Converter health state detection method, apparatus and device, and storage medium

Abstract

The invention discloses a converter health state detection method, device and equipment and a storage medium, and the method comprises the steps: obtaining the real-time working data of a converter, and enabling the real-time working data to serve as the input data of a converter health detection model; performing feature extraction on the real-time working data, and performing feature fusion on the extracted feature data to generate fused feature data; and according to the fusion feature data, generating an output result of the converter health detection model, and taking the output result as a health state detection result of the converter. According to the technical scheme, the recognition precision of the converter health detection model during health state detection of the converter with weak and complex features or unobvious fault phenomena is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of converter fault identification, and in particular to a converter health status detection method, device, equipment and storage medium. Background Art

[0002] With the development of power electronics technology, the power level of three-phase converters has been continuously improved. However, due to the complex working environment of the converter, the long-term operation of the converter makes its failure frequency higher and higher. Once a failure occurs, it will bring significant losses and pose a very serious safety hazard.

[0003] In recent years, with the development of artificial intelligence technology, the fault diagnosis method of artificial intelligence algorithm has gradually gained attention. The existing technical solution directly inputs the three-phase current data of the converter into the diagnostic model, and the diagnostic model directly extracts the original feature data of the three-phase current data. However, when identifying the health status of the converter, due to the small difference in the characteristics of the output current of converters with different health levels, it is difficult for the diagnostic model to extract the effective features in the current characteristics, resulting in low accuracy of the model in identifying the health status of the converter. Summary of the invention

[0004] The present invention provides a method for detecting the health status of a converter, so as to improve the recognition accuracy of the health status of the converter.

[0005] According to one aspect of the present invention, a method for detecting a health status of a converter is provided, the method comprising:

[0006] Acquire real-time operating data of the converter, and use the real-time operating data as input data of the converter health detection model; wherein the real-time operating data includes three-phase output voltage data and three-phase output current data;

[0007] Extracting features from the real-time working data, and fusing the extracted feature data to generate fused feature data; wherein the fused feature data includes current feature information and voltage feature information;

[0008] An output result of the converter health detection model is generated according to the fused feature data, and the output result is used as a health status detection result of the converter.

[0009] According to another aspect of the present invention, there is provided a device for detecting the health status of a converter, the device comprising:

[0010] A converter data acquisition module, used to acquire real-time operating data of the converter, and use the real-time operating data as input data of the converter health detection model; wherein the real-time operating data includes three-phase output voltage data and three-phase output current data;

[0011] A feature fusion module, used to extract features from the real-time working data, and perform feature fusion on the extracted feature data to generate fused feature data; wherein the fused feature data includes current feature information and voltage feature information;

[0012] The health status determination module is used to generate an output result of the converter health detection model according to the fused feature data, and use the output result as the converter health status detection result.

[0013] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0014] at least one processor; and

[0015] a memory communicatively connected to the at least one processor; wherein,

[0016] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the converter health status detection method described in any embodiment of the present invention.

[0017] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the converter health status detection method described in any embodiment of the present invention when executed.

[0018] According to another aspect of the present invention, a computer program product is provided. The computer program product comprises a computer program. When the computer program is executed by a processor, the method for detecting the health status of a converter according to any embodiment of the present invention is implemented.

[0019] The technical solution of the embodiment of the present invention, by performing feature fusion on the three-phase output voltage data and the three-phase output current data of the converter, identifies and detects the health status of the converter based on the fused feature data, thereby avoiding the problem that the detection model has difficulty in extracting effective features from the current features due to the small feature differences between the output currents of converters with different health levels, resulting in low accuracy in identifying the health status of the rectifier, and achieves the recognition accuracy of the converter health status detection model when performing health status detection on converters with weak, complex features or unclear fault phenomena.

[0020] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0022] Figure 1 is a flow chart of a method for detecting the health status of a converter provided according to Embodiment 1 of the present invention;

[0023] Figure 2 is a flow chart of a method for detecting the health status of a converter provided according to Embodiment 2 of the present invention;

[0024] Figure 3a is a flow chart of a method for detecting the health status of a converter provided according to Embodiment 3 of the present invention;

[0025] Figure 3b is an architecture diagram of a variable step-size weighted model provided according to Embodiment 3 of the present invention;

[0026] Figure 3c is an architecture diagram of a reverse attention layer provided according to Embodiment 3 of the present invention;

[0027] Figure 4 is a structural schematic diagram of a converter health status detection device provided according to a fourth embodiment of the present invention;

[0028] Figure 5 It is a structural schematic diagram of an electronic device for implementing the converter health status detection method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0029] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0030] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0031] Embodiment 1

[0032] Figure 1 A flowchart of a method for detecting the health status of a converter is provided for the first embodiment of the present invention. This embodiment is applicable to the case of performing health detection on a converter. The method can be executed by a converter health detection device. The converter health detection device can be implemented in the form of hardware and / or software. The converter health detection device can be configured in various general computing devices, such as a terminal device integrated with a converter health detection model. Figure 1 As shown, the method includes:

[0033] S110, acquiring real-time operating data of the converter, and using the real-time operating data as input data of a converter health detection model.

[0034] The real-time working data may include real-time three-phase output voltage data and three-phase output current data during the operation of the converter.

[0035] The converter health detection model may refer to a pre-trained neural network model. Optionally, the converter health detection model includes at least one variable step-size weighted model, which may include a first convolution layer and a second convolution layer whose number of convolution kernels and convolution kernel moving step sizes are mutually inverse, a normalization layer, a first activation function layer, and a reverse attention layer. Different variable step-size weighted models have different model parameters. Exemplarily, when the variable step-size weighted models are different, the number of convolution kernels and the convolution kernel moving step sizes of the convolution layers in the variable step-size weighted models are also different. It should be noted that the first convolution layer and the second convolution layer whose number of convolution kernels and convolution kernel moving step sizes are mutually inverse may refer to that if the convolution kernel size of the first convolution layer is a and the convolution kernel moving step size is b, then the convolution kernel size of the second convolution layer is b and the convolution kernel moving step size is a.

[0036] In the embodiment of the present invention, the real-time detection of the health status of the converter can be performed by collecting the real-time working data of the converter. Optionally, the real-time working data of the converter can be periodically collected in real time at a preset collection frequency, and the collection frequency can be adaptively set according to those skilled in the art.

[0037] S120: extract features from the real-time working data, and perform feature fusion on the extracted feature data to generate fused feature data.

[0038] The fused characteristic data may include current characteristic information and voltage characteristic information.

[0039] Specifically, preliminary feature extraction may be performed on the three-phase output voltage data and the three-phase output current data, and further feature extraction and fusion may be performed on the extracted voltage feature data and current feature data to generate fused feature data.

[0040] S130. Generate an output result of a converter health detection model according to the fused feature data, and use the output result as a converter health status detection result.

[0041] Specifically, the converter health detection model can detect the health status of the converter by extracting the fused feature data generated by the real-time working data of the converter, and output the health status detection result of the converter.

[0042] It should be noted that since the fused feature data carries the current feature information and voltage feature information of the converter, the converter health detection model can extract more comprehensive converter working feature information, thereby improving the recognition accuracy of the converter health status.

[0043] The technical solution of the embodiment of the present invention, by performing feature fusion on the three-phase output voltage data and the three-phase output current data of the converter, identifies and detects the health status of the converter based on the fused feature data, thereby avoiding the problem that the detection model has difficulty in extracting effective features from the current features due to the small feature differences between the output currents of converters with different health levels, resulting in low accuracy in identifying the health status of the rectifier, and achieves the recognition accuracy of the converter health status detection model when performing health status detection on converters with weak, complex features or unclear fault phenomena.

[0044] Embodiment 2

[0045] Figure 2This is a flowchart of a method for detecting the health status of a converter provided in the second embodiment of the present invention. This embodiment further refines the above embodiment and provides specific steps for extracting features from real-time working data, fusing the extracted feature data, and generating fused feature data. It should be noted that for the undetailed parts of the embodiment of the present invention, please refer to the relevant descriptions of other embodiments, which will not be repeated here. Figure 2 As shown, the method includes:

[0046] S210: Acquire real-time operating data of the converter, and use the real-time operating data as input data of a converter health detection model.

[0047] S220 , respectively extracting features from the three-phase output voltage data and the three-phase output current data to obtain instantaneous amplitude features and instantaneous frequency features of the three-phase output voltage data and the three-phase output current data.

[0048] S230, concatenating the instantaneous amplitude feature and the instantaneous frequency feature to generate fused feature data.

[0049] S240: Generate an output result of the converter health detection model according to the fused feature data, and use the output result as a health status detection result of the converter.

[0050] In an embodiment of the present invention, the converter health detection model can first perform Hilbert transform on the real-time working data, perform initial feature extraction, and obtain the instantaneous amplitude characteristics and instantaneous frequency characteristics of the three-phase output voltage data and the three-phase output current data, and then the instantaneous amplitude characteristics and instantaneous frequency characteristics of the three-phase output voltage data and the three-phase output current data can be spliced. Exemplarily, the instantaneous amplitude characteristics and instantaneous frequency characteristics of the three-phase output voltage data and the three-phase output current data obtained by Hilbert transform are changed from three-phase data to a 6-column feature data, and the feature information of the three-phase output voltage data and the three-phase output current data is fused by performing feature splicing, and the generated fused feature data can be a 12-column feature data.

[0051] Optionally, before concatenating the instantaneous amplitude feature and the instantaneous frequency feature, the method further includes: sampling and reducing the dimension of the instantaneous amplitude feature and the instantaneous frequency feature at a preset sampling frequency.

[0052] Specifically, the instantaneous amplitude feature and the instantaneous frequency feature may be sampled at intervals at a preset sampling frequency, and data dimension reduction may be performed on the instantaneous amplitude feature and the instantaneous frequency feature.

[0053] The dimension of the feature data is reduced through interval sampling, which reduces the amount of data while ensuring that important features in the feature data are not lost, thereby improving data processing efficiency.

[0054] The technical solution of the embodiment of the present invention extracts features from three-phase current signals and three-phase voltage signals, and then performs data dimension reduction on the extracted feature data through interval sampling. This reduces the amount of feature data while ensuring that feature data in real-time working data is not lost, thereby achieving the integration of voltage features and current features while ensuring the integrity of feature data.

[0055] Embodiment 3

[0056] FIG3 is a flow chart of a method for detecting the health status of a converter provided in the third embodiment of the present invention. This embodiment further refines the above embodiment and provides specific steps for further feature extraction of fused feature data. It should be noted that, for the undetailed parts of the embodiment of the present invention, reference can be made to the relevant descriptions of other embodiments, which will not be repeated here. Figure 3a As shown, the method includes:

[0057] S310: Input the fused feature data into the first convolution layer and the second convolution layer respectively to perform feature extraction operations to obtain first feature convolution parameters and second feature convolution parameters.

[0058] S320, inputting the first feature convolution parameter and the second feature convolution parameter into the normalization layer and the first activation function layer in sequence to perform feature normalization operation and feature nonlinear operation processing to obtain a first feature tensor and a second feature tensor.

[0059] S330. Input the first feature tensor and the second feature tensor into the reverse attention layer respectively for feature extraction operation to obtain the first feature tensor parameters and the second feature tensor parameters.

[0060] S340. Perform weighted processing on the first feature tensor parameter and the second feature tensor parameter according to the mean value of the first feature tensor parameter and the mean value of the second feature tensor parameter, and sum the weighted first feature tensor parameter and the second feature tensor parameter to complete the feature extraction operation on the fused feature data.

[0061] In the embodiment of the present invention, the variable step-size weighted model can be as follows: Figure 3bAs shown, in order to obtain more fine-grained feature data, the fused feature data can be respectively input into at least one variable step-size weighted model in the converter health detection model, and the fused feature data input into the variable step-size weighted model will enter the first convolution layer and the second convolution layer respectively for feature extraction operations to obtain the first feature convolution parameters output by the first convolution layer and the second feature convolution parameters output by the second convolution layer; thereafter, the first feature convolution parameters and the second feature convolution parameters will be sequentially input into the normalization layer and the first activation function layer for feature normalization operations and feature nonlinear operations to obtain the first feature tensor and the second feature tensor corresponding to the first feature convolution parameters, respectively. The second feature tensor corresponding to the feature convolution parameters; the reverse attention layer will perform feature cleaning on the first feature tensor and the second feature tensor, and perform further feature extraction, so that important features are highlighted and minor features are suppressed, and the first feature tensor parameters corresponding to the first feature tensor and the second feature tensor parameters corresponding to the second feature tensor are obtained; finally, the first feature tensor parameters and the second feature tensor parameters can be weighted according to the mean of the first feature tensor parameters and the mean of the second feature tensor parameters, and the weighted first feature tensor parameters and the second feature tensor parameters can be summed to complete the feature extraction operation of the fused feature data. Optionally, the first activation function layer can be a ReLU activation function layer.

[0062] Optionally, the first feature tensor parameter and the second feature tensor parameter are weighted according to the mean of the first feature tensor parameter and the mean of the second feature tensor parameter, including: multiplying the first feature tensor parameter and the second feature tensor parameter with the mean of the first feature tensor sequence parameter and the mean of the second feature tensor sequence parameter, respectively, to complete the weighted processing of the first feature tensor sequence parameter and the second feature tensor sequence parameter.

[0063] By performing weighted processing on the first characteristic tensor sequence parameters and the second characteristic tensor sequence parameters, the converter health detection model can better identify subtle features.

[0064] Optional, such as Figure 3c As shown, the reverse attention layer may include a third convolution layer, a second activation function layer, and a third activation function layer. It should be noted that the third activation function layer may be a ReLU activation function layer, and the second activation function layer may be a Softmax activation function layer.

[0065] Optionally, the first feature tensor and the second feature tensor are respectively input into the reverse attention layer for feature sequence information extraction to obtain the first feature tensor parameter and the second feature tensor parameter, including: inputting the first feature tensor and the second feature tensor into the third convolution layer for feature extraction operation, inputting the first feature tensor and the second feature tensor after feature extraction into the second activation function layer for feature conversion, and obtaining the weight ratio tensors of the first feature tensor and the second feature tensor respectively; negating the weight ratio tensors of the first feature tensor and the second feature tensor, respectively summing them with their respective means, and inputting the summation results into the third activation function layer for feature negative value removal operation, and obtaining the weight factor of the first feature tensor and the weight factor of the second feature tensor respectively; negating the weight factor, and multiplying the negated weight factor with the corresponding first feature tensor and second feature tensor and then adding them to obtain the first feature tensor parameter and the second feature tensor parameter. It should be noted that in the embodiment of the present invention, negation can refer to negation of positive and negative signs.

[0066] In the embodiment of the present invention, the application of the reverse attention layer can highlight the important features in the feature data and suppress the secondary features, thereby effectively improving the recognition accuracy of the converter health status by the converter health detection model.

[0067] In an embodiment of the present invention, after completing the feature extraction operation on the fused feature data, the feature extraction parameters of the fused feature number output by at least one variable step-size weighted model can be spliced, and the converter health detection model outputs the recognition result of the converter health status based on the spliced ​​feature extraction parameters.

[0068] The technical solution of the embodiment of the present invention, by adopting a variable step-size weighted model with different numbers of convolution kernels and different convolution kernel moving steps, effectively extracts multi-scale features of fused feature data, so that the converter health detection model can better identify tiny feature data, thereby improving the recognition accuracy of the converter health detection model.

[0069] Embodiment 4

[0070] Figure 4 This is a schematic diagram of the structure of a converter health status detection device provided by Embodiment 4 of the present invention. Figure 4 As shown, the device comprises:

[0071] The converter data acquisition module 410 is used to acquire the real-time operating data of the converter and use the real-time operating data as input data of the converter health detection model; wherein the real-time operating data includes three-phase output voltage data and three-phase output current data;

[0072] The feature fusion module 420 is used to extract features from the real-time working data, and perform feature fusion on the extracted feature data to generate fused feature data; wherein the fused feature data includes current feature information and voltage feature information;

[0073] The health status determination module 430 is used to generate an output result of the converter health detection model according to the fused feature data, and use the output result as the converter health status detection result.

[0074] The technical solution of the embodiment of the present invention, by performing feature fusion on the three-phase output voltage data and the three-phase output current data of the converter, identifies and detects the health status of the converter based on the fused feature data, thereby avoiding the problem that the detection model has difficulty in extracting effective features from the current features due to the small feature differences between the output currents of converters with different health levels, resulting in low accuracy in identifying the health status of the rectifier, and achieves the recognition accuracy of the converter health status detection model when performing health status detection on converters with weak, complex features or unclear fault phenomena.

[0075] Optionally, the feature fusion module 420 includes:

[0076] The feature fusion unit is used to extract features from the three-phase output voltage data and the three-phase output current data respectively to obtain instantaneous amplitude features and instantaneous frequency features of the three-phase output voltage data and the three-phase output current data; and to splice the instantaneous amplitude features and the instantaneous frequency features to generate fused feature data.

[0077] Optionally, the feature fusion module 420 further includes:

[0078] The feature dimension reduction unit is used to perform sampling and dimension reduction on the instantaneous amplitude feature and the instantaneous frequency feature at a preset sampling frequency.

[0079] Optionally, the converter health detection model includes at least one variable step-size weighted model, which includes a first convolution layer and a second convolution layer in which the number of convolution kernels and the convolution kernel movement step are mutually inverse, a normalization layer, a first activation function layer, and a reverse attention layer.

[0080] Optionally, the device further includes a feature extraction module; the feature extraction module includes:

[0081] A convolution processing unit, used for inputting the fused feature data into the first convolution layer and the second convolution layer respectively to perform feature extraction operations, so as to obtain first feature convolution parameters and second feature convolution parameters;

[0082] A normalization unit, used for inputting the first feature convolution parameter and the second feature convolution parameter into a normalization layer and a first activation function layer in sequence to perform feature normalization operation and feature nonlinear operation processing, so as to obtain a first feature tensor and a second feature tensor;

[0083] An attention layer unit, used for inputting the first feature tensor and the second feature tensor into a reverse attention layer for feature extraction to obtain a first feature tensor parameter and a second feature tensor parameter;

[0084] A feature extraction unit is used to perform weighted processing on the first feature tensor parameter and the second feature tensor parameter according to the mean value of the first feature tensor parameter and the mean value of the second feature tensor parameter, and sum the weighted first feature tensor parameter and the second feature tensor parameter to complete the feature extraction operation on the fused feature data.

[0085] Optionally, the feature extraction unit may be specifically used for:

[0086] The first feature tensor parameter and the second feature tensor parameter are multiplied by the mean of the first feature tensor sequence parameter and the mean of the second feature tensor sequence parameter, respectively, to complete weighted processing of the first feature tensor sequence parameter and the second feature tensor sequence parameter.

[0087] Optionally, the reverse attention layer may include a third convolutional layer, a second activation function layer, and a third activation function layer.

[0088] Optionally, the attention layer unit can be specifically used to:

[0089] The first feature tensor and the second feature tensor are respectively input into the third convolutional layer for feature extraction operation, and the first feature tensor and the second feature tensor after feature extraction are input into the second activation function layer for feature conversion, so as to obtain weight ratio tensors of the first feature tensor and the second feature tensor respectively; after negating the weight ratio tensors of the first feature tensor and the second feature tensor, they are summed with their respective means, and the summation results are input into the third activation function layer for feature negative value removal operation, so as to obtain weight factors of the first feature tensor and weight factors of the second feature tensor respectively; the weight factors are negated, and the negated weight factors are multiplied with the corresponding first feature tensor and second feature tensor and then added to obtain first feature tensor parameters and second feature tensor parameters.

[0090] The converter health status detection device provided in the embodiment of the present invention can execute the converter health status detection method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0091] Embodiment 5

[0092] Figure 5 A schematic diagram of an electronic device 510 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0093] like Figure 5 As shown, the electronic device 510 includes at least one processor 511, and a memory connected to the at least one processor 511 in communication, such as a read-only memory (ROM) 512, a random access memory (RAM) 513, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 511 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 512 or the computer program loaded from the storage unit 518 to the random access memory (RAM) 513. In the RAM 513, various programs and data required for the operation of the electronic device 510 can also be stored. The processor 511, the ROM 512, and the RAM 513 are connected to each other via a bus 514. An input / output (I / O) interface 515 is also connected to the bus 514.

[0094] A number of components in the electronic device 510 are connected to the I / O interface 515, including: an input unit 516, such as a keyboard, a mouse, etc.; an output unit 517, such as various types of displays, speakers, etc.; a storage unit 518, such as a disk, an optical disk, etc.; and a communication unit 519, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 519 allows the electronic device 510 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0095] The processor 511 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 511 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 511 executes the various methods and processes described above, such as a method for detecting the health status of a converter.

[0096] In some embodiments, the converter health status detection method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 518. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 510 via the ROM 512 and / or the communication unit 519. When the computer program is loaded into the RAM 513 and executed by the processor 511, one or more steps of the converter health status detection method described above may be performed. Alternatively, in other embodiments, the processor 511 may be configured to execute the converter health status detection method in any other appropriate manner (e.g., by means of firmware).

[0097] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0098] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0099] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0100] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0101] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0102] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0103] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0104] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for detecting the health status of a converter, characterized in that: include: Acquire real-time operating data of the converter, and use the real-time operating data as input data of the converter health detection model; wherein the real-time operating data includes three-phase output voltage data and three-phase output current data; Extracting features from the real-time working data, and fusing the extracted feature data to generate fused feature data; wherein the fused feature data includes current feature information and voltage feature information; An output result of the converter health detection model is generated according to the fused feature data, and the output result is used as a health status detection result of the converter.

2. The method according to claim 1, characterized in that Extracting features from the real-time working data and fusing the extracted feature data to generate fused feature data includes: Performing feature extraction on the three-phase output voltage data and the three-phase output current data respectively to obtain instantaneous amplitude features and instantaneous frequency features of the three-phase output voltage data and the three-phase output current data; The instantaneous amplitude feature and the instantaneous frequency feature are concatenated to generate fused feature data.

3. The method according to claim 2, characterized in that Before splicing the instantaneous amplitude feature and the instantaneous frequency feature, the method further includes: The instantaneous amplitude feature and the instantaneous frequency feature are sampled and dimensionally reduced at a preset sampling frequency.

4. The method according to claim 1, characterized in that: The converter health detection model includes at least one variable step-size weighted model, wherein the variable step-size weighted model includes a first convolution layer and a second convolution layer, in which the number of convolution kernels and the convolution kernel moving step sizes are mutually inverse, a normalization layer, a first activation function layer, and a reverse attention layer; Accordingly, the feature extraction operation process of the fused feature data includes: Inputting the fused feature data into the first convolution layer and the second convolution layer respectively to perform feature extraction operations to obtain first feature convolution parameters and second feature convolution parameters; Inputting the first feature convolution parameter and the second feature convolution parameter into a normalization layer and a first activation function layer in sequence to perform feature normalization operation and feature nonlinear operation processing, so as to obtain a first feature tensor and a second feature tensor; Inputting the first feature tensor and the second feature tensor into the reverse attention layer for feature extraction, respectively, to obtain a first feature tensor parameter and a second feature tensor parameter; According to the mean value of the first feature tensor parameter and the mean value of the second feature tensor parameter, the first feature tensor parameter and the second feature tensor parameter are weighted, and the weighted first feature tensor parameter and the second feature tensor parameter are summed to complete the feature extraction operation of the fused feature data.

5. The method according to claim 4, characterized in that According to the mean value of the first feature tensor sequence parameter and the mean value of the second feature tensor sequence parameter, weighted processing is performed on the first feature tensor sequence parameter and the second feature tensor sequence parameter, including: The first feature tensor parameter and the second feature tensor parameter are multiplied by the mean of the first feature tensor sequence parameter and the mean of the second feature tensor sequence parameter, respectively, to complete weighted processing of the first feature tensor sequence parameter and the second feature tensor sequence parameter.

6. The method according to claim 4, characterized in that The reverse attention layer includes a third convolutional layer, a second activation function layer and a third activation function layer; Correspondingly, the first feature tensor and the second feature tensor are respectively input into the reverse attention layer to extract feature sequence information, and the first feature tensor parameters and the second feature tensor parameters are obtained, including: Inputting the first feature tensor and the second feature tensor into the third convolution layer for feature extraction, respectively, inputting the first feature tensor and the second feature tensor after feature extraction into the second activation function layer for feature conversion, and obtaining weight proportion tensors of the first feature tensor and the second feature tensor respectively; After negating the weight proportion tensors of the first feature tensor and the second feature tensor, sum them with their respective means, and input the summation results into the third activation function layer for feature negative value removal operation, so as to obtain the weight factor of the first feature tensor and the weight factor of the second feature tensor respectively; The weight factor is inverted, and the inverted weight factor is multiplied by the corresponding first feature tensor and second feature tensor and then added to obtain the first feature tensor parameter and the second feature tensor parameter.

7. A converter health status detection device, characterized in that: include: A converter data acquisition module, used to acquire real-time operating data of the converter, and use the real-time operating data as input data of the converter health detection model; wherein the real-time operating data includes three-phase output voltage data and three-phase output current data; A feature fusion module, used to extract features from the real-time working data, and perform feature fusion on the extracted feature data to generate fused feature data; wherein the fused feature data includes current feature information and voltage feature information; The health status determination module is used to generate an output result of the converter health detection model according to the fused feature data, and use the output result as the converter health status detection result.

8. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the converter health status detection method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the converter health status detection method according to any one of claims 1 to 7 when executed.

10. A computer program product, characterized in that It comprises a computer program, which, when executed by a processor, implements the method for detecting the health status of a converter according to any one of claims 1 to 6.

Images