Nuclear power equipment data augmentation method and device based on 5G, equipment and medium

Abstract

The invention discloses a 5G-based nuclear power equipment data augmentation method, apparatus and device, and a medium. The method comprises the steps of obtaining a first training sample set; determining similar equipment of the target nuclear power equipment; obtaining a second training sample set; converting the second training sample set into a third training sample set; the third training sample set comprises a plurality of third training samples, and each third training sample comprises target augmented data which is obtained by converting second sampling data and has the same feature distribution with the first sampling data, and a second health state category label corresponding to the second sampling data; and training a health state evaluation model by using the first training sample set and the third training sample set. The training data sample of the target nuclear power equipment is expanded by means of a small amount of related data of the target nuclear power equipment and a large amount of complete data of similar equipment, and the training sample data requirement of a health state evaluation model is met.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a 5G-based nuclear power equipment data augmentation method, device, equipment and medium. Background Art

[0002] A nuclear power plant consists of nearly 300 systems and over 6,000 sets of mechanical equipment. Among them, rotating equipment such as pumps, motors, and fans exist in multiple process systems, are widely distributed, and have many types and quantities. At the same time, rotating equipment is the main carrier of energy conversion and transmission, and plays an important role in the operation of the system.

[0003] Rotating equipment is prone to performance degradation in long-term operation. In order to ensure continuous and reliable operation of the equipment and avoid permanent damage caused by failures, it is necessary to monitor and evaluate the equipment. However, due to the extreme environments such as radiation, high temperature, and high pressure in nuclear power scenarios, it is difficult to install sensors and there are few detection points. The monitoring data available in current nuclear power scenarios is relatively scarce and cannot meet the data volume requirements for equipment status monitoring and evaluation. Summary of the invention

[0004] This application aims to propose a 5G-based nuclear power equipment data augmentation method, device, equipment and medium, which can expand the training data samples of the target nuclear power equipment with the help of a small amount of target nuclear power equipment-related data and a large amount of complete data of similar equipment, so as to meet the training sample data requirements of the health status assessment model.

[0005] According to the first aspect of the present application, a 5G-based nuclear power equipment data augmentation method includes:

[0006] Acquire a first training sample set; the first training sample set includes a plurality of first training samples of the target nuclear power equipment whose health status is to be detected, each of the first training samples includes a first sampling data of the target nuclear power equipment and a first health status category label corresponding to the first sampling data;

[0007] Identify similar equipment to the target nuclear power equipment;

[0008] Acquire a second training sample set; the second training sample set includes a plurality of second training samples of the similar device, each of the second training samples includes a second sampling data of the similar device, and a second health status category label corresponding to the second sampling data;

[0009] Converting the second training sample set into a third training sample set; the third training sample set includes a plurality of third training samples, each of the third training samples includes a target augmented data converted from the second sampled data and having the same feature distribution as the first sampled data, and a second health status category label corresponding to the second sampled data;

[0010] The first training sample set and the third training sample set are used to train a health status assessment model; the health status assessment model is used to assess the health status of the target nuclear power equipment.

[0011] According to some embodiments of the present application, before converting the second training sample set into a third training sample set, the method further includes:

[0012] Acquire a fourth training sample set; the fourth training sample set includes a plurality of fourth training samples, each of the fourth training samples includes a third sampling data of the similar device and fourth sampling data of the target nuclear power device corresponding to the third sampling data;

[0013] Using the fourth training sample set, training the first neural network until a training stop condition of the first neural network is met, thereby obtaining a trained first neural network;

[0014] The converting the second training sample set into a third training sample set comprises:

[0015] The second training sample is input into the trained first neural network to obtain a third training sample corresponding to the second training sample.

[0016] According to some embodiments of the present application, the step of training the first neural network by using the fourth training sample set until a training stop condition of the first neural network is met to obtain the trained first neural network includes:

[0017] For the i-th fourth training sample in the fourth training sample set, input the i-th fourth training sample into the first neural network to obtain the i-th augmented data corresponding to the i-th fourth training sample; wherein i is a positive integer;

[0018] Determine a first loss situation corresponding to the i-th fourth training sample according to the fourth sampling data of the i-th fourth training sample and the i-th augmented data;

[0019] When the first loss situation does not meet the first training stop condition of the first neural network, adjust the parameters of the first neural network to obtain the updated first neural network, update i to i+1, and return to input the second training sample set into the first neural network until the first training stop condition is met.

[0020] According to some embodiments of the present application, determining a first loss situation corresponding to the i-th fourth training sample according to the fourth sampling data of the i-th fourth training sample and the i-th augmented data includes:

[0021] Inputting the fourth sampling data of the i-th fourth training sample and the i-th augmented data into a first discriminant neural network to obtain a corresponding first discriminant result;

[0022] According to the first discrimination result, a first discrimination accuracy rate and a first loss value are obtained, wherein the first discrimination accuracy rate represents the similarity of feature distribution between the plurality of the fourth sampling data and the plurality of the i-th augmented data;

[0023] The first loss condition satisfies the first training stop condition, including that the first discrimination accuracy is within a first accuracy range.

[0024] According to some embodiments of the present application, when the target loss value satisfies the first training stop condition of the first neural network, the first neural network is trained by using the fourth training sample set until the training stop condition of the first neural network is met to obtain the trained first neural network, and further includes:

[0025] For the jth fourth training sample in the fourth training sample set, input the jth fourth training sample into the first neural network to be trained to obtain the jth augmented data corresponding to the jth fourth training sample; wherein j is a positive integer;

[0026] Inputting the j-th augmented data into a second neural network to obtain the j-th restored data of the similar device; wherein the network structure of the second neural network is mirror-symmetric to the network structure of the first neural network;

[0027] Determining a second loss situation corresponding to the i-th fourth training sample according to the third sampling data of the i-th fourth training sample and the j-th restored data;

[0028] When the second loss situation does not meet the second training stop condition of the first neural network, the parameters of the first neural network and the second neural network are adjusted to obtain the updated first neural network and the second neural network, and j is updated to j+1, and the j-th fourth training sample is input into the first neural network to be trained to obtain the j-th augmented data corresponding to the j-th fourth training sample until the second training stop condition is met.

[0029] According to some embodiments of the present application, determining the second loss situation corresponding to the i-th fourth training sample according to the third sampling data of the i-th fourth training sample and the j-th restored data includes:

[0030] Inputting the third sampling data of the i-th fourth training sample and the j-th restored data into a second discriminant neural network to obtain a corresponding second discriminant result;

[0031] According to the second discrimination result, a second discrimination accuracy rate and a second loss value are obtained, wherein the second discrimination accuracy rate represents the similarity between the feature distributions of the plurality of the third sampling data and the plurality of the j-th restored data;

[0032] The second loss situation satisfies the second training stop condition, including that the first discrimination accuracy and the second discrimination accuracy are within a second accuracy range, and the first loss value is less than a first threshold, and the second loss value is less than a second threshold.

[0033] According to some embodiments of the present application, the target nuclear power equipment is a rotating equipment having a bearing, and determining similar equipment to the target nuclear power equipment includes:

[0034] Determine a type of other rotating equipment including bearings as the first similar equipment;

[0035] Acquire the sampled data volume of the first similar device;

[0036] When the sample data volume is greater than or equal to a sample volume threshold, the first similar device is determined as the similar device; wherein the sample volume threshold is obtained based on parameters of the health status assessment model.

[0037] According to the second aspect of the present application, a 5G-based nuclear power equipment data augmentation device includes:

[0038] A first acquisition module is used to acquire a first training sample set; the first training sample set includes a plurality of first training samples of the target nuclear power equipment to be tested for health status, each of the first training samples includes a first sampling data of the target nuclear power equipment, and a first health status category label corresponding to the first sampling data;

[0039] A determination module, used to determine similar equipment to the target nuclear power equipment;

[0040] A second acquisition module is used to acquire a second training sample set; the second training sample set includes multiple second training samples of the similar device, each of the second training samples includes a second sampling data of the similar device, and a second health status category label corresponding to the second sampling data;

[0041] a data augmentation module, configured to convert the second training sample set into a third training sample set; the third training sample set includes a plurality of third training samples, each of the third training samples includes a target augmented data converted from the second sampled data and having the same feature distribution as the first sampled data, and a second health status category label corresponding to the second sampled data;

[0042] A training module is used to train a health status assessment model using the first training sample set and the third training sample set; the health status assessment model is used to assess the health status of the target nuclear power equipment.

[0043] An electronic device according to an embodiment of the third aspect of the present application includes a processor and a memory, wherein the memory stores programs or instructions that can be run on the processor, and when the program or instructions are executed by the processor, the steps of the 5G-based nuclear power equipment data augmentation method as described in any one of the embodiments of the second aspect above are implemented.

[0044] According to the computer-readable storage medium of the fourth aspect embodiment of the present application, computer executable instructions are stored, and the computer executable instructions are used to execute the 5G-based nuclear power equipment data augmentation method as described in the first aspect embodiment above.

[0045] In the embodiment of the present application, by converting similar equipment-related data with sufficient data volume, augmented data with the same characteristic distribution as the target nuclear power equipment-related data is obtained, so that the training data samples of the target nuclear power equipment can be expanded with the help of a small amount of target nuclear power equipment-related data and a large amount of complete data of similar equipment, so as to meet the training sample data requirements of the health status assessment model, and meet the data volume requirements while ensuring data accuracy. In addition, the 5G-based nuclear power equipment data augmentation method of the present application obtains a health status assessment model by training based on the expanded training data samples, and performs 5G-based nuclear power equipment data augmentation. The health status assessment model obtains sufficient training data support, which can achieve better assessment results.

[0046] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent from the description, or may be understood by practicing the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0048] Figure 1 It is a flowchart of an embodiment of the 5G-based nuclear power equipment data augmentation method of the present application;

[0049] Figure 2 is a schematic diagram of the network structure of a first neural network of an embodiment of the 5G-based nuclear power equipment data augmentation method of the present application;

[0050] Figure 3 It is a schematic diagram of the training process of the first neural network of an embodiment of the 5G-based nuclear power equipment data augmentation method of the present application;

[0051] Figure 4 It is a structural schematic diagram of an embodiment of a 5G-based nuclear power equipment data augmentation device of the present application;

[0052] Figure 5 It is a schematic diagram of the hardware structure of an embodiment of the electronic device of the present application. DETAILED DESCRIPTION

[0053] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be understood as limiting the present application.

[0054] In the description of this application, if there is a description of first, second, etc., it is only for the purpose of distinguishing technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0055] In the description of the present application, it should be understood that the descriptions involving orientation, such as the orientation or positional relationship indicated as up, down, etc., are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0056] In the description of this application, it should be noted that, unless otherwise clearly defined, terms such as setting, installing, connecting, etc. should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meaning of the above terms in this application based on the specific content of the technical solution.

[0057] The technical solution of the present application will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described below are only part of the embodiments of the present application, not all of the embodiments.

[0058] Figure 1 This is a flow chart of an embodiment of a method for augmenting nuclear power equipment data based on 5G according to an embodiment of the present application. Figure 1 , the embodiments of the present application are further described.

[0059] like Figure 1 As shown, the embodiment of the present application proposes a 5G-based nuclear power equipment data augmentation method, and the 5G-based nuclear power equipment data augmentation method includes:

[0060] Step 101, obtaining a first training sample set; the first training sample set includes a plurality of first training samples of target nuclear power equipment whose health status is to be detected, each first training sample includes a first sampling data of the target nuclear power equipment, and a first health status category label corresponding to the first sampling data;

[0061] Step 102: determine similar equipment to the target nuclear power equipment;

[0062] Step 103: Obtain a second training sample set; the second training sample set includes a plurality of second training samples of similar devices, each second training sample includes a second sampled data of the similar device, and a second health status category label corresponding to the second sampled data;

[0063] Step 104: convert the second training sample set into a third training sample set; the third training sample set includes a plurality of third training samples, each of which includes a target augmented data converted from the second sampled data and having the same feature distribution as the first sampled data, and a second health status category label corresponding to the second sampled data;

[0064] Step 105: Use the first training sample set and the third training sample set to train a health status assessment model; the health status assessment model is used to assess the health status of the target nuclear power equipment.

[0065] In an embodiment of the present application, by converting similar equipment-related data with sufficient data volume, augmented data with the same characteristic distribution as the target nuclear power equipment-related data is obtained, and the training data samples of the target nuclear power equipment are expanded with the help of a small amount of target nuclear power equipment-related data and a large amount of complete data of similar equipment, so as to meet the training sample data requirements of the health status assessment model, and meet the data volume requirements while ensuring data accuracy. In addition, the 5G-based nuclear power equipment data augmentation method of the present application obtains a health status assessment model by training based on the expanded training data samples, and performs 5G-based nuclear power equipment data augmentation. The health status assessment model obtains sufficient training data support and can achieve better assessment results. In addition, there is no need to carry out the data collection process by building or simulating actual scenarios, which greatly reduces the difficulty of data acquisition, reduces the application cost of data-driven models, and improves data utilization.

[0066] The target nuclear power equipment may be any nuclear power equipment that needs to be evaluated for health status, and the corresponding health status evaluation model is used to evaluate the health status of the nuclear power equipment.

[0067] The target nuclear power equipment may be a rotating device in a nuclear power plant, having a bearing structure, and the bearing of the target nuclear power equipment may be a rolling bearing. Specifically, the target nuclear power equipment may be a motor device, a fan device or a pump device that is currently widely used in nuclear power plants.

[0068] For example, the target nuclear power equipment mentioned above is a motor equipment, and the health status assessment model is used to assess the health status of the motor equipment.

[0069] The above-mentioned similar equipment refers only to a type of rotating equipment with sufficient data, which can be a rotating equipment containing a bearing, specifically, a rotating equipment containing a rolling bearing, or a rolling bearing.

[0070] The first sampling data and the second sampling data may refer to vibration data during the constant power or constant speed operation of the equipment, and the source of the vibration data is collected by a vibration sensor that takes the bearings of the equipment as the measurement object; specifically, the first sampling data may be obtained by collecting nuclear power rotating equipment by a vibration sensor, and the second collected data may be obtained by collecting similar equipment by a vibration sensor.

[0071] The first sampling data or the second sampling data may be a set of vibration data collected by the vibration sensor within a collection duration; the multiple collection data included in a training sample may be multiple sets of vibration data collected multiple times. The relevant parameters of the above collection process may be set according to the equipment conditions to make the collected data more accurate, and the relevant parameters may include collection duration, collection frequency, etc.; the above vibration data may be amplitude data.

[0072] For example, the data acquisition parameters of the target nuclear power equipment may include: acquisition frequency f tar , single acquisition duration t tar，sample , sampling interval time t tar，break , the speed of the target rotating device is w tar , the number of vibrations contained in each first training sample set is

[0073]

[0074] Data collection parameters for similar devices may include: collection frequency f sim , single acquisition duration t sim，sample , sampling interval time t sim，break , the speed of the target rotating device is w sim , the number of vibrations contained in each second training sample set is

[0075] The above health status category label is used to judge the health status of the corresponding device according to the vibration data. Specifically, the above first health status category label corresponds to the first sampling data, and the above second health status category label corresponds to the second sampling data.

[0076] The first training sample set may include a plurality of first training samples, the second training sample set may include a plurality of second training samples, and the first training samples and the second training samples may consist of vibration data and corresponding labels.

[0077] For example, the first training sample and the second training sample, as labeled data samples, can be expressed as (X, Y), where X is the vibration data and Y is the health status category label corresponding to the vibration data, which should be divided into at least three categories: "normal", "degradation", and "fault".

[0078] For example, the data format of the first training sample of the target nuclear power equipment is: tar =(X tar , Y tar );

[0079] in, y i ∈(C1,C2,......,C n ), n≥3;

[0080] Among them, x 1i is the first sampling data of the target nuclear power equipment, including a set of amplitude data collected by the target nuclear power equipment within a collection duration, N tar is the sample size corresponding to the target nuclear power equipment, which can be determined according to the number of target nuclear power equipment and the data collection situation. 1iis the corresponding amplitude data collected, N x，tar is the number of amplitude data; y 1i is the corresponding first health status category label, n is the number of label categories, n can be set to 3, then C1, C2, and C3 are normal, decay, and failure, respectively, that is, y 1i ∈(normal, decay, fault);

[0081] The data format of the second training sample of similar equipment is: D sim =(X sim , Y sim );

[0082] in, y i ∈(C1,C2,......,C n ), n≥3;

[0083] Among them, x 2i is the second sampling data of the similar device, including a set of amplitude data collected by the similar device within a collection duration, N sim is the sample size corresponding to similar devices, which can be determined according to the number of similar devices and the data collection situation. 2i is the corresponding amplitude data collected, N x，sim is the number of amplitude data; y 2i is the corresponding second health status category label, n is the number of label categories, n can be set to 3, then C1, C2, and C3 are normal, decay, and failure, respectively, that is, y 2i ∈(normal, decay, fault).

[0084] The above-mentioned label may be a health status category determined based on vibration data.

[0085] For example, the maximum amplitude of the “normal” label sample is A norm , the amplitude of the "decay" label A d ∈(A norm , 5A norm ], the amplitude A of the "fault" label f >5A norm The amplitude A of the above “decay” label d and the amplitude A of the "fault" label f It can refer to all values ​​of a set of amplitude data contained in a sample of data, or it can refer to the mean or median.

[0086] The amplitude-related data may be absolute values.

[0087] The first training sample set may be a vibration data set with labels related to target nuclear power equipment, and the second training sample set may be a vibration data set with labels related to similar equipment.

[0088] The above-mentioned third training sample set can be an augmented sample set converted from a similar equipment-related data set and having the same characteristic distribution as the target nuclear power equipment-related data set, which ensures the similarity with the target nuclear power equipment-related data set and has the data volume level of the similar equipment-related data set, and can meet the requirements of the training quantity required for the health status assessment model of nuclear power rotating equipment.

[0089] The above-mentioned health status assessment model can be a deep learning model that can be used to perform multi-classification tasks. Specifically, the network layer of the model can be a fully connected network, a convolutional network, a recurrent neural network, etc. The above-mentioned model can be trained by supervised learning, and the training process can be trained by inputting the above-mentioned first training sample set and the above-mentioned third training sample set.

[0090] The above health status assessment model is used to assess the health status of the target nuclear power equipment. The trained health status assessment model can output the assessed health status category of the equipment based on the input vibration data.

[0091] In some implementations, before converting the second training sample set into the third training sample set, the method further includes:

[0092] Acquire a fourth training sample set; the fourth training sample set includes a plurality of fourth training samples, each of which includes a third sampled data of a similar device and fourth sampled data of a target nuclear power device corresponding to the third sampled data;

[0093] Using the fourth training sample set, training the first neural network until a training stop condition of the first neural network is met, thereby obtaining a trained first neural network;

[0094] Converting the second training sample set into the third training sample set includes:

[0095] The second training sample is input into the trained first neural network to obtain a third training sample corresponding to the second training sample.

[0096] In this embodiment, the first neural network can be used to convert the second training sample set into the third training sample set, so as to convert the similar equipment-related data set into an augmented data set having the same feature distribution as the target nuclear power equipment-related data set, thereby better realizing data augmentation through neural network technology.

[0097] The above-mentioned third sampling data can be a part of multiple second sampling data in the second training sample set, and the above-mentioned fourth sampling data can be a part of multiple first sampling data in the first training sample set; the above-mentioned fourth sampling data corresponding to the third sampling data can be understood as the corresponding third sampling data and the fourth sampling data corresponding to the same health status category label.

[0098] The fourth training sample set mentioned above can be understood as including relevant vibration data of similar equipment and relevant vibration data of target nuclear power equipment. A plurality of fourth training samples are correspondingly formed according to label conditions and can be used to train the first neural network.

[0099] In some implementations, using the fourth training sample set to train the first neural network until a training stop condition of the first neural network is met, thereby obtaining a trained first neural network, comprises:

[0100] For the i-th fourth training sample in the fourth training sample set, input the i-th fourth training sample into the first neural network to obtain the i-th augmented data corresponding to the i-th fourth training sample; wherein i is a positive integer;

[0101] Determine a first loss situation corresponding to the i-th fourth training sample according to the fourth sampling data and the i-th augmented data of the i-th fourth training sample;

[0102] When the first loss situation does not meet the first training stop condition of the first neural network, the parameters of the first neural network are adjusted to obtain an updated first neural network, and i is updated to i+1, and the second training sample set is returned to the first neural network until the first training stop condition is met.

[0103] In this implementation, the first neural network is trained by inputting the fourth training sample set, and the parameters are continuously adjusted and the first neural network is updated according to the loss situation until the training stop condition is met, so as to obtain a trained first neural network. Updating the neural network using the loss situation can improve the accuracy and robustness of the model to a certain extent, and can also adapt to different training environments and data types, and has good generalization ability and practicality.

[0104] like Figure 2 As shown, in some embodiments, the first neural network G1 (θ g1 The network structure of ) can be as follows: It includes three main network structures: input layer, hidden layer, and output layer. The number of input layers is 1, and the number of neurons is the same as the number of vibrations of similar equipment data samples N. x，sim Same, number of hidden layers Lh∈[4,8], Lh∈N + , the number of neurons in the hidden layer is unlimited, the number of output layers is 1, and the number of neurons is the same as the number of generated augmented data samples Nx，tar same.

[0105] In some embodiments, the first neural network G1 (θ g1 ) can be a fully connected neural network, and the neural network activation function is act g1 =max(0.01x, x), where x is the output result of the neuron linear operation.

[0106] In some implementations, determining a first loss situation corresponding to the i-th fourth training sample according to the fourth sampling data and the i-th augmented data of the i-th fourth training sample includes:

[0107] Inputting the fourth sampling data and the i-th augmented data of the i-th fourth training sample into the first discriminant neural network to obtain a corresponding first discriminant result;

[0108] According to the first discrimination result, a first discrimination accuracy rate and a first loss value are obtained, wherein the first discrimination accuracy rate represents the similarity between the feature distributions of the plurality of fourth sampling data and the plurality of i-th augmented data;

[0109] The first loss condition satisfies a first training stop condition, including that the first discrimination accuracy is within a first accuracy range.

[0110] In this implementation, the loss situation is obtained through a discriminant neural network. The use of a discriminant neural network not only provides a clear optimization goal and an efficient optimization method, but also has good adaptability and flexibility, which helps to improve the performance and generalization ability of the model.

[0111] The fourth sampling data mentioned above is the relevant vibration data of the target nuclear power equipment, and the i-th augmented data is the augmented data output after the first neural network converts the relevant vibration data of the input similar equipment.

[0112] The first discriminant neural network can determine the similarity of the characteristic distribution between the relevant vibration data of the target nuclear power equipment and the generated augmented data, and further determine the loss of the first neural network.

[0113] The first discriminant neural network D1(θ d1 ) can identify the generated augmented data D tar，gen Compared with the actual target nuclear power equipment data D tar Whether they have the same feature distribution; the output result of the network, that is, the first discrimination result can be a binary classification value (0, 1). If the distribution is similar, the output is 0, and if the distribution is very different, the output is 1. The first discriminant neural network D1(θ d1), specifically, the loss function is a cross entropy loss function, and further, the loss function is a binary cross entropy loss function, and then the first discrimination accuracy and the first loss value can be obtained, wherein the first discrimination accuracy is the binary classification discrimination accuracy;

[0114] The first training stop condition may be that the training is stopped when the first discrimination accuracy rate reaches a certain accuracy rate range. Specifically, the first accuracy rate range may be within a range greater than 90%.

[0115] In some embodiments, when the first loss situation does not satisfy the first training stop condition of the first neural network, adjusting the parameters of the first neural network to obtain the updated first neural network can be performed by updating the parameters of the first neural network through an Adam gradient descent algorithm.

[0116] like Figure 3 As shown, in some embodiments, when the target loss value satisfies the first training stop condition of the first neural network, the first neural network is trained using the fourth training sample set until the training stop condition of the first neural network is met to obtain a trained first neural network, further comprising:

[0117] For the jth fourth training sample in the fourth training sample set, input the jth fourth training sample into the first neural network to be trained to obtain the jth augmented data corresponding to the jth fourth training sample; wherein j is a positive integer;

[0118] Inputting the j-th augmented data into the second neural network to obtain the j-th restored data of the similar device; wherein the network structure of the second neural network is mirror-symmetric to the network structure of the first neural network;

[0119] Determine a second loss situation corresponding to the i-th fourth training sample according to the third sampling data and the j-th restored data of the i-th fourth training sample;

[0120] When the second loss situation does not meet the second training stop condition of the first neural network, the parameters of the first neural network and the second neural network are adjusted to obtain updated first neural network and second neural network, and j is updated to j+1, and the j-th fourth training sample is input into the first neural network to be trained to obtain the j-th augmented data corresponding to the j-th fourth training sample until the second training stop condition is met.

[0121] In this embodiment, by building a second neural network D2 (θ d2 ), the generated augmented data D sim,gen Restore to restore data similar to similar device data, that is, D sim,gen =G2(D tar，gen ,θg2 ), where the second neural network G2(θ g2 ) and the first neural network G1(θ g1 ) mirror symmetry; then by building the second discriminant neural network D2 (θ d2 ), determine the restored data D sim,gen Similar to the actual equipment data D sim Whether they have the same feature distribution. It further provides a clear optimization goal and an efficient optimization method for the first neural network, and has good adaptability and flexibility, which helps to improve the performance and generalization ability of the model.

[0122] In some implementations, determining a second loss situation corresponding to the i-th fourth training sample according to the third sampling data and the j-th restored data of the i-th fourth training sample includes:

[0123] Inputting the third sampling data and the jth restored data of the i-th fourth training sample into the second discriminant neural network to obtain a corresponding second discriminant result;

[0124] According to the second discrimination result, a second discrimination accuracy rate and a second loss value are obtained, wherein the second discrimination accuracy rate represents the similarity of feature distribution between the plurality of third sampling data and the plurality of j-th restored data;

[0125] Among them, the second loss situation satisfies the second training stop condition, including that the first discrimination accuracy and the second discrimination accuracy are within the second accuracy range, and the first loss value is less than the first threshold, and the second loss value is less than the second threshold.

[0126] In this implementation, the second loss situation satisfies the second training stop condition and needs to simultaneously meet the requirements of the discrimination accuracy and the loss value, thereby further achieving better training of the model.

[0127] The output result of the second discriminant neural network can be a binary classification value (0, 1). If the distribution is similar, the output is 0, and if the distribution is very different, the output is 1. Calculate the neural network D2(θ d2 ), specifically, the loss function is the cross entropy loss function, further, the loss function is the binary cross entropy loss function, and then the second discrimination accuracy and the second loss value can be obtained, where the second discrimination accuracy is the binary classification discrimination accuracy.

[0128] The second training stop condition includes that the first discrimination accuracy and the second discrimination accuracy are within the second accuracy range, and the first loss value is less than the first threshold value, and the second loss value is less than the second threshold value, which can be the first discriminant neural network D1 (θ d1 )’s first discriminant accuracy and the second discriminant neural network D2(θd2 ) are all greater than or equal to 95% and satisfy the following formula:

[0129]

[0130] Among them A tar,min is the minimum vibration amplitude of the target device data used for training, A sim,min is the minimum vibration amplitude of similar equipment data used in training.

[0131] In some embodiments, when the second loss condition does not satisfy the second training stop condition of the first neural network, adjusting the parameters of the first neural network and the second neural network to obtain the updated first neural network and the second neural network can be performed by respectively updating the first neural network G1(θ g1 ) and the second neural network G2(θ g2 ) parameters.

[0132] In some embodiments, the first discriminant neural network D1 (θ d1 ) and the second discriminant neural network D2(θ d2 ) is a nonlinear classifier network. The basic structure of the network can be a fully connected network or a convolutional network. The training process of both is supervised training. The initial training termination condition is that the binary classification accuracy of the neural network obtained by 10 consecutive test set tests is in the range of 75%-85%, where the test set is divided from the data sample set corresponding to the training.

[0133] In some embodiments, the target nuclear power equipment is a rotating equipment having a bearing, and determining similar equipment to the target nuclear power equipment includes:

[0134] Determine a type of other rotating equipment including bearings as the first similar equipment;

[0135] Obtaining the amount of sampled data of the first similar device;

[0136] When the amount of sampled data is greater than or equal to a sample size threshold, the first similar device is determined as a similar device; wherein the sample size threshold is obtained based on parameters of the health status assessment model.

[0137] In this embodiment, other rotating devices with sufficient sampling data volume are selected to be determined as similar devices, which can provide a sufficient sample volume basis for the data augmentation process. By selecting a type of other rotating devices that can provide a sufficient amount of sampling data as similar devices and using the data of such devices, rather than using data from multiple types of other rotating devices to piece together to meet the sampling data volume requirement, the data of the similar devices provided can be of higher quality, thereby enabling data augmentation to achieve better results.

[0138] The bearing mentioned above may be a rolling bearing.

[0139] The above sampled data volume is greater than or equal to the sample volume threshold, which is shown in the following formula:

[0140]

[0141] Where n is the number of categories of the evaluation task, that is, the number of health status categories, e is the minimum achievable error of the health status evaluation model, e∈(0, 0.1), is the estimated discrimination accuracy of the health status assessment model during the training process, and a is the learning rate of the iterative training process of the health status assessment model, a∈(0.01,1), d is the learning decay rate, d∈(-1,0), and the relevant parameters of the above health status assessment model can be obtained in advance.

[0142] The 5G-based nuclear power equipment data augmentation method provided in the embodiment of the present application may be executed by a 5G-based nuclear power equipment data augmentation device 200. In the embodiment of the present application, the 5G-based nuclear power equipment data augmentation method performed by the 5G-based nuclear power equipment data augmentation device 200 is taken as an example to illustrate the 5G-based nuclear power equipment data augmentation device 200 provided in the embodiment of the present application.

[0143] See also Figure 4 , is a schematic diagram of a 5G-based nuclear power equipment data augmentation device 200 provided in an embodiment of the present application. Figure 4 As shown, the 5G-based nuclear power equipment data augmentation device 200 includes:

[0144] The first acquisition module 201 is used to acquire a first training sample set; the first training sample set includes a plurality of first training samples of the target nuclear power equipment to be tested for health status, each first training sample includes a first sampling data of the target nuclear power equipment, and a first health status category label corresponding to the first sampling data;

[0145] A determination module 202, used to determine similar equipment to the target nuclear power equipment;

[0146] The second acquisition module 203 is used to acquire a second training sample set; the second training sample set includes multiple second training samples of similar devices, each second training sample includes a second sampled data of the similar device, and a second health status category label corresponding to the second sampled data;

[0147] The data augmentation module 204 is used to convert the second training sample set into a third training sample set; the third training sample set includes a plurality of third training samples, each of which includes a target augmented data converted from the second sampled data and having the same feature distribution as the first sampled data, and a second health status category label corresponding to the second sampled data;

[0148] The training module 205 is used to train a health status assessment model using the first training sample set and the third training sample set; the health status assessment model is used to assess the health status of the target nuclear power equipment.

[0149] In some implementations, the data augmentation module 204 may be used to:

[0150] Acquire a fourth training sample set; the fourth training sample set includes a plurality of fourth training samples, each of which includes a third sampled data of a similar device and fourth sampled data of a target nuclear power device corresponding to the third sampled data;

[0151] Using the fourth training sample set, training the first neural network until a training stop condition of the first neural network is met, thereby obtaining a trained first neural network;

[0152] Converting the second training sample set into the third training sample set includes:

[0153] The second training sample is input into the trained first neural network to obtain a third training sample corresponding to the second training sample.

[0154] In some implementations, the data augmentation module 204 may be used to:

[0155] For the i-th fourth training sample in the fourth training sample set, input the i-th fourth training sample into the first neural network to obtain the i-th augmented data corresponding to the i-th fourth training sample; wherein i is a positive integer;

[0156] Determine a first loss situation corresponding to the i-th fourth training sample according to the fourth sampling data and the i-th augmented data of the i-th fourth training sample;

[0157] When the first loss situation does not meet the first training stop condition of the first neural network, the parameters of the first neural network are adjusted to obtain an updated first neural network, and i is updated to i+1, and the second training sample set is returned to the first neural network until the first training stop condition is met.

[0158] In some implementations, the data augmentation module 204 may be used to:

[0159] Inputting the fourth sampling data and the i-th augmented data of the i-th fourth training sample into the first discriminant neural network to obtain a corresponding first discriminant result;

[0160] According to the first discrimination result, a first discrimination accuracy rate and a first loss value are obtained, wherein the first discrimination accuracy rate represents the similarity between the feature distributions of the plurality of fourth sampling data and the plurality of i-th augmented data;

[0161] The first loss condition satisfies a first training stop condition, including that the first discrimination accuracy is within a first accuracy range.

[0162] In some implementations, the data augmentation module 204 may also be used to:

[0163] For the jth fourth training sample in the fourth training sample set, input the jth fourth training sample into the first neural network to be trained to obtain the jth augmented data corresponding to the jth fourth training sample; wherein j is a positive integer;

[0164] Inputting the j-th augmented data into the second neural network to obtain the j-th restored data of the similar device; wherein the network structure of the second neural network is mirror-symmetric to the network structure of the first neural network;

[0165] Determine a second loss situation corresponding to the i-th fourth training sample according to the third sampling data and the j-th restored data of the i-th fourth training sample;

[0166] When the second loss situation does not meet the second training stop condition of the first neural network, the parameters of the first neural network and the second neural network are adjusted to obtain updated first neural network and second neural network, and j is updated to j+1, and the j-th fourth training sample is input into the first neural network to be trained to obtain the j-th augmented data corresponding to the j-th fourth training sample until the second training stop condition is met.

[0167] In some implementations, the data augmentation module 204 may be used to:

[0168] Inputting the third sampling data and the jth restored data of the i-th fourth training sample into the second discriminant neural network to obtain a corresponding second discriminant result;

[0169] According to the second discrimination result, a second discrimination accuracy rate and a second loss value are obtained, wherein the second discrimination accuracy rate represents the similarity of feature distribution between the plurality of third sampling data and the plurality of j-th restored data;

[0170] Among them, the second loss situation satisfies the second training stop condition, including that the first discrimination accuracy and the second discrimination accuracy are within the second accuracy range, and the first loss value is less than the first threshold, and the second loss value is less than the second threshold.

[0171] In some implementations, the determination module 202 may be configured to:

[0172] Determine a type of other rotating equipment including bearings as the first similar equipment;

[0173] Obtaining the amount of sampled data of the first similar device;

[0174] When the amount of sampled data is greater than or equal to a sample size threshold, the first similar device is determined as a similar device; wherein the sample size threshold is obtained based on parameters of the health status assessment model.

[0175] Since the 5G-based nuclear power equipment data augmentation device 200 adopts all the technical solutions of the 5G-based nuclear power equipment data augmentation method of the above-mentioned embodiment, it at least has all the beneficial effects brought by the technical solutions of the above-mentioned embodiment, which will not be described in detail here.

[0176] Figure 5 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present application.

[0177] The electronic device may include a processor 301 and a memory 302 storing computer program instructions.

[0178] Specifically, the processor 301 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0179] The memory 302 may include a large capacity memory for data or instructions. By way of example and not limitation, the memory 302 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In appropriate cases, the memory 302 may include a removable or non-removable (or fixed) medium. In appropriate cases, the memory 302 may be inside or outside the integrated gateway disaster recovery device. In a specific embodiment, the memory 302 is a non-volatile solid-state memory.

[0180] In some embodiments, the memory 302 may include a read-only memory (ROM), a random access memory (RAM), a magnetic disk storage medium device, an optical storage medium device, a flash memory device, an electrical, optical or other physical / tangible memory storage device. Thus, in general, the memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., a memory device) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to an aspect of the present disclosure.

[0181] The processor 301 reads and executes computer program instructions stored in the memory 302 to implement any one of the 5G-based nuclear power equipment data augmentation methods in the above embodiments.

[0182] In one example, the electronic device may further include a communication interface 303 and a bus 310. Figure 5 As shown, the processor 301, the memory 302, and the communication interface 303 are connected via a bus 310 and communicate with each other.

[0183] The communication interface 303 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiments of the present application.

[0184] Bus 310 includes hardware, software or both, and the parts of online data flow billing equipment are coupled to each other. For example, but not limitation, bus may include accelerated graphics port (AGP) or other graphics bus, enhanced industrial standard architecture (EISA) bus, front-end bus (FSB), hypertransport (HT) interconnection, industrial standard architecture (ISA) bus, infinite bandwidth interconnection, low pin count (LPC) bus, memory bus, micro channel architecture (MCA) bus, peripheral component interconnection (PCI) bus, PCI-Express (PCI-X) bus, serial advanced technology attachment (SATA) bus, video electronics standard association local (VLB) bus or other suitable bus or two or more of these combinations. In appropriate cases, bus 310 may include one or more buses. Although the present application embodiment describes and shows a specific bus, the present application considers any suitable bus or interconnection.

[0185] The electronic device can execute the 5G-based nuclear power equipment data augmentation method in the embodiment of the present application, thereby realizing the combination Figure 1 and Figure 4 The present invention describes a 5G-based nuclear power equipment data augmentation method and device.

[0186] In addition, in combination with the 5G-based nuclear power equipment data augmentation method in the above embodiment, the embodiment of the present application can provide a computer storage medium for implementation. The computer storage medium stores computer program instructions; when the computer program instructions are executed by the processor, any one of the 5G-based nuclear power equipment data augmentation methods in the above embodiment is implemented.

[0187] It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.

[0188] The functional blocks shown in the above-described block diagram can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present application are programs or code segments that are used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link by a data signal carried in a carrier. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.

[0189] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, this application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiment, or in a different order from the embodiment, or several steps can be performed simultaneously.

[0190] Aspects of the present disclosure are described above with reference to the flowchart and / or block diagram of the method, device (system) and computer program product according to the embodiment of the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device to produce a machine so that these instructions executed by the processor of the computer or other programmable data processing device enable the implementation of the function / action specified in one or more boxes of the flowchart and / or block diagram. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field programmable logic circuit. It can also be understood that each box in the block diagram and / or flowchart and the combination of boxes in the block diagram and / or flowchart can also be implemented by dedicated hardware that performs a specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions.

[0191] The above is only a specific implementation of the present application. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. It should be understood that the protection scope of the present application is not limited to this. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in this application, and these modifications or replacements should be included in the protection scope of this application.

Claims

1. A 5G-based nuclear power equipment data augmentation method, characterized in that: include: Obtaining a first training sample set; The first training sample set includes a plurality of first training samples of the target nuclear power equipment whose health status is to be detected, each of the first training samples includes a first sampling data of the target nuclear power equipment and a first health status category label corresponding to the first sampling data; Identify similar equipment to the target nuclear power equipment; Acquire a second training sample set; the second training sample set includes a plurality of second training samples of the similar device, each of the second training samples includes a second sampling data of the similar device, and a second health status category label corresponding to the second sampling data; Converting the second training sample set into a third training sample set; the third training sample set includes a plurality of third training samples, each of the third training samples includes a target augmented data converted from the second sampled data and having the same feature distribution as the first sampled data, and a second health status category label corresponding to the second sampled data; The first training sample set and the third training sample set are used to train a health status assessment model; the health status assessment model is used to assess the health status of the target nuclear power equipment.

2. The method according to claim 1, characterized in that Before converting the second training sample set into a third training sample set, the method further includes: Acquire a fourth training sample set; the fourth training sample set includes a plurality of fourth training samples, each of the fourth training samples includes a third sampling data of the similar device and fourth sampling data of the target nuclear power device corresponding to the third sampling data; Using the fourth training sample set, training the first neural network until a training stop condition of the first neural network is met, thereby obtaining a trained first neural network; The converting the second training sample set into a third training sample set comprises: The second training sample is input into the trained first neural network to obtain a third training sample corresponding to the second training sample.

3. The method according to claim 2, characterized in that The step of training the first neural network by using the fourth training sample set until a training stop condition of the first neural network is met to obtain the trained first neural network includes: For the i-th fourth training sample in the fourth training sample set, input the i-th fourth training sample into the first neural network to obtain the i-th augmented data corresponding to the i-th fourth training sample; wherein i is a positive integer; Determine a first loss situation corresponding to the i-th fourth training sample according to the fourth sampling data of the i-th fourth training sample and the i-th augmented data; When the first loss situation does not meet the first training stop condition of the first neural network, adjust the parameters of the first neural network to obtain the updated first neural network, update i to i+1, and return to input the second training sample set into the first neural network until the first training stop condition is met.

4. The method according to claim 3, characterized in that The determining, according to the fourth sampling data of the i-th fourth training sample and the i-th augmented data, a first loss situation corresponding to the i-th fourth training sample comprises: Inputting the fourth sampling data of the i-th fourth training sample and the i-th augmented data into a first discriminant neural network to obtain a corresponding first discriminant result; According to the first discrimination result, a first discrimination accuracy rate and a first loss value are obtained, wherein the first discrimination accuracy rate represents the similarity of feature distribution between the plurality of the fourth sampling data and the plurality of the i-th augmented data; The first loss condition satisfies the first training stop condition, including that the first discrimination accuracy is within a first accuracy range.

5. The method according to claim 4, characterized in that When the target loss value satisfies the first training stop condition of the first neural network, the first neural network is trained by using the fourth training sample set until the training stop condition of the first neural network is satisfied to obtain the trained first neural network, and further includes: For the jth fourth training sample in the fourth training sample set, input the jth fourth training sample into the first neural network to be trained to obtain the jth augmented data corresponding to the jth fourth training sample; wherein j is a positive integer; Inputting the j-th augmented data into a second neural network to obtain the j-th restored data of the similar device; wherein the network structure of the second neural network is mirror-symmetric to the network structure of the first neural network; Determining a second loss situation corresponding to the i-th fourth training sample according to the third sampling data of the i-th fourth training sample and the j-th restored data; When the second loss situation does not meet the second training stop condition of the first neural network, the parameters of the first neural network and the second neural network are adjusted to obtain the updated first neural network and the second neural network, and j is updated to j+1, and the j-th fourth training sample is returned to the first neural network to be trained to obtain the j-th augmented data corresponding to the j-th fourth training sample, until the second training stop condition is met.

6. The method according to claim 5, characterized in that The determining, according to the third sampling data of the i-th fourth training sample and the j-th restored data, a second loss situation corresponding to the i-th fourth training sample comprises: Inputting the third sampling data of the i-th fourth training sample and the j-th restored data into a second discriminant neural network to obtain a corresponding second discriminant result; According to the second discrimination result, a second discrimination accuracy rate and a second loss value are obtained, wherein the second discrimination accuracy rate represents the similarity between the feature distributions of the plurality of the third sampling data and the plurality of the j-th restored data; The second loss situation satisfies the second training stop condition, including that the first discrimination accuracy and the second discrimination accuracy are within a second accuracy range, and the first loss value is less than a first threshold, and the second loss value is less than a second threshold.

7. The method according to claim 1, characterized in that The target nuclear power equipment is a rotating equipment having a bearing, and the step of determining similar equipment to the target nuclear power equipment includes: Determine a type of other rotating equipment including bearings as the first similar equipment; Acquire the sampled data volume of the first similar device; When the sample data volume is greater than or equal to a sample volume threshold, the first similar device is determined as the similar device; wherein the sample volume threshold is obtained based on parameters of the health status assessment model.

8. A 5G-based nuclear power equipment data augmentation device, characterized in that: include: A first acquisition module, used to acquire a first training sample set; The first training sample set includes a plurality of first training samples of the target nuclear power equipment whose health status is to be detected, each of the first training samples includes a first sampling data of the target nuclear power equipment and a first health status category label corresponding to the first sampling data; A determination module, used to determine similar equipment to the target nuclear power equipment; A second acquisition module is used to acquire a second training sample set; the second training sample set includes multiple second training samples of the similar device, each of the second training samples includes a second sampling data of the similar device, and a second health status category label corresponding to the second sampling data; a data augmentation module, configured to convert the second training sample set into a third training sample set; the third training sample set includes a plurality of third training samples, each of the third training samples includes a target augmented data converted from the second sampled data and having the same feature distribution as the first sampled data, and a second health status category label corresponding to the second sampled data; A training module is used to train a health status assessment model using the first training sample set and the third training sample set; the health status assessment model is used to assess the health status of the target nuclear power equipment.

9. An electronic device, characterized in that: It includes a processor and a memory, the memory stores a program or instruction that can be run on the processor, and when the program or instruction is executed by the processor, the steps of the 5G-based nuclear power equipment data augmentation method as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the 5G-based nuclear power equipment data augmentation method as described in any one of claims 1 to 7.

Images