×3 convolution layer, a first activation block, and a feature fusion layer. The step of extracting features from the second action feature map by the first residual module to obtain a third action feature map includes: Performing feature extraction on the second action feature map through a 3×1 convolutional layer to obtain a first feature map; Performing nonlinear mapping on the first feature map using a first activation block to obtain a second feature map; wherein the first activation block includes a batch normalization layer and a DY-ReLU activation layer connected in sequence; Perform feature extraction on the second feature map through a 1×3 convolutional layer to obtain a third feature map; Performing nonlinear mapping on the third feature map using the first activation block to obtain a fourth feature map; A feature fusion layer is used to perform feature fusion on the second action feature map and the fourth feature map to obtain a third action feature map.

5. The circuit breaker electromagnet status evaluation method according to claim 1, characterized in that: The second residual module includes a first extraction branch, a second extraction branch and a feature fusion layer. The step of extracting features from the third action feature map by the second residual module to obtain a fourth action feature map includes: Performing feature extraction on the third action feature map through the first extraction branch to obtain a fifth feature map; wherein the first extraction branch includes a 3×3 convolutional layer, a batch normalization layer, a Relu activation layer, a 3×3 convolutional layer, a batch normalization layer, and a Relu activation layer connected in sequence; Performing feature extraction on the third action feature map through the second extraction branch to obtain a sixth feature map; wherein the second extraction branch includes a 3×3 convolutional layer, a batch normalization layer, and a Relu activation layer connected in sequence; A feature fusion layer is used to perform feature fusion on the fifth feature map and the sixth feature map to obtain a fourth action feature map.

6. The circuit breaker electromagnet status evaluation method according to claim 3, characterized in that: The step of extracting features from the first state feature map by a third residual module to obtain a second state feature map includes: Performing feature extraction on the first state feature map through the third extraction branch to obtain a seventh feature map; wherein the third extraction branch includes a 3×3 convolutional layer, a batch normalization layer, a Relu activation layer, a 3×3 convolutional layer, a batch normalization layer, and a Relu activation layer connected in sequence; The first state feature map and the seventh feature map are subjected to feature fusion measurement layer fusion to obtain a second state feature map.

7. A circuit breaker electromagnet status assessment system, used to implement the circuit breaker electromagnet status assessment method according to any one of claims 1 to 6, characterized in that: include: An acquisition module is used to obtain the action data of the circuit breaker electromagnet in the open and close states, and perform image conversion on the action data to obtain an electromagnet action image; an execution module, configured to input the electromagnet motion image into a pre-trained electromagnet state assessment model, wherein the electromagnet state assessment model includes a backbone network and a classification network; An extraction module, configured to extract features from the electromagnet motion image through the backbone network to obtain a state feature map; An evaluation module is used to use the classification network to perform a state evaluation on the state characteristic diagram to obtain a state evaluation result corresponding to the circuit breaker electromagnet.

8. An electronic device, characterized in that: The method comprises a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the processor executes the steps of the circuit breaker electromagnet state evaluation method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed, the circuit breaker electromagnet state evaluation method according to any one of claims 1 to 6 is implemented.

Images