Multi-virtual-source fusion nuclear power circulating pump unit health state identification method and device

Abstract

The invention provides a multi-virtual-source fusion nuclear power circulating pump unit health state identification method and device, and the method comprises the following steps: obtaining an acceleration signal of a nuclear power circulating pump unit, preprocessing the acceleration signal through kernel independent component analysis to obtain a motor virtual vibration source signal, a pump virtual vibration source signal and a planetary gearbox virtual vibration source signal; and inputting the motor virtual vibration source signal, the pump virtual vibration source signal and the planetary gearbox virtual vibration source signal into a multi-source fusion health recognition model to obtain the health state of the nuclear power circulating pump unit. According to the scheme, the acceleration signal of the nuclear power circulating pump unit is analyzed through improved nuclear independent component analysis, fault features in the original vibration signal are better restored, and the accuracy of detecting the health state of the nuclear power circulating pump unit is improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence, and in particular to a method and device for identifying the health status of a nuclear power circulating pump unit by integrating multiple virtual sources. Background Art

[0002] The nuclear power circulating pump unit is a key equipment in the conventional island circulating water cooling system of a nuclear power plant. Its main function is to cool the exhaust steam discharged by the turbine. If a pump unit fails during operation, the turbine unit will be reduced in load by 50%. In severe cases, it will cause the turbine to shut down or even the reactor to shut down. The nuclear power circulating pump unit has a complex structure and many subsystems. The potential fault location during its operation may occur in any component of its subsystem. The existing technology only uses the signal of a single sensor to perceive and identify the fault characteristics of other subsystems far away from its installation location. Therefore, it is very important to carry out health status identification of nuclear power circulating pump units using multi-sensor signals.

[0003] Nuclear power circulating pump units have been operating in a fault-free state for a long time, resulting in an extreme lack of effective fault data that can be used for fault identification; at the same time, the drive and transmission structure of nuclear power circulating pump units is complex, and the vibrations of the high-voltage motor, planetary gearbox and circulating seawater pump are coupled with each other, resulting in that when a sub-component fails, the fault characteristic frequency is easily interfered by the frequencies of other vibration sources, which poses a major challenge to the health status identification of nuclear power circulating pump units.

[0004] In summary, how to identify the health status of nuclear power circulating pump units through virtual source data under data imbalance is an urgent problem to be solved. Summary of the invention

[0005] The embodiment of the present application provides a method and device for identifying the health status of a nuclear power circulating pump unit by fusion of multiple virtual sources. The acceleration signal of the nuclear power circulating pump unit is analyzed by improved nuclear independent component analysis, so as to better restore the fault characteristics in the original vibration signal and improve the accuracy of detecting the health status of the nuclear power circulating pump unit.

[0006] In a first aspect, an embodiment of the present application provides a method for identifying the health status of a nuclear power circulating pump unit by fusion of multiple virtual sources, the method comprising: The acceleration signal of the nuclear power circulating pump unit is obtained, and the acceleration signal is preprocessed by nuclear independent component analysis to obtain the motor virtual vibration source signal, the pump virtual vibration source signal and the planetary gearbox virtual vibration source signal, wherein the preprocessing method based on nuclear independent component analysis is: Mapping the acceleration signal to a high-dimensional linear space to obtain a mapping vector, constructing a motor reference signal, a pump reference signal and a planetary gearbox reference signal, and obtaining the item with the highest frequency similarity to the motor reference signal in the mapping vector based on the constructed threshold function as a motor virtual vibration source signal, obtaining the item with the highest frequency similarity to the pump reference signal as a pump virtual vibration source signal, and obtaining the item with the highest frequency similarity to the planetary gearbox reference signal as a planetary gearbox virtual vibration source signal; The virtual vibration source signals of the motor, the pump and the planetary gearbox are input into the multi-source fusion health identification model to obtain the health status of the nuclear power circulating pump unit.

[0007] In a second aspect, an embodiment of the present application provides a health status identification device for a nuclear power circulating pump unit with multi-virtual source fusion, comprising: The acquisition device is used to acquire the acceleration signal of the nuclear power circulating pump unit, and pre-process the acceleration signal through nuclear independent component analysis to obtain the motor virtual vibration source signal, the pump virtual vibration source signal and the planetary gearbox virtual vibration source signal, wherein the pre-processing method based on nuclear independent component analysis is: Mapping the acceleration signal to a high-dimensional linear space to obtain a mapping vector, constructing a motor reference signal, a pump reference signal and a planetary gearbox reference signal, and obtaining the item with the highest frequency similarity to the motor reference signal in the mapping vector based on the constructed threshold function as a motor virtual vibration source signal, obtaining the item with the highest frequency similarity to the pump reference signal as a pump virtual vibration source signal, and obtaining the item with the highest frequency similarity to the planetary gearbox reference signal as a planetary gearbox virtual vibration source signal; The health status identification device is used to input the motor virtual vibration source signal, the pump virtual vibration source signal and the planetary gearbox virtual vibration source signal into the multi-source fusion health identification model to obtain the health status of the nuclear power circulating pump unit.

[0008] In a third aspect, an embodiment of the present application provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute a method for identifying the health status of a nuclear power circulating pump unit with multi-virtual source fusion.

[0009] In a fourth aspect, an embodiment of the present application provides a readable storage medium, in which a computer program is stored. The computer program includes a program code for controlling a process to execute a process, and the process includes a method for identifying the health status of a nuclear power circulating pump unit with multi-virtual source fusion.

[0010] The main contributions and innovations of the present invention are as follows: The embodiment of the present application uses an improved kernel independent component analysis algorithm to efficiently separate the motor virtual vibration source signal, pump virtual vibration source signal and planetary gearbox virtual vibration source signal in the nuclear power circulating pump unit, which helps to more accurately obtain the operating status information of different components and provides a more reliable basis for subsequent fault diagnosis; the embodiment of the present application constructs a multi-source fusion recognition model, fully integrating the information of the motor virtual vibration source signal, pump virtual vibration source signal and planetary gearbox virtual vibration source signal, thereby improving the accuracy of judging the health status of the nuclear power circulating pump unit; this scheme uses a label smoothing regularization loss function to complete the training of the multi-source fusion recognition model, ensuring that the network overfitting problem can be effectively solved in the case of data imbalance, improving the generalization ability of the model, thereby improving the accuracy of single sensor signal fault classification, and making the model more reliable and practical in practical applications.

[0011] Details of one or more embodiments of the present application are set forth in the following drawings and description to make other features, objects, and advantages of the present application more readily apparent. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings: Figure 1 It is a flow chart of a method for identifying the health status of a nuclear power circulating pump unit by integrating multiple virtual sources according to an embodiment of the present application; Figure 2 is a structural schematic diagram of a multi-source fusion health recognition model according to an embodiment of the present application; Figure 3 is a structural schematic diagram of a convolution block according to an embodiment of the present application; Figure 4 is a structural schematic diagram of a feature fusion layer according to an embodiment of the present application; Figure 5 is a structural schematic diagram of a global-local feature extraction structure according to an embodiment of the present application; Figure 6 It is a structural block diagram of a health status identification device for a nuclear power circulating pump unit with multi-virtual source fusion according to an embodiment of the present application; Figure 7 It is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0013] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with one or more embodiments of this specification. Instead, they are merely examples of devices and methods consistent with some aspects of one or more embodiments of this specification as detailed in the appended claims.

[0014] It should be noted that: in other embodiments, the steps of the corresponding method are not necessarily performed in the order shown and described in this specification. In some other embodiments, the steps included in the method may be more or less than those described in this specification. In addition, a single step described in this specification may be decomposed into multiple steps for description in other embodiments; and multiple steps described in this specification may be combined into a single step for description in other embodiments.

[0015] Embodiment 1 The embodiment of the present application provides a method for identifying the health status of a nuclear power circulating pump unit by fusion of multiple virtual sources. The acceleration signal of the nuclear power circulating pump unit is analyzed by improved nuclear independent component analysis, so as to better restore the fault characteristics in the original vibration signal and improve the accuracy of detecting the health status of the nuclear power circulating pump unit. Specifically, reference is made to Figure 1 , the method comprising: The acceleration signal of the nuclear power circulating pump unit is obtained, and the acceleration signal is preprocessed by nuclear independent component analysis to obtain the motor virtual vibration source signal, the pump virtual vibration source signal and the planetary gearbox virtual vibration source signal, wherein the preprocessing method based on nuclear independent component analysis is: Mapping the acceleration signal to a high-dimensional linear space to obtain a mapping vector, constructing a motor reference signal, a pump reference signal and a planetary gearbox reference signal, and obtaining the item with the highest frequency similarity to the motor reference signal in the mapping vector based on the constructed threshold function as a motor virtual vibration source signal, obtaining the item with the highest frequency similarity to the pump reference signal as a pump virtual vibration source signal, and obtaining the item with the highest frequency similarity to the planetary gearbox reference signal as a planetary gearbox virtual vibration source signal; The virtual vibration source signals of the motor, the pump and the planetary gearbox are input into the multi-source fusion health identification model to obtain the health status of the nuclear power circulating pump unit.

[0016] In this solution, acceleration sensors are arranged at the motor end and the pump end of the nuclear power circulating pump unit to obtain the acceleration signal of the nuclear power circulating pump unit.

[0017] Specifically, the vibration signal of the nuclear power circulating pump unit during operation is obtained as the acceleration signal through the acceleration sensors arranged at the motor end and the pump end, and the acceleration signal is saved in a sequence.

[0018] Specifically, the acceleration signal is expressed as ,in, Represents the data points obtained at different times when the nuclear power circulating pump unit is running.

[0019] Furthermore, in nuclear power circulating pumps, there are differences in vibration characteristics, working conditions, etc. at the motor end and the pump end during the operation of the unit. Therefore, this solution sets an acceleration sensor at the motor end and the pump end to independently obtain the operating information of the motor and the pump, including vibration amplitude, frequency, etc., so as to more comprehensively reflect the working conditions of the motor and the pump, which is helpful to detect potential faults in a timely manner.

[0020] In some embodiments, the acceleration signal is mapped to a high-dimensional linear space based on a nonlinear kernel function to obtain a mapping vector.

[0021] Specifically, a nonlinear kernel function in a polynomial form is constructed, and the kernel function value of any data point in the acceleration signal and other data points is calculated based on the nonlinear kernel function, so as to map the acceleration signal to a high-dimensional linear space to obtain a mapping vector.

[0022] Specifically, the formula of the nonlinear kernel function is expressed as:

[0023] Where a and b are constants, d is the polynomial order, , is the acceleration signal The operation rule of the nonlinear kernel function is: calculate any data point in the acceleration signal With other data points The kernel function value between .

[0024] Specifically, by constructing a nonlinear kernel function to convert the acceleration signal into a mapping vector, it is more convenient to obtain high-dimensional information from the mapping vector, which helps to separate multiple vibration source signals.

[0025] In some embodiments, the motor shaft speed signal and the pump shaft speed signal in the nuclear power circulating pump unit are obtained by an eddy current sensor, and a square wave pulse sequence is constructed based on the motor shaft speed signal to obtain a motor reference signal. A square wave pulse sequence is constructed based on the pump shaft speed signal to obtain a pump reference signal, and the meshing frequency of the planetary gearbox is used as the planetary gearbox reference signal.

[0026] Specifically, the motor shaft speed signal and the pump shaft speed signal obtained are speed signals in the form of TTL level, and then the corresponding rotation frequency can be calculated, and then the square wave pulse sequence can be constructed using the rotation frequency and the square wave function of Matlab. The meshing frequency of the planetary gearbox can be calculated according to the theoretical calculation formula of the meshing frequency of the planetary gearbox, which is expressed as follows:

[0027] in, is the motor shaft speed, is the number of teeth on the inner ring gear, is the number of planetary gear teeth, is the number of sun gear teeth.

[0028] Specifically, since the vibration source signals received by the motor end and the pump end of this scheme include multiple interference signals, this scheme constructs a motor reference signal, a pump reference signal and a planetary gearbox reference signal. The motor reference signal is constructed by the motor shaft speed signal, and is used to represent the vibration frequency of the motor. The pump reference signal is constructed by the pump shaft speed signal, and is used to represent the vibration frequency of the pump. The planetary gearbox reference signal is constructed by the planetary gearbox meshing frequency theoretical calculation formula, and is used to represent the vibration frequency of the planetary gearbox, so as to find the items closest to the motor reference signal, the pump reference signal and the planetary gearbox reference signal in the mapping vector.

[0029] Furthermore, in the step of obtaining the item with the highest frequency similarity to the motor reference signal as the motor virtual vibration source signal, a threshold function is constructed to constrain the mapping vector with the motor reference signal to obtain a motor constraint target function, and the motor constraint target function is solved to obtain the motor virtual vibration source signal.

[0030] Furthermore, in the step of obtaining the item with the highest frequency similarity to the pump reference signal as the pump virtual vibration source signal, a threshold function is constructed to constrain the mapping vector with the pump reference signal to obtain a pump constraint objective function, and the pump constraint objective function is solved to obtain the pump virtual vibration source signal.

[0031] Furthermore, in the step of obtaining the item with the highest frequency similarity to the planetary gearbox reference signal as the planetary gearbox virtual vibration source signal, a threshold function is constructed to constrain the mapping vector with the planetary gearbox reference signal to obtain a planetary gearbox constraint objective function, and the planetary gearbox constraint objective function is solved to obtain the planetary gearbox virtual vibration source signal.

[0032] Specifically, the essence of solving the motor constraint objective function, the pump constraint objective function and the planetary gearbox constraint objective function is to find the signal in the mapping vector that is closest in frequency to the motor reference signal, the pump reference signal and the planetary gearbox reference signal respectively.

[0033] Specifically, due to the complex structure of the nuclear power circulating pump unit, the vibration signal generated during its operation will be affected by many factors, and the vibrations of each component are coupled with each other, so it is impossible to find the motor virtual vibration source signal, pump virtual vibration source signal and planetary gearbox virtual vibration source signal by direct comparison. In this scheme, by solving the motor constraint objective function, pump constraint objective function and planetary gearbox constraint objective function, it is possible to better extract signals that are closer to various reference signal frequencies from the complex mixed signal, thereby improving the subsequent identifiability of faults and more accurately judging the health status of the nuclear power circulating pump unit.

[0034] Specifically, the similarity between the items in the mapping vector and the reference signal is expressed by the following formula:

[0035] in, represents any item in the mapping vector, is a reference signal, such as a motor reference signal, a pump reference signal, and a planetary gearbox reference signal, for and The similarity between them, E is the mathematical expectation.

[0036] Specifically, this solution uses the Lagrange multiplier method to solve the motor constraint objective function, the pump constraint objective function and the planetary gearbox constraint objective function.

[0037] Exemplarily, the constraint objective function can be expressed by the following formula:

[0038] in, is any item in the mapping vector, expressed as a Gram matrix, is the difference between the mean square error similarity function and the threshold function, and They are used to normalize the items in the mapping vector And the reference signal , E is the mathematical expectation, is the objective function constructed, λ is the eigenvalue to be solved, and solving the minimum value by Lagrange multiplier method can speed up the solution and provide more effective system response speed for subsequent deployment on hardware.

[0039] The structure of the multi-source fusion health recognition model of this scheme and the corresponding three signals as three-way input branches are introduced first, and then the process of training data is mentioned.

[0040] In this scheme, the structural diagram of the multi-source fusion health recognition model is as follows: Figure 2 As shown, the multi-source fusion health identification model consists of a motor feature extraction network, a pump feature extraction network, a planetary gearbox feature extraction network, a feature fusion layer and a classification layer, wherein the motor feature extraction network performs feature extraction on the motor virtual vibration source signal to obtain the motor vibration source feature, the pump feature extraction network performs feature extraction on the pump virtual vibration source signal to obtain the pump vibration source feature, the planetary gearbox feature extraction network performs feature extraction on the planetary gearbox virtual vibration source signal to obtain the planetary gearbox vibration source feature, and in the feature fusion layer, the motor vibration source feature, the pump vibration source feature and the planetary gearbox vibration source feature are fused to obtain the fusion feature, and finally the fusion feature is input into the classification layer to obtain the health status of the nuclear power circulating pump unit.

[0041] Correspondingly, the motor virtual vibration source signal, the pump virtual vibration source signal and the planetary gearbox virtual vibration source signal are input into the multi-source fusion health identification model to output the health status, wherein the motor virtual vibration source signal is input into the motor feature extraction network to extract the motor vibration source characteristics, the pump virtual vibration source signal is input into the pump feature extraction network to extract the pump vibration source characteristics, and the planetary gearbox virtual vibration source signal is input into the planetary gearbox feature extraction network to extract the planetary gearbox vibration source characteristics. The motor vibration source characteristics, pump vibration source characteristics and planetary gearbox vibration source characteristics are fused in the feature fusion layer and input into the classification layer for classification to obtain the health status.

[0042] Specifically, the health status of a nuclear power circulating pump unit includes one or any combination of motor shaft misalignment, motor shaft bending, motor bar breakage, planetary gearbox sun gear tooth root cracks, pitting, wear, tooth breakage, impeller mass imbalance, and impeller damage failures.

[0043] In some specific embodiments, the multi-source fusion health identification model can accurately determine whether the nuclear power circulating pump unit has problems such as motor shaft misalignment, motor shaft bending, motor bar breakage, planetary gearbox sun gear root cracks, pitting, wear, tooth breakage, impeller mass imbalance and impeller damage, thereby obtaining the health status of the nuclear power circulating pump unit.

[0044] Furthermore, the motor feature extraction network, the pump feature extraction network, and the planetary gearbox feature extraction network have the same structure, and are all composed of a first convolution block, a second convolution block, a third convolution block, a residual block, and a self-attention mechanism layer connected in series. The input of the feature extraction network passes through the first convolution block, the second convolution block, the third convolution block, the residual block, and the self-attention mechanism layer in sequence before being output.

[0045] Specifically, the structural diagram of the convolution block is as follows Figure 3 As shown, the first convolution block, the second convolution block, and the second convolution block all include a convolution layer, a ReLU activation layer, and a maximum pooling layer.

[0046] In some specific embodiments, the structural diagram of the feature fusion layer is as follows: Figure 4 As shown, the feature fusion layer includes a motor branch, a pump branch, a planetary gearbox branch, a multi-layer perceptron and a Softmax activation function. The motor vibration source feature is input into the motor branch to extract global features and local features to obtain a first global feature and a first local feature. The pump vibration source feature is input into the pump branch to extract global features and local features to obtain a second global feature and a second local feature. The planetary gearbox vibration source feature is input into the planetary gearbox branch to extract global features and local features to obtain a third global feature and a third local feature. The first global feature, the second global feature and the third global feature are connected in series to obtain a multi-source global feature. The first local feature, the second local feature and the third local feature are connected in series to obtain a multi-source local feature. The multi-source global feature and the multi-source local feature are respectively input into the multi-layer perceptron to obtain a global feature map and a local feature map. The global feature map and the local feature map are added and then output by the Softmax activation function to obtain a fusion feature.

[0047] Specifically, the structures of the motor branch, the pump branch, and the planetary gearbox branch are all global-local feature extraction structures. The structural schematic diagram of the global-local feature extraction structure is shown in FIG. Figure 5 As shown, the global-local feature extraction structure includes a global feature extraction branch and a local feature extraction branch, and outputs the input features to the global feature extraction branch and the local feature extraction branch respectively to obtain global features and local features.

[0048] That is, in the motor branch, the motor vibration source feature is input into the global feature extraction branch and the local feature extraction branch respectively to obtain the first global feature and the first local feature; In the pump branch, the pump vibration source feature is input into the global feature extraction branch and the local feature extraction branch respectively to obtain a second global feature and a second local feature; In the planetary gearbox branch, the vibration source characteristics of the planetary gearbox are respectively input into the global feature extraction branch and the local feature extraction branch to obtain the third global feature and the third local feature.

[0049] Furthermore, the global feature extraction branch is composed of a global average pooling layer, a first convolutional global layer, a first global batch normalization layer, a second global convolutional layer, and a second global batch normalization layer in series, and the local feature extraction branch is composed of a first local convolutional layer, a first local batch normalization layer, a second local convolutional layer, and a second local batch normalization layer in series, wherein the convolution layers in the global feature extraction branch and the local feature extraction branch are all 1×1 convolutions.

[0050] In some specific embodiments, label smoothing regularization loss is used as the loss function when training the multi-source fusion health recognition architecture.

[0051] Specifically, the formula for label smoothing regularization loss is expressed as:

[0052] Among them, LSR represents the label smoothing regularization loss, To predict the label, is the true label, is the smoothing factor, To pass and uniform distribution The true label obtained, N is the number of health states, L is the cross entropy loss function, and the expression of L is .

[0053] Specifically, this scheme uses label smoothing regularization loss as the loss function to solve the overfitting problem of the network under data imbalance and improve the generalization ability. Therefore, this scheme can significantly improve the accuracy of single sensor signal fault classification, has high accuracy and reliability, and can meet the requirements of high accuracy in identifying the health status of nuclear power circulating pump units in practical applications.

[0054] Regarding the training method of the multi-source fusion health recognition model of this solution: This scheme obtains the virtual vibration source signals of the motor, pump and planetary gearbox of the nuclear power circulating pump unit in different health states as training samples, and uses the training samples to train the constructed multi-source fusion health recognition architecture to obtain a multi-source fusion health recognition model.

[0055] Furthermore, an imbalance factor is constructed during the training process, and at least one imbalanced training subset is constructed according to the training samples based on the imbalance factor, and the imbalanced training subset is used to train the multi-source fusion health recognition architecture, wherein the imbalance factor is used to control the ratio of healthy samples to unhealthy samples in each imbalanced training subset.

[0056] Exemplarily, the formula of the imbalance factor is as follows:

[0057] in, is the imbalance factor, is the number of healthy samples, is the number of unhealthy samples.

[0058] In this scheme, four imbalanced training subsets were constructed, and the imbalance factors in the four imbalanced training subsets were set to 1, 0.2, 0.1 and 0.05, respectively.

[0059] Specifically, constructing an unbalanced training subset through an imbalance factor can more comprehensively evaluate the stability and generalization ability of the multi-source fusion health recognition model. For example, if the model trained on a data set with an imbalance factor of 0.05 can achieve good performance, it means that the model has certain advantages in processing severely unbalanced data and can more effectively identify faulty samples in minority categories.

[0060] Embodiment 2 Based on the same idea, refer to Figure 6 The present application also proposes a multi-virtual source fusion nuclear power circulating pump unit health status identification device, including: The acquisition device is used to acquire the acceleration signal of the nuclear power circulating pump unit, and pre-process the acceleration signal through nuclear independent component analysis to obtain the motor virtual vibration source signal, the pump virtual vibration source signal and the planetary gearbox virtual vibration source signal, wherein the pre-processing method based on nuclear independent component analysis is: Mapping the acceleration signal to a high-dimensional linear space to obtain a mapping vector, constructing a motor reference signal, a pump reference signal and a planetary gearbox reference signal, and obtaining the item with the highest frequency similarity to the motor reference signal in the mapping vector based on the constructed threshold function as a motor virtual vibration source signal, obtaining the item with the highest frequency similarity to the pump reference signal as a pump virtual vibration source signal, and obtaining the item with the highest frequency similarity to the planetary gearbox reference signal as a planetary gearbox virtual vibration source signal; The health status identification device is used to input the motor virtual vibration source signal, the pump virtual vibration source signal and the planetary gearbox virtual vibration source signal into the multi-source fusion health identification model to obtain the health status of the nuclear power circulating pump unit.

[0061] Embodiment 3 This embodiment also provides an electronic device, referring to Figure 7 , comprises a memory 404 and a processor 402, wherein the memory 404 stores a computer program, and the processor 402 is configured to run the computer program to execute the steps in any of the above method embodiments.

[0062] Specifically, the processor 402 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.

[0063] Among them, the memory 404 may include a large-capacity memory for data or instructions. For example, but not limitation, the memory 404 may include a hard disk drive (HDD), a floppy disk drive, a solid-state drive (SSD), 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 404 may include a removable or non-removable (or fixed) medium. In appropriate cases, the memory 404 may be inside or outside the data processing device. In a specific embodiment, the memory 404 is a non-volatile memory. In a specific embodiment, the memory 404 includes a read-only memory (ROM) and a random access memory (RAM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (Programmable Read-Only Memory, PROM for short), an erasable PROM (Erasable Programmable Read-Only Memory, EPROM for short), an electrically erasable PROM (Electrically Erasable Programmable Read-Only Memory, EEPROM for short), an electrically alterable ROM (Electrically Alterable Read-Only Memory, EAROM for short) or a flash memory (FLASH) or a combination of two or more of these. In appropriate circumstances, the RAM may be a static random access memory (SRAM) or a dynamic random access memory (DRAM), wherein the DRAM may be a fast page mode dynamic random access memory 404 (FPMDRAM), an extended data output dynamic random access memory (EDODRAM), a synchronous dynamic random access memory (SDRAM), etc.

[0064] The memory 404 may be used to store or cache various data files required for processing and / or communication, as well as possible computer program instructions executed by the processor 402 .

[0065] The processor 402 reads and executes the computer program instructions stored in the memory 404 to implement any one of the health status identification methods of a nuclear power circulating pump unit with multi-virtual source fusion in the above embodiments.

[0066] Optionally, the electronic device may further include a transmission device 406 and an input / output device 408 , wherein the transmission device 406 is connected to the processor 402 , and the input / output device 408 is connected to the processor 402 .

[0067] The transmission device 406 can be used to receive or send data via a network. Specific examples of the above-mentioned network may include a wired or wireless network provided by a communication provider of the electronic device. In one example, the transmission device includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 406 can be a radio frequency (Radio Frequency, referred to as RF) module, which is used to communicate with the Internet wirelessly.

[0068] The input / output device 408 is used to input or output information. In this embodiment, the input information may be an acceleration signal, a reference signal, etc., and the output information may be the health status of a nuclear power circulating pump unit, etc.

[0069] Optionally, in this embodiment, the processor 402 may be configured to perform the following steps through a computer program: The acceleration signal of the nuclear power circulating pump unit is obtained, and the acceleration signal is preprocessed by nuclear independent component analysis to obtain the motor virtual vibration source signal, the pump virtual vibration source signal and the planetary gearbox virtual vibration source signal, wherein the preprocessing method based on nuclear independent component analysis is: Mapping the acceleration signal to a high-dimensional linear space to obtain a mapping vector, constructing a motor reference signal, a pump reference signal and a planetary gearbox reference signal, and obtaining the item with the highest frequency similarity to the motor reference signal in the mapping vector based on the constructed threshold function as a motor virtual vibration source signal, obtaining the item with the highest frequency similarity to the pump reference signal as a pump virtual vibration source signal, and obtaining the item with the highest frequency similarity to the planetary gearbox reference signal as a planetary gearbox virtual vibration source signal; The virtual vibration source signals of the motor, the pump and the planetary gearbox are input into the multi-source fusion health identification model to obtain the health status of the nuclear power circulating pump unit.

[0070] It should be noted that the specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementation modes, and this embodiment will not be described in detail here.

[0071] In general, various embodiments may be implemented in hardware or dedicated circuits, software, logic, or any combination thereof. Some aspects of the invention may be implemented in hardware, while other aspects may be implemented in firmware or software that may be executed by a controller, microprocessor, or other computing device, but the invention is not limited thereto. Although various aspects of the invention may be shown and described as block diagrams, flow charts, or using some other graphical representation, it should be understood that, as non-limiting examples, the boxes, devices, systems, techniques, or methods described herein may be implemented in hardware, software, firmware, dedicated circuits or logic, general-purpose hardware or controllers or other computing devices, or some combination thereof.

[0072] Embodiments of the present invention may be implemented by computer software that is executable by a data processor of a mobile device, such as in a processor entity, or by hardware, or by a combination of software and hardware. Computer software or programs (also referred to as program products) including software routines, applets and / or macros may be stored in any device-readable data storage medium, and they include program instructions for performing specific tasks. A computer program product may include one or more computer executable components configured to perform an embodiment when the program is run. One or more computer executable components may be at least one software code or a portion thereof. In addition, at this point, it should be noted that, for example, Figure 7 Any block of the logic flow in the program may represent program steps, or interconnected logic circuits, blocks and functions, or a combination of program steps and logic circuits, blocks and functions. The software may be stored on physical media such as memory chips or storage blocks implemented within a processor, magnetic media such as hard disks or floppy disks, and optical media such as, for example, DVDs and their data variants, CDs, etc. Physical media are non-transitory media.

[0073] Those skilled in the art should understand that the technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0074] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for identifying the health status of a nuclear power circulating pump unit by fusion of multiple virtual sources, characterized in that: The following steps are involved: The acceleration signal of the nuclear power circulating pump unit is obtained, and the acceleration signal is preprocessed by nuclear independent component analysis to obtain the motor virtual vibration source signal, the pump virtual vibration source signal and the planetary gearbox virtual vibration source signal, wherein the preprocessing method based on nuclear independent component analysis is: Mapping the acceleration signal to a high-dimensional linear space to obtain a mapping vector, constructing a motor reference signal, a pump reference signal and a planetary gearbox reference signal, and obtaining the item with the highest frequency similarity to the motor reference signal in the mapping vector based on the constructed threshold function as a motor virtual vibration source signal, obtaining the item with the highest frequency similarity to the pump reference signal as a pump virtual vibration source signal, and obtaining the item with the highest frequency similarity to the planetary gearbox reference signal as a planetary gearbox virtual vibration source signal; The motor virtual vibration source signal, pump virtual vibration source signal and planetary gearbox virtual vibration source signal are input into the multi-source fusion health identification model to obtain the health status of the nuclear power circulating pump unit.

2. A method for identifying the health status of a nuclear power circulating pump unit by fusion of multiple virtual sources according to claim 1, characterized in that: A nonlinear kernel function in a polynomial form is constructed, and a kernel function value of any data point in the acceleration signal and other data points is calculated based on the nonlinear kernel function to map the acceleration signal to a high-dimensional linear space.

3. The method for identifying the health status of a nuclear power circulating pump unit by fusion of multiple virtual sources according to claim 1 is characterized in that: The motor shaft speed signal and pump shaft speed signal in the nuclear power circulating pump unit are obtained by eddy current sensors. A square wave pulse sequence is constructed based on the motor shaft speed signal to obtain the motor reference signal. A square wave pulse sequence is constructed based on the pump shaft speed signal to obtain the pump reference signal. The meshing frequency of the planetary gearbox is used as the planetary gearbox reference signal.

4. The method for identifying the health status of a nuclear power circulating pump unit by fusion of multiple virtual sources according to claim 1 is characterized in that: A motor constraint target function is obtained by constraining a mapping vector with a motor reference signal by constructing a threshold function, and a motor virtual vibration source signal is obtained by solving the motor constraint target function; a pump constraint target function is obtained by constraining a mapping vector with a pump reference signal by constructing a threshold function, and a pump virtual vibration source signal is obtained by solving the pump constraint target function; a planetary gearbox constraint target function is obtained by constraining a mapping vector with a planetary gearbox reference signal by constructing a threshold function, and a planetary gearbox virtual vibration source signal is obtained by solving the planetary gearbox constraint target function.

5. The method for identifying the health status of a nuclear power circulating pump unit by fusion of multiple virtual sources according to claim 1 is characterized in that: The multi-source fusion health identification model is composed of a motor feature extraction network, a pump feature extraction network, a planetary gearbox feature extraction network, a feature fusion layer and a classification layer. The motor virtual vibration source signal, the pump virtual vibration source signal and the planetary gearbox virtual vibration source signal are input into the multi-source fusion health identification model to output the health status, wherein the motor virtual vibration source signal is input into the motor feature extraction network to extract the motor vibration source feature, the pump virtual vibration source signal is input into the pump feature extraction network to extract the pump vibration source feature, the planetary gearbox virtual vibration source signal and the planetary gearbox feature extraction network extract the planetary gearbox vibration source feature, and the motor vibration source feature, the pump vibration source feature and the planetary gearbox vibration source feature are fused in the feature fusion layer and input into the classification layer to classify and obtain the health status.

6. A method for identifying the health status of a nuclear power circulating pump unit by fusion of multiple virtual sources according to claim 5, characterized in that: The feature fusion layer includes a motor branch, a pump branch, a planetary gearbox branch, a multi-layer perceptron and a Softmax activation function. The motor vibration source feature is input into the motor branch to extract global features and local features to obtain a first global feature and a first local feature. The pump vibration source feature is input into the pump branch to extract global features and local features to obtain a second global feature and a second local feature. The planetary gearbox vibration source feature is input into the planetary gearbox branch to extract global features and local features to obtain a third global feature and a third local feature. The first global feature, the second global feature and the third global feature are connected in series to obtain a multi-source global feature. The first local feature, the second local feature and the third local feature are connected in series to obtain a multi-source local feature. The multi-source global feature and the multi-source local feature are respectively input into the multi-layer perceptron to obtain a global feature map and a local feature map. The global feature map and the local feature map are added and then output by the Softmax activation function to obtain a fusion feature.

7. The method for identifying the health status of a nuclear power circulating pump unit by fusion of multiple virtual sources according to claim 1, characterized in that: The multi-source fusion health recognition model is obtained by obtaining the motor virtual vibration source signals, pump virtual vibration source signals and planetary gearbox virtual vibration source signals of the nuclear power circulating pump unit in different health states as training samples, and using the training samples to train the constructed multi-source fusion health recognition architecture.

8. The method for identifying the health status of a nuclear power circulating pump unit by fusion of multiple virtual sources according to claim 7 is characterized in that: An imbalance factor is constructed during the training process, at least one imbalanced training subset is constructed according to the training samples based on the imbalance factor, and the imbalanced training subset is used to train the multi-source fusion health recognition architecture, wherein the imbalance factor is used to control the ratio of healthy samples to unhealthy samples in each imbalanced training subset.

9. A multi-virtual source fusion nuclear power circulating pump unit health status identification device, characterized in that: include: The acquisition device is used to acquire the acceleration signal of the nuclear power circulating pump unit, and pre-process the acceleration signal through nuclear independent component analysis to obtain the motor virtual vibration source signal, the pump virtual vibration source signal and the planetary gearbox virtual vibration source signal, wherein the pre-processing method based on nuclear independent component analysis is: Mapping the acceleration signal to a high-dimensional linear space to obtain a mapping vector, constructing a motor reference signal, a pump reference signal and a planetary gearbox reference signal, and obtaining the item with the highest frequency similarity to the motor reference signal in the mapping vector based on the constructed threshold function as a motor virtual vibration source signal, obtaining the item with the highest frequency similarity to the pump reference signal as a pump virtual vibration source signal, and obtaining the item with the highest frequency similarity to the planetary gearbox reference signal as a planetary gearbox virtual vibration source signal; The health status identification device is used to input the motor virtual vibration source signal, the pump virtual vibration source signal and the planetary gearbox virtual vibration source signal into the multi-source fusion health identification model to obtain the health status of the nuclear power circulating pump unit.

10. A readable storage medium, characterized in that: The readable storage medium stores a computer program, which includes a program code for controlling a process to execute a process, and the process includes a method for identifying the health status of a nuclear power circulating pump unit with multi-virtual source fusion according to any one of claims 1-8.

Images