Self-supervised vibration abnormity identification method and device for water pump unit

Abstract

The invention provides a self-supervised vibration anomaly recognition method and device for a water pump unit, and the method comprises the steps: obtaining vibration data during the normal operation of the water pump unit, and building a training data set based on the vibration data; inputting the training data set into the self-encoding network, and constructing an initial vibration anomaly recognition model based on the self-encoding network; constructing a loss function of the initial vibration anomaly recognition model, and performing self-supervised training on the initial vibration anomaly recognition model based on the training data set to obtain a vibration anomaly recognition model; obtaining to-be-detected water pump unit data, and inputting the to-be-detected water pump unit data into the vibration anomaly recognition model to obtain a vibration anomaly recognition score; and comparing the vibration anomaly identification score with the preset anomaly score threshold, outputting the operation state of the water pump unit, and completing the training of the vibration anomaly identification model through the vibration data of the water pump unit during normal shift operation, thereby reducing the model training cost, and improving the accuracy of the vibration anomaly identification model.

Description

Technical Field

[0001] The invention belongs to the technical field of water pump units, and in particular relates to a method and a device for identifying abnormal self-monitoring vibration of a water pump unit. Background Art

[0002] Water is one of the important resources for human production and life. Safe water supply is an important guarantee for people's daily life and the foundation of social and economic development. In the water supply process, water pump units, as common pressurizing equipment, can provide power for tap water transportation and are widely used in water plants, pressure stations and other water distribution nodes. Their safe and stable operation is an important guarantee for efficient and reliable water supply. Therefore, it is of great engineering value to detect the operating status of water pump units, identify abnormal operating status of water pump units in time, and avoid production safety accidents.

[0003] Regarding the operating status detection of water pump units, Chinese patent document 202410757413.8 provides "A method, system and storage medium for identifying the operating status of a water pump". This method establishes a water pump fault diagnosis model through dynamic simulation, convolutional neural network and transfer learning method to achieve water pump fault diagnosis. However, the above-mentioned literature method needs to obtain labeled fault data by establishing a water pump dynamics model, conducting fault simulation experiments, etc. in order to establish an effective fault diagnosis model. In actual engineering applications, due to the scarcity of faults, it is difficult and costly to obtain fault data. In addition, this method can only diagnose faults with relevant simulation or measured data, and cannot diagnose unknown faults that have not been reproduced in the dynamic model and fault simulation experiment, and it is difficult to apply to water pump units with complex structures. Summary of the invention

[0004] In order to overcome the deficiencies of the prior art, the present invention provides a method and device for identifying abnormal vibration of a water pump unit by self-monitoring, so as to solve the problems of high cost and inaccurate diagnosis results of abnormal vibration of a water pump unit in the prior art.

[0005] One embodiment of the present invention provides a method for identifying abnormal vibration of a water pump unit by self-supervision, comprising the following steps:

[0006] Acquire vibration data of the water pump unit during normal operation, and establish a training data set based on the vibration data;

[0007] Inputting the training data set into an autoencoder network, and constructing an initial vibration anomaly recognition model based on the autoencoder network;

[0008] Constructing a loss function of the initial vibration anomaly recognition model, and performing self-supervised training on the initial vibration anomaly recognition model based on the training data set to obtain a vibration anomaly recognition model;

[0009] Acquire the water pump unit data to be tested, and input the water pump unit data to be tested into the vibration abnormality recognition model to obtain a vibration abnormality recognition score;

[0010] The vibration anomaly recognition score is compared with a preset anomaly score threshold, and the operating status of the water pump unit is output.

[0011] In this embodiment, the present invention can complete the training of the model only through the vibration data during normal operation, which significantly reduces the data collection cost of establishing the vibration anomaly recognition model and shortens the model development cycle. The maximum mean difference is added to the loss function to guide the model to reconstruct the reconstructed samples consistent with the original input distribution, thereby improving the robustness of the model training process. The model is trained by self-supervised learning, and normal samples are preprocessed by data enhancement methods such as random masking and noise addition, so that the autoencoder network can reconstruct the original normal samples from the normal samples after random masking and noise addition, thereby improving the anti-interference performance of the model and avoiding false alarms.

[0012] In one embodiment, obtaining vibration data of a water pump unit during normal operation and establishing a training data set based on the vibration data include:

[0013] Vibration data of the water pump unit during normal operation is collected by a vibration sensor and a data acquisition device installed on the water pump unit, and a training data set is established based on the vibration data.

[0014] In one embodiment, the training data set is input into an autoencoder network, and an initial vibration anomaly recognition model based on the autoencoder network is constructed, including:

[0015] Performing data enhancement on the training data set as input samples, and inputting the enhanced input samples into the autoencoder network;

[0016] Extracting features from the enhanced input sample using an encoder to obtain a feature vector;

[0017] Obtaining a reconstructed sample of the enhanced input sample from the feature vector through a decoder;

[0018] A mean square error between the input sample and the reconstructed sample is calculated, and an initial vibration abnormality recognition model is constructed based on the mean square error as the vibration abnormality recognition score.

[0019] In one embodiment, constructing the loss function of the initial vibration anomaly recognition model includes:

[0020] Obtaining a mean square error and a maximum mean difference between the input sample and the reconstructed sample;

[0021] Determining the loss function based on the mean square error and the maximum mean difference;

[0022] The loss function determined based on the mean square error and the maximum mean difference is specifically determined by the following formula:

[0023]

[0024] Where L is the loss function, N is the number of samples, MSE is the mean square error, MMD is the maximum mean error, X={x 1 ,…,x N} is the input sample, To reconstruct the samples, α and β are preset values.

[0025] In one embodiment, the initial vibration anomaly recognition model is self-supervised trained based on the training data set to obtain the vibration anomaly recognition model, including:

[0026] Step S510, initializing the initial vibration abnormality recognition model;

[0027] Step S520, randomly collecting a training data subset from the training data set, and using the training data subset as the input sample;

[0028] Step S530, before the input sample is input into the encoder, randomly using several preset data enhancement methods to process the sample to obtain an enhanced sample, and using batch stochastic gradient descent to train the encoder and decoder until all samples in the training data subset traverse the preset data enhancement method;

[0029] Step S540, repeating steps S520-S530 until the entire training data set is traversed;

[0030] Step S550, repeating step S530 until the number of iterations reaches a preset number of iterations, thereby obtaining the vibration anomaly recognition model.

[0031] In one embodiment, the preset anomaly score threshold is obtained through a historical anomaly score sequence and a first formula, wherein the first formula is specifically:

[0032] ε t =μ(s 0～t-1 )+3σ(s 0～t-1 );

[0033] Among them, μ is the function for finding the mean, σ is the function for finding the standard deviation, and s 0～t-1 It is the historical anomaly score sequence from time 0 to time t-1.

[0034] In one embodiment, the vibration anomaly identification score is compared with a preset anomaly score threshold, and the operating status of the water pump unit is output, including:

[0035] When the vibration abnormality identification score is greater than the preset abnormality score threshold, the operation state of the water pump unit is abnormal vibration;

[0036] When the vibration abnormality identification score is less than or equal to the preset abnormality score threshold, the operating state of the water pump unit is normal vibration.

[0037] One embodiment of the present invention further provides a water pump unit self-supervision vibration abnormality identification device, comprising:

[0038] A training data set establishment module, used to obtain vibration data of the water pump unit during normal operation, and establish a training data set based on the vibration data;

[0039] An initial model building module, used for inputting the training data set into the autoencoder network and building an initial vibration anomaly recognition model based on the autoencoder network;

[0040] A self-supervised training module, used for constructing a loss function of the initial vibration anomaly recognition model, and performing self-supervised training on the initial vibration anomaly recognition model based on the training data set to obtain a vibration anomaly recognition model;

[0041] A score acquisition module, used for acquiring the water pump unit data to be tested, and inputting the water pump unit data to be tested into the vibration abnormality recognition model to obtain a vibration abnormality recognition score;

[0042] The comparison module is used to compare the vibration abnormality recognition score with a preset abnormality score threshold and output the operating status of the water pump unit.

[0043] In this embodiment, the advantages and beneficial effects of the water pump unit self-supervision vibration abnormality identification method have been explained above and will not be repeated here. Moreover, since the water pump unit self-supervision vibration abnormality identification device uses the water pump unit self-supervision vibration abnormality identification method, the water pump unit self-supervision vibration abnormality identification device also has the same advantages and beneficial effects.

[0044] One of the embodiments of the present invention further provides a readable storage medium, in which a computer program is stored. The computer program includes program instructions. When the program instructions are executed by a processor of an electronic device, the processor executes the steps of the self-supervisory vibration abnormality identification method for a water pump unit as described in any one of the above.

[0045] One of the embodiments of the present invention also provides an electronic device, including: a processor and a memory, the memory is used to store computer program code, the computer program code includes computer instructions, when the processor executes the computer instructions, the electronic device executes the steps of the water pump unit self-supervision vibration abnormality identification method as described in any one of the above. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. 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 the structures shown in these drawings without paying creative work.

[0047] Figure 1 A schematic flow chart of a method for identifying abnormal vibration of a water pump unit provided by one embodiment of the present invention;

[0048] Figure 2 A schematic diagram of a process for obtaining a reconstructed sample according to one embodiment of the present invention;

[0049] Figure 3 Another schematic diagram of a flow chart of a method for identifying abnormal vibration of a water pump unit provided by one embodiment of the present invention;

[0050] Figure 4 A schematic diagram of the working principle of an electronic device provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions 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 are within the scope of protection of the present invention.

[0052] It should be noted that if the embodiments of the present invention involve directional indications (such as up, down, left, right, front, back, etc.), the directional indications are only used to explain the relative position relationship, movement status, etc. between the components in a certain specific posture. If the specific posture changes, the directional indication will also change accordingly.

[0053] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present invention, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features limited to "first" and "second" may explicitly or implicitly include at least one of the features. In addition, if "and / or" or "and / or" appears in the full text, its meaning includes three parallel solutions. Taking "A and / or B" as an example, it includes solution A, solution B, or solutions that satisfy both A and B. In addition, the technical solutions between the various embodiments can be combined with each other, but it must be based on the ability of ordinary technicians in this field to implement. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0054] See also Figure 1-Figure 4 One embodiment of the present invention provides a method for identifying abnormal vibration of a water pump unit by self-supervision, comprising the following steps:

[0055] S100, obtaining vibration data of a water pump unit during normal operation, and establishing a training data set based on the vibration data.

[0056] In this embodiment, the vibration data during normal operation may be collected within a fixed continuous time, or may be multiple sets of vibration data collected within different time periods, and the collected vibration data during normal operation is established as a training data set.

[0057] In one embodiment, obtaining vibration data of a water pump unit during normal operation and establishing a training data set based on the vibration data include:

[0058] Vibration data of the water pump unit during normal operation is collected by a vibration sensor and a data acquisition device installed on the water pump unit, and a training data set is established based on the vibration data.

[0059] In this embodiment, the vibration sensor is arranged near the bearing seat of the water pump unit, and is used to convert the vibration signal during the operation of the unit into a voltage signal, and at the same time send the vibration signal to the data acquisition device to realize the collection process of vibration data during the normal operation of the water pump unit. It should be noted that technical personnel in this field can flexibly select different vibration data acquisition devices according to actual conditions, including but not limited to the vibration sensors and data acquisition devices mentioned in this embodiment.

[0060] S200: input the training data set into an autoencoder network, and construct an initial vibration anomaly recognition model based on the autoencoder network.

[0061] In this embodiment, an autoencoder network is used to construct a vibration anomaly recognition model. Therefore, first, an initial model is constructed by inputting a training data set into the autoencoder network, that is, an initial vibration anomaly recognition model based on the autoencoder network. This model is a preliminary recognition model and needs to be further optimized and adjusted before it can be put into vibration anomaly recognition work.

[0062] In one embodiment, the training data set is input into an autoencoder network, and an initial vibration anomaly recognition model based on the autoencoder network is constructed, including:

[0063] Performing data enhancement on the training data set as input samples, and inputting the enhanced input samples into the autoencoder network;

[0064] Extracting features from the enhanced input sample using an encoder to obtain a feature vector;

[0065] Obtaining a reconstructed sample of the enhanced input sample from the feature vector through a decoder;

[0066] A mean square error between the input sample and the reconstructed sample is calculated, and an initial vibration abnormality recognition model is constructed based on the mean square error as the vibration abnormality recognition score.

[0067] In this embodiment, the self-encoding network includes an encoder and a decoder, such as Figure 2 The figure shows a flow chart of obtaining a reconstructed sample from an input sample proposed in this embodiment. As can be seen from the figure, before the input sample is input to the encoder, the input sample needs to be data enhanced, where for the input sample x∈R n×m (n is the number of vibration signal channels, m is the length of the vibration signal), the model first uses the encoder E to extract the features of the sample x and obtains the feature z = E(x); then the decoder D is used to obtain the reconstruction of the input sample x from the feature z. Calculate the mean square error between the input sample x and the reconstructed sample as the anomaly score:

[0068]

[0069] It should be noted that in the process of self-supervised training, the training samples are data-enhanced before entering the encoder E, that is, DA stands for data augmentation.

[0070] S300 , constructing a loss function of the initial vibration anomaly recognition model, and performing self-supervised training on the initial vibration anomaly recognition model based on the training data set to obtain a vibration anomaly recognition model.

[0071] In this embodiment, the loss function or cost function is a function that maps the value of a random event or its related random variables to a non-negative real number to represent the "risk" or "loss" of the random event. In applications, the loss function is usually associated with the optimization problem as a learning criterion, that is, the model is solved and evaluated by minimizing the loss function. Specifically, the loss function has the following functions: Measuring model performance: The loss function is used to measure the difference between the model's prediction results and the true label. A good model should be able to minimize the loss, that is, the prediction result is as close to the actual label as possible, so as to complete the task more accurately. Training optimization: The training process of machine learning and deep learning models is an optimization problem, and the goal is to minimize the loss function. By continuously adjusting the parameters of the model, the loss function is gradually reduced, and the prediction ability of the model is gradually improved. Commonly used optimization algorithms such as gradient descent update model parameters by calculating the gradient of the loss function.

[0072] In one embodiment, constructing the loss function of the initial vibration anomaly recognition model includes:

[0073] Obtaining a mean square error and a maximum mean difference between the input sample and the reconstructed sample;

[0074] Determining the loss function based on the mean square error and the maximum mean difference;

[0075] The loss function determined based on the mean square error and the maximum mean difference is specifically determined by the following formula:

[0076]

[0077] Where L is the loss function, N is the number of samples, MSE is the mean square error, MMD is the maximum mean error, X={x 1 ,…,x N} is the input sample, To reconstruct the samples, α and β are preset values.

[0078] In this embodiment, the maximum mean difference is added to the loss function to guide the model to reconstruct reconstructed samples that are consistent with the original input distribution, thereby improving the robustness of the model training process to ensure that the reconstructed samples are consistent with the sample distribution of the original input and to ensure the accuracy of the model. Therefore, the reconstructed samples and the model are optimized and adjusted through the loss function to achieve the above purpose.

[0079] In one embodiment, the initial vibration anomaly recognition model is self-supervised trained based on the training data set to obtain the vibration anomaly recognition model, including:

[0080] Step S510, initializing the initial vibration abnormality recognition model;

[0081] Step S520, randomly collecting a training data subset from the training data set, and using the training data subset as the input sample;

[0082] Step S530, before the input sample is input into the encoder, randomly using several preset data enhancement methods to process the sample to obtain an enhanced sample, and using batch stochastic gradient descent to train the encoder and decoder until all samples in the training data subset traverse the preset data enhancement method;

[0083] Step S540, repeating steps S520-S530 until the entire training data set is traversed;

[0084] Step S550, repeating step S530 until the number of iterations reaches a preset number of iterations, thereby obtaining the vibration anomaly recognition model.

[0085] In this embodiment, several preset data enhancement methods include at least random masking, noise addition and other data enhancement methods. At the same time, in actual applications, technical personnel in this field can limit the traversal types of data enhancement methods according to actual conditions, and do not need to strictly follow the above methods.

[0086] S400, obtaining water pump unit data to be tested, and inputting the water pump unit data to be tested into the vibration abnormality recognition model to obtain a vibration abnormality recognition score.

[0087] In this embodiment, the water pump unit data to be tested is the real-time data of the current water pump unit. By importing the real-time data into the vibration anomaly recognition model, the vibration anomaly recognition score of the water pump unit can be automatically obtained, thereby realizing real-time monitoring of the water pump unit.

[0088] In one embodiment, the preset anomaly score threshold is obtained through a historical anomaly score sequence and a first formula, wherein the first formula is specifically:

[0089] ε t =μ(s 0～t-1 )+3σ(s 0～t-1 );

[0090] Among them, μ is the function for finding the mean, σ is the function for finding the standard deviation, and s 0～t-1 It is the historical anomaly score sequence from time 0 to time t-1.

[0091] S500: Compare the vibration abnormality recognition score with a preset abnormality score threshold, and output the operating status of the water pump unit.

[0092] In one embodiment, the vibration abnormality identification score is compared with a preset abnormality score threshold, and the operating status of the water pump unit is output, including:

[0093] When the vibration abnormality identification score is greater than the preset abnormality score threshold, the operation state of the water pump unit is abnormal vibration;

[0094] When the vibration abnormality identification score is less than or equal to the preset abnormality score threshold, the operating state of the water pump unit is normal vibration.

[0095] In this embodiment, there are two comparison results. When the vibration abnormality recognition score is greater than the preset abnormality score threshold, the operating state of the water pump unit is abnormal vibration. At this time, an alarm signal can be sent, or an abnormal picture or information can be displayed through the display device. On the contrary, the water pump unit operates normally.

[0096] In summary, if Figure 3 As shown, the present application obtains the measured input sample x at the tth moment. t , input the autoencoder network to obtain the reconstructed sample Calculate the input sample x t and reconstructed samples Anomaly score And the anomaly score threshold ε at time t t In contrast, if s t >ε t , then the water pump unit has abnormal vibration, otherwise, the water pump unit is operating normally.

[0097] This application has the following advantages:

[0098] (1) In the process of model training, the present invention does not require abnormal data to participate in the training, and only normal data can be used to complete the model training, which significantly reduces the data collection cost of establishing an intelligent vibration anomaly recognition model and shortens the model development cycle;

[0099] (2) The present invention adds the maximum mean difference to the loss function to guide the model to reconstruct samples that are consistent with the original input distribution, thereby improving the robustness of the model training process.

[0100] (3) The present invention utilizes a self-supervised learning training model and uses data enhancement methods such as random masking and noise addition to pre-process normal samples, so that the autoencoder network can reconstruct the original normal samples from the normal samples after random masking and noise addition, thereby improving the anti-interference performance of the model and avoiding false alarms.

[0101] One embodiment of the present invention further provides a water pump unit self-supervision vibration abnormality identification device, comprising:

[0102] A training data set establishment module, used to obtain vibration data of the water pump unit during normal operation, and establish a training data set based on the vibration data;

[0103] An initial model building module, used for inputting the training data set into the autoencoder network and building an initial vibration anomaly recognition model based on the autoencoder network;

[0104] A self-supervised training module, used for constructing a loss function of the initial vibration anomaly recognition model, and performing self-supervised training on the initial vibration anomaly recognition model based on the training data set to obtain a vibration anomaly recognition model;

[0105] A score acquisition module, used for acquiring the water pump unit data to be tested, and inputting the water pump unit data to be tested into the vibration abnormality recognition model to obtain a vibration abnormality recognition score;

[0106] The comparison module is used to compare the vibration abnormality recognition score with a preset abnormality score threshold and output the operating status of the water pump unit.

[0107] One of the embodiments of the present invention further provides a readable storage medium, in which a computer program is stored. The computer program includes program instructions. When the program instructions are executed by a processor of an electronic device, the processor executes the steps of the self-supervisory vibration abnormality identification method for a water pump unit as described in any one of the above.

[0108] One of the embodiments of the present invention also provides an electronic device, including: a processor and a memory, the memory is used to store computer program code, the computer program code includes computer instructions, when the processor executes the computer instructions, the electronic device executes the steps of the water pump unit self-supervision vibration abnormality identification method as described in any one of the above.

[0109] See also Figure 4 , Figure 4 A schematic diagram of the hardware structure of an electronic device provided by an embodiment of the present invention.

[0110] The electronic device 2 includes a processor 21, a memory 22, an input device 23, and an output device 24. The processor 21, the memory 22, the input device 23, and the output device 24 are coupled via a connector, and the connector includes various interfaces, transmission lines, or buses, etc., which are not limited in the embodiments of the present invention. It should be understood that in various embodiments of the present invention, coupling refers to mutual connection in a specific manner, including direct connection or indirect connection through other devices, for example, through various interfaces, transmission lines, buses, etc.

[0111] The processor 21 may be a combination of one or more central processing units (CPUs) and graphics processing units (GPUs). When the processor 21 is a CPU, the GPU may be a single-core CPU or a multi-core CPU. Optionally, the processor 21 may be a processor group consisting of multiple CPUs and GPUs, and the multiple processors are coupled to each other via one or more buses. Optionally, the processor may also be other types of processors, etc., which are not limited in the embodiments of the present invention.

[0112] The memory 22 can be used to store computer program instructions and various computer program codes including program codes for executing the scheme of the present invention. Optionally, the memory includes but is not limited to random access memory (RAM), read-only memory (ROM), erasable programmable read only memory (EPROM), or portable read only memory (CD-ROM), which is used for related instructions and data.

[0113] The input device 23 is used to input data and / or signals, and the output device 24 is used to output data and / or signals. The output device 23 and the input device 24 can be independent devices or an integrated device.

[0114] It is understandable that in the embodiment of the present invention, the memory 22 is not only used to store related instructions, and the embodiment of the present invention does not limit the specific data stored in the memory.

[0115] Understandably, Figure 4 Only a simplified design of an electronic device is shown. In practical applications, the electronic device may also include other necessary components, including but not limited to any number of input / output devices, processors, memories, etc., and all video analysis devices that can implement the embodiments of the present invention are within the protection scope of the present invention.

[0116] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0117] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. Those skilled in the art can also clearly understand that the descriptions of the various embodiments of the present invention have different focuses. For the convenience and brevity of description, the same or similar parts may not be repeated in different embodiments. Therefore, for parts not described or not described in detail in a certain embodiment, refer to the records of other embodiments.

[0118] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0119] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0120] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0121] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented by software, it can be implemented in whole or in part in the form of a computer program to be sealed. The computer program to be sealed includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted through the computer-readable storage medium. The computer instructions can be transmitted from a website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (digital subscriber line, DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media integrations. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a digital versatile disc (DVD)), or a semiconductor medium (eg, a solid state disk (SSD)).

[0122] A person skilled in the art can understand that to implement all or part of the processes in the above-mentioned embodiments, the processes can be completed by a computer program to instruct the relevant hardware, and the program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above-mentioned method embodiments. The aforementioned storage medium includes: a read-only memory (ROM) or a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0123] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. All equivalent structural changes made by using the contents of the present invention specification and drawings under the inventive concept of the present invention, or directly / indirectly applied in other related technical fields are included in the patent protection scope of the present invention.

Claims

1. A method for identifying abnormal vibration of a water pump unit by self-supervision, characterized in that: The following steps are involved: Acquire vibration data of the water pump unit during normal operation, and establish a training data set based on the vibration data; Inputting the training data set into an autoencoder network, and constructing an initial vibration anomaly recognition model based on the autoencoder network; Constructing a loss function of the initial vibration anomaly recognition model, and performing self-supervised training on the initial vibration anomaly recognition model based on the training data set to obtain a vibration anomaly recognition model; Acquire the water pump unit data to be tested, and input the water pump unit data to be tested into the vibration abnormality recognition model to obtain a vibration abnormality recognition score; The vibration anomaly recognition score is compared with a preset anomaly score threshold, and the operating status of the water pump unit is output.

2. The method for identifying abnormal vibration of a pump unit according to claim 1, characterized in that: Obtaining vibration data of the water pump unit during normal operation, and establishing a training data set based on the vibration data, including: Vibration data of the water pump unit during normal operation is collected by a vibration sensor and a data acquisition device installed on the water pump unit, and a training data set is established based on the vibration data.

3. The method for identifying abnormal vibration of a pump unit according to claim 2, characterized in that: The training data set is input into the autoencoder network, and an initial vibration anomaly recognition model based on the autoencoder network is constructed, including: Performing data enhancement on the training data set as input samples, and inputting the enhanced input samples into the autoencoder network; Extracting features from the enhanced input sample using an encoder to obtain a feature vector; Obtaining a reconstructed sample of the enhanced input sample from the feature vector through a decoder; A mean square error between the input sample and the reconstructed sample is calculated, and an initial vibration abnormality recognition model is constructed based on the mean square error as the vibration abnormality recognition score.

4. The method for identifying abnormal vibration of a pump unit according to claim 3, characterized in that: Constructing the loss function of the initial vibration anomaly recognition model includes: Obtaining a mean square error and a maximum mean difference between the input sample and the reconstructed sample; Determining the loss function based on the mean square error and the maximum mean difference; The loss function determined based on the mean square error and the maximum mean difference is specifically determined by the following formula: Where L is the loss function, N is the number of samples, MSE is the mean square error, MMD is the maximum mean error, X = {x1, ..., x N } is the input sample, To reconstruct the samples, α and β are preset values.

5. The method for identifying abnormal vibration of a pump unit according to claim 4, characterized in that: The initial vibration anomaly recognition model is self-supervised trained based on the training data set to obtain a vibration anomaly recognition model, including: Step S510, initializing the initial vibration abnormality recognition model; Step S520, randomly collecting a training data subset from the training data set, and using the training data subset as the input sample; Step S530, before the input sample is input into the encoder, randomly using several preset data enhancement methods to process the sample to obtain an enhanced sample, and using batch stochastic gradient descent to train the encoder and decoder until all samples in the training data subset traverse the preset data enhancement method; Step S540, repeating steps S520-S530 until the entire training data set is traversed; Step S550, repeating step S530 until the number of iterations reaches a preset number of iterations, thereby obtaining the vibration anomaly recognition model.

6. The method for identifying abnormal vibration of a pump unit according to claim 1, characterized in that: The preset anomaly score threshold is obtained through the historical anomaly score sequence and the first formula, and the first formula is specifically: e t =μ(s 0～t-1 )+3σ(s 0～t-1 ); Among them, μ is the function for finding the mean, σ is the function for finding the standard deviation, and s 0～t-1 It is the historical anomaly score sequence from time 0 to time t-1.

7. The method for identifying abnormal vibration of a pump unit according to claim 1, characterized in that: The vibration anomaly identification score is compared with a preset anomaly score threshold, and the operating status of the water pump unit is output, including: When the vibration abnormality identification score is greater than the preset abnormality score threshold, the operation state of the water pump unit is abnormal vibration; When the vibration abnormality identification score is less than or equal to the preset abnormality score threshold, the operating state of the water pump unit is normal vibration.

8. A water pump unit self-monitoring vibration abnormality identification device, characterized in that: include: A training data set establishment module, used to obtain vibration data of the water pump unit during normal operation, and establish a training data set based on the vibration data; An initial model building module, used for inputting the training data set into the autoencoder network and building an initial vibration anomaly recognition model based on the autoencoder network; A self-supervised training module, used for constructing a loss function of the initial vibration anomaly recognition model, and performing self-supervised training on the initial vibration anomaly recognition model based on the training data set to obtain a vibration anomaly recognition model; A score acquisition module, used for acquiring the water pump unit data to be tested, and inputting the water pump unit data to be tested into the vibration abnormality recognition model to obtain a vibration abnormality recognition score; The comparison module is used to compare the vibration abnormality recognition score with a preset abnormality score threshold and output the operating status of the water pump unit.

9. An electronic device, comprising: A processor and a memory, characterized in that the memory is used to store computer program codes, the computer program codes include computer instructions, and when the processor executes the computer instructions, the electronic device executes the steps of the method for identifying self-supervisory vibration anomalies of a pump unit as described in any one of claims 1-7.

10. A readable storage medium having a computer program stored therein, characterized in that: The computer program includes program instructions, and when the program instructions are executed by a processor of an electronic device, the processor executes the steps of the method for identifying abnormal vibration of a water pump unit self-supervising as described in any one of claims 1-7.

Images