Abnormal monitoring method and device for multi-channel lubricating oil abrasive particles and storage medium

Abstract

The invention discloses a multi-channel lubricating oil abrasive particle abnormity monitoring method and device and a storage medium. The method comprises the following steps: constructing a first auto-encoder and a second auto-encoder; historical health data of the multi-channel lubricating oil abrasive particles are obtained; training a first auto-encoder and a second auto-encoder by using the historical health data; optimizing the trained first auto-encoder and second auto-encoder based on an adversarial learning strategy; and carrying out abnormity monitoring on the multi-channel lubricating oil abrasive particles by utilizing the optimized combined model of the first auto-encoder and the second auto-encoder. According to the built double-auto-encoder combined model, historical health data can be fully utilized, complex features hidden in the historical health data can be deeply excavated, through adversarial learning training, the performance of the double-auto-encoder model is further improved, high-performance multi-channel abrasive particle abnormity monitoring is achieved, and the monitoring accuracy and reliability are effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of lubricating oil state recognition, and particularly relates to an abnormal monitoring method, device and storage medium for multi-channel lubricating oil abrasive particles. Background Art

[0002] Lubricating oil plays a crucial role in mechanical equipment. It is similar to the "blood" of mechanical equipment and contains rich information about the operating state of the equipment. Among them, the abrasive particle information is directly related to the wear state of mechanical equipment. Therefore, real-time monitoring of the abrasive particle information in lubricating oil is of great significance for understanding and evaluating the wear state of mechanical equipment. In the context of the current wave of artificial intelligence and the digital transformation of enterprises, there is an urgent need for a method that can intelligently and automatically monitor the abrasive particle information in lubricating oil. The existing monitoring methods mainly apply fixed thresholds directly to the original data for monitoring, and this method has problems of false alarms or missed alarms. At the same time, a large amount of historical health data has been deposited at the engineering site and not fully utilized. These data contain rich information and will play an important role in health management if reasonably utilized.

[0003] The existing patent CN117892190A discloses a method and system for unmanned aerial vehicle health assessment based on a multi-channel model. The method includes: S1, determining the input parameters of the unmanned aerial vehicle engine health assessment model; S2, collecting the sensing data of the unmanned aerial vehicle engine under the working state, performing standardization and structuring processing to form a data set; S3, establishing a multi-channel neural network model; S4, training the multi-channel neural network model; S5, using the multi-channel neural network model to evaluate the health state of the unmanned aerial vehicle.

[0004] The existing patent CN118094284A discloses a method, device and storage medium for journal anomaly detection. The method includes the steps of: collecting normal journal and abnormal journal data to obtain a data set; performing a first processing on the data set to obtain first-processed data; performing a second processing on the first pre-processed data to obtain second-processed data, and dividing the second-processed data into a training set and a test set; the second processing adopts the combination of multiple sampling techniques; constructing a compound autoencoder model and introducing an attention layer therein; performing deep training on the autoencoder model by using an adaptive regularization method and the training set to obtain a trained model; using the test set and the AUC-PRC index to evaluate the performance of the trained model.

[0005] In summary, the above two existing patents have not solved the problems of false alarms or missed alarms that occur when directly applying fixed thresholds to the original data in the technical field of lubricating oil state recognition. Summary of the Invention

[0006] Based on the above technical problems, the present invention proposes an abnormal monitoring method, device and storage medium for multi-channel lubricating oil abrasives, which solves the problems of false alarms or missed alarms existing in the prior art when directly applying a fixed threshold to the original data for monitoring.

[0007] An abnormal monitoring method for multi-channel lubricating oil abrasives, the method includes:

[0008] Construct a first autoencoder and a second autoencoder;

[0009] Obtain the historical health data of multi-channel lubricating oil abrasives;

[0010] Use the historical health data to train the first autoencoder and the second autoencoder;

[0011] Optimize the trained first autoencoder and second autoencoder based on an adversarial learning strategy;

[0012] Use the combined model of the optimized first autoencoder and second autoencoder to perform abnormal monitoring on multi-channel lubricating oil abrasives.

[0013] Further, both the first autoencoder and the second autoencoder include: an encoder and a decoder, and the encoder and the decoder are used to extract features from multi-channel lubricating oil abrasive data.

[0014] Further, both the first autoencoder and the second autoencoder adopt a fully connected neural network.

[0015] Further, using the historical health data to train the first autoencoder and the second autoencoder includes:

[0016] Use the historical health data to optimize the internal parameters of the first autoencoder and the second autoencoder according to a preset loss function.

[0017] Further, the loss function corresponding to the first autoencoder is: The loss function corresponding to the second autoencoder is: Wherein, W represents the historical health data, AE A (W) represents the data reconstructed by W after passing through the first autoencoder, AE B (W) represents the data reconstructed by W after passing through the second autoencoder, λ and γ are regularization term coefficients, ||AE A ||2, ||AE B ||2 are the regularization terms of the first encoder and the second encoder respectively.

[0018] Further, optimizing the trained first autoencoder and second autoencoder based on an adversarial learning strategy includes:

[0019] Optimize the trained first autoencoder and second autoencoder according to the loss function based on the adversarial learning strategy. The loss function corresponding to the trained first autoencoder is The loss function corresponding to the trained second autoencoder is

[0020] Furthermore, use the combined model of the optimized first autoencoder and second autoencoder to perform anomaly monitoring on the multi-channel lubricating oil wear particles, including:

[0021] Obtain the real-time operation data of the multi-channel lubricating oil wear particles;

[0022] Use the combined model of the optimized first autoencoder and second autoencoder to determine the anomaly score of the real-time operation data through Formula 1. Formula 1 is where W′ represents the real-time operation data, AE A (W′) represents the data after W′ is reconstructed by the first autoencoder, AE B (AE A (W′)) represents the result after AE A (W′) is reconstructed by the second autoencoder, and α and β are coefficients.

[0023] An anomaly monitoring device for multi-channel lubricating oil wear particles, including:

[0024] A construction module for constructing a first autoencoder and a second autoencoder;

[0025] An acquisition module for acquiring the historical health data of the multi-channel lubricating oil wear particles;

[0026] A training module for training the first autoencoder and the second autoencoder using the historical health data;

[0027] An optimization module for optimizing the trained first autoencoder and second autoencoder based on the adversarial learning strategy;

[0028] An anomaly monitoring module for performing anomaly monitoring on the multi-channel lubricating oil wear particles using the combined model of the optimized first autoencoder and second autoencoder.

[0029] A computer-readable storage medium, characterized in that the computer-readable storage medium includes a stored computer program, wherein the computer program can implement the above method when run by an electronic device.

[0030] A computer program product, including a computer program, which implements the steps of the above method when executed by a processor.

[0031] An electronic device includes a memory and a processor, wherein a computer program is stored in the memory and the processor is configured to execute the above method through the computer program.

[0032] Based on the above technical solution, the present invention has at least the following beneficial effects:

[0033] The dual autoencoder combination model constructed by the present invention can make full use of historical health data, deeply explore the complex features hidden therein, effectively integrate multi-channel wear data, and reduce information conflicts. This model architecture not only re-endows a large amount of historical health data with value and application potential, but also further enhances the learning ability of the model through adversarial learning strategies, making it no longer overly sensitive to noise and changes in the input data, so as to better adapt to unknown data. Through adversarial learning training, the performance of the dual autoencoder model can be further improved, realizing high-performance multi-channel wear anomaly monitoring, and effectively improving the accuracy and reliability of monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The accompanying drawings constituting a part of the present invention are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the accompanying drawings:

[0035] Figure 1 A flow chart of a method for monitoring abnormality of multi-channel lubricating oil wear particles in one embodiment of the present invention;

[0036] Figure 2 It is a schematic diagram of autoencoder reconstruction learning training of a multi-channel lubricating oil wear particle abnormality monitoring method in the present invention;

[0037] Figure 3 A schematic diagram of dual encoder adversarial training of a multi-channel lubricating oil wear particle abnormality monitoring method in the present invention;

[0038] Figure 4 A schematic diagram of abnormality score calculation of a multi-channel lubricating oil wear particle abnormality monitoring method in the present invention;

[0039] Figure 5 It is an operation data diagram of a multi-channel lubricating oil wear particle abnormality monitoring method in the present invention;

[0040] Figure 6 It is a schematic diagram of 3sigma threshold monitoring of optimized algorithm operation data of a multi-channel lubricating oil wear particle abnormality monitoring method in the present invention;

[0041] Figure 7 A schematic diagram of abnormal scores of operating data obtained by a multi-channel lubricating oil wear particle abnormal monitoring method of the present invention;

[0042] Figure 8Schematic diagram of an abnormal monitoring device for multi-channel lubricating oil abrasive particles in an embodiment of the present invention;

[0043] Figure 9 Block diagram of the computer system structure of an electronic device for implementing the embodiments of the present application;

[0044] Figure 10 Schematic diagram of an electronic device for abnormal monitoring of multi-channel lubricating oil abrasive particles. Detailed implementation manners

[0045] It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments may be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0046] The following further describes the present invention in detail with reference to specific embodiments, and these embodiments should not be construed as limiting the scope claimed by the present invention.

[0047] Embodiment

[0048] To solve the problems of false alarms or missed alarms existing in the prior art when directly applying fixed thresholds to the original data for monitoring, the present invention proposes an abnormal monitoring method, device and storage medium for multi-channel lubricating oil abrasive particles. In the present solution, the method and device construct a dual encoder joint driving model to make full use of historical health data and mine the hidden features of historical data; the adversarial learning strategy can further improve the learning ability of the model, enabling it to obtain better robustness and generalization ability; the dual encoder joint driving model based on the adversarial learning strategy can enhance the advantages of the autoencoder model and achieve more accurate multi-channel abrasive particle abnormal monitoring.

[0049] According to one aspect of the embodiments of the present application, an abnormal monitoring method for multi-channel lubricating oil abrasive particles is provided. As Figure 1 shown, the above method includes the following steps:

[0050] S1, construct a first autoencoder and a second autoencoder.

[0051] Further, both the first autoencoder and the second autoencoder include: an encoder and a decoder, and the encoder and the decoder are used for feature extraction of multi-channel lubricating oil abrasive particle data. Further, both the first autoencoder and the second autoencoder adopt a fully connected neural network to implement encoding and decoding of data. Specifically, the encoder consists of multiple layers. From the input layer to the output layer, the number of nodes in each layer decreases; the decoder also consists of multiple layers. From the input layer to the output layer, the number of nodes in each layer increases. It should be noted that the number of nodes in the output layer of the encoder is the same as the number of nodes in the input layer of the decoder to ensure the complete transmission and restoration of information.

[0052] S2. Obtain the historical health data of multi-channel lubricating oil abrasive particles.

[0053] S3. Use the historical health data to train the first autoencoder and the second autoencoder.

[0054] In an embodiment of the present invention, using the historical health data to train the first autoencoder and the second autoencoder includes: using the historical health data to optimize the internal parameters of the first autoencoder and the second autoencoder according to a preset loss function. Through this step, the first autoencoder and the second autoencoder can have good data reconstruction capabilities. As Figure 2 shows the process of minimizing the error between the encoder input and output results, so that the first autoencoder A and the second autoencoder B have good historical health data reconstruction capabilities.

[0055] Furthermore, the loss function corresponding to the first autoencoder is: The loss function corresponding to the second autoencoder is: where W represents the historical health data, AE A (W) represents the data after W is reconstructed by the first autoencoder, AE B (W) represents the data after W is reconstructed by the second autoencoder, λ and γ are regularization term coefficients, ||AE A ||2, ||AE B ||2 are the regularization terms of the first encoder and the second encoder respectively. As Figure 3 shows the process of adversarial training of the two autoencoders. During the training process, the historical health data reconstruction ability of the first autoencoder A becomes stronger and stronger, and the discrimination ability of the second autoencoder B also becomes stronger and stronger.

[0056] S4. Optimize the trained first autoencoder and second autoencoder based on the adversarial learning strategy.

[0057] The autoencoder can learn the complex features of the historical health data, fully excavate and utilize a large amount of existing historical health data, and make it valuable and applicable again. The goal of the autoencoder is to learn a compact representation of the data, but this representation may be too sensitive to noise and changes in the input data. By introducing the idea of adversarial learning, the autoencoder can learn a more robust feature representation and better adapt to unknown data at the same time.

[0058] In an embodiment of the present invention, optimizing the trained first autoencoder and second autoencoder based on the adversarial learning strategy includes: optimizing the trained first autoencoder and second autoencoder according to the loss function of the adversarial learning strategy.

[0059] Specifically, the loss function corresponding to the trained first autoencoder in this embodiment is The loss function corresponding to the trained second autoencoder is It should be understood that the optimization of the trained first autoencoder and second autoencoder based on the adversarial learning strategy is not limited to the method provided in this embodiment.

[0060] S5. Use the combined model of the optimized first autoencoder and second autoencoder to perform anomaly monitoring on multi-channel lubricating oil abrasive particles.

[0061] Furthermore, using the combined model of the optimized first autoencoder and second autoencoder to perform anomaly monitoring on multi-channel lubricating oil abrasive particles includes the following process:

[0062] S501. Obtain the real-time operation data of multi-channel lubricating oil abrasive particles.

[0063] S502. Use the combined model of the optimized first autoencoder and second autoencoder to determine the anomaly score of the real-time operation data through Formula 1.

[0064] Load the real-time operation data of the multi-channel lubricating oil abrasive particles, and then use the combined model of the optimized first autoencoder and second autoencoder to determine the anomaly score of the real-time operation data through Formula 1. Specifically, the formula adopted in this embodiment is as follows: Among them, W′ represents the real-time operation data, AE A (W′) represents the data after W′ is reconstructed by the first autoencoder, AE B (AE A (W′)) represents the result after AE A (W′) is reconstructed by the second autoencoder, and α and β are coefficients, between 0 and 1. As Figure 4 shows the process of finally calculating the anomaly score according to the operation data.

[0065] As Figure 5 shows the operation data of a certain channel of the abrasive particles, and some data have obvious anomalies. As Figure 6 shows the use of the 3sigma method to perform anomaly monitoring on the operation data, and its effect is relatively poor, with a large number of false alarms. As Figure 7 shows the use of the 3sigma method to perform anomaly monitoring on the anomaly score of the operation data, and the effect is very good, reducing a large number of false alarms. This method effectively fuses the multi-channel abrasive particle data by using the dual autoencoder network, and can reduce information conflict. At the same time, the adversarial learning training strategy can further improve the performance of the dual autoencoder model, effectively improve the model performance, and achieve high-reliability and high-performance multi-channel abrasive particle anomaly monitoring.

[0066] It should be noted that for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that this application is not limited by the described action sequence, because according to this application, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0067] According to another aspect of the embodiments of the present application, the present invention also provides an abnormal monitoring device for multi-channel lubricating oil abrasive particles. As Figure 8 shown, the device includes: a construction module 801, an acquisition module 802, a training module 803, an optimization module 804, and an abnormal monitoring module 805. The functions of each module will be introduced below.

[0068] The construction module 801 is used to construct a first autoencoder and a second autoencoder.

[0069] The acquisition module 802 is used to acquire historical health data of multi-channel lubricating oil abrasive particles.

[0070] The training module 803 is used to train the first autoencoder and the second autoencoder by using the historical health data.

[0071] The optimization module 804 is used to optimize the trained first autoencoder and second autoencoder based on an adversarial learning strategy.

[0072] The abnormal monitoring module 805 is used to perform abnormal monitoring on multi-channel lubricating oil abrasive particles by using the combined model of the optimized first autoencoder and second autoencoder.

[0073] As an optional solution, the above device is also used for: both the first autoencoder and the second autoencoder include: an encoder and a decoder, and the encoder and the decoder are used to extract features from multi-channel lubricating oil abrasive particle data.

[0074] As an optional solution, the above device is also used for: both the first autoencoder and the second autoencoder adopt a fully connected neural network.

[0075] As an optional solution, the above device is also used for: training the first autoencoder and the second autoencoder by using the historical health data, including:

[0076] Using the historical health data to optimize the internal parameters of the first autoencoder and the second autoencoder according to a preset loss function.

[0077] As an optional solution, the loss function corresponding to the first autoencoder is: The loss function corresponding to the second autoencoder is as follows: Where W represents historical health data, and AE A (W) represents the data after W is reconstructed by the first autoencoder, and AE B (W) represents the data after W is reconstructed by the second autoencoder. λ and γ are regularization term coefficients, and ||AE A ||2 and ||AE B ||2 are the regularization terms of the first encoder and the second encoder respectively.

[0078] As an alternative solution, the above device is also used to optimize the trained first autoencoder and second autoencoder based on an adversarial learning strategy, including:

[0079] Optimize the trained first autoencoder and second autoencoder according to the loss function of the adversarial learning strategy. The loss function corresponding to the trained first autoencoder is The loss function corresponding to the trained second autoencoder is

[0080] As an alternative solution, the above device is also used to perform anomaly monitoring on multi-channel lubricating oil wear particles by using the combined model of the optimized first autoencoder and second autoencoder, including:

[0081] Obtain the real-time operation data of multi-channel lubricating oil wear particles;

[0082] Use the combined model of the optimized first autoencoder and second autoencoder to determine the anomaly score of the real-time operation data through Formula 1. Formula 1 is Where W′ represents the real-time operation data, and AE A (W′) represents the data after W′ is reconstructed by the first autoencoder, and AE B (AE A (W′)) represents the result after AE A (W′) is reconstructed by the second autoencoder. α and β are coefficients.

[0083] In the embodiments of the present application, the term "module" or "unit" refers to a computer program with a predetermined function or a part of a computer program, which works together with other related parts to achieve a predetermined goal, and can be fully or partially implemented by using software, hardware (such as a processing circuit or a memory), or a combination thereof. Similarly, one processor (or multiple processors or memories) can be used to implement one or more modules or units. In addition, each module or unit can be a part of the overall module or unit that includes the function of the module or unit.

[0084] Regarding the device in the above embodiments, the specific manner in which each module performs operations has been described in detail in the embodiments related to the method, and will not be elaborated here.

[0085] According to one aspect of the present application, there is provided a computer program product, which includes a computer program.

[0086] The serial numbers of the embodiments of the present application above are only for description and do not represent the superiority or inferiority of the embodiments.

[0087] Figure 9 The block diagram of the computer system of the electronic device for implementing the embodiments of the present application is shown.

[0088] It should be noted that Figure 9 The computer system 1100 of the electronic device shown is only an example and should not impose any limitations on the functions and usage scope of the embodiments of the present application.

[0089] As Figure 9 shown, the computer system 1100 includes a central processing unit 1101 (Central Processing Unit, CPU), which can perform various appropriate actions and processes according to the program stored in the read-only memory 1102 (Read-Only Memory, ROM) or the program loaded from the storage section 1108 into the random access memory 1103 (Random Access Memory, RAM). In the random access memory 1103, various programs and data required for system operation are also stored. The central processing unit 1101, the read-only memory 1102, and the random access memory 1103 are connected to each other via a bus 1104. The input / output interface 1105 (Input / Output interface, i.e., I / O interface) is also connected to the bus 1104.

[0090] The following components are connected to the input / output interface 1105: an input section 1106 including a keyboard, a mouse, etc.; an output section 1107 including such as a cathode ray tube (Cathode Ray Tube, CRT), a liquid crystal display (Liquid Crystal Display, LCD), etc. and a speaker, etc.; a storage section 1108 including a hard disk, etc.; and a communication section 1109 including a network interface card such as a local area network card, a modem, etc. The communication section 1109 performs communication processing via a network such as the Internet. A drive 1110 is also connected to the input / output interface 1105 as needed. A removable medium 1111, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 1110 as needed so that the computer program read from it can be installed into the storage section 1108 as needed.

[0091] In particular, according to the embodiments of the present application, the processes described in each method flowchart can be implemented as computer software programs. For example, the embodiments of the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program codes for executing the methods shown in the flowcharts. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 1109, and / or installed from the removable medium 1111. When the computer program is executed by the central processing unit 1101, various functions defined in the system of the present application are executed.

[0092] In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 1109, and / or installed from the removable medium 1111. When the computer program is executed by the central processing unit 1101, various functions provided by the embodiments of the present application are executed.

[0093] According to another aspect of the embodiments of the present application, an electronic device for abnormal monitoring of multi-channel lubricating oil abrasive particles is further provided. In this embodiment, the electronic device is taken as an example of a terminal device for illustration. As Figure 10 shown, the electronic device includes a memory 1202 and a processor 1204. A computer program is stored in the memory 1202, and the processor 1204 is configured to execute the steps in any one of the above method embodiments through the computer program.

[0094] Optionally, in this embodiment, the above electronic device may be at least one network device among multiple network devices in a computer network.

[0095] Optionally, in this embodiment, the above processor may be configured to execute the methods in the various embodiments of the present application through a computer program.

[0096] Optionally, those of ordinary skill in the art can understand that Figure 10 the structure shown is only schematic, Figure 10 and it does not limit the structure of the above electronic device. For example, the electronic device may further include more or fewer components (such as a network interface, etc.) than those shown in Figure 10 , or have a different configuration from that shown in Figure 10 .

[0097] Among them, the memory 1202 can be used to store software programs and modules, such as the program instructions / modules corresponding to the abnormal monitoring method and device of multi-channel lubricating oil abrasive particles in the embodiments of the present application. The processor 1204 executes various functional applications and data processing by running the software programs and modules stored in the memory 1202, that is, implements the above-mentioned abnormal monitoring method of multi-channel lubricating oil abrasive particles. The memory 1202 can include a high-speed random access memory, and can also include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some instances, the memory 1202 can further include a memory remotely set relative to the processor 1204, and these remote memories can be connected to the terminal through a network. Examples of the above network include but are not limited to the Internet, enterprise intranet, local area network, mobile communication network, and their combinations. Among them, the memory 1202 can specifically but not limitedly be used to store real-time running data information. As an example, as Figure 10 shown, the above-mentioned memory 1202 can include but is not limited to the construction module 801, acquisition module 802, training module 803, optimization module 804, and abnormal monitoring module 805 in the above-mentioned abnormal monitoring device of multi-channel lubricating oil abrasive particles. In addition, it can also include but is not limited to other module units in the above-mentioned device, which will not be elaborated in this example.

[0098] Optionally, the above-mentioned transmission device 1206 is used to receive or send data via a network. Specific examples of the above-mentioned network can include a wired network and a wireless network. In one instance, the transmission device 1206 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices and routers through a network cable, so as to communicate with the Internet or local area network. In one instance, the transmission device 1206 is a radio frequency (Radio Frequency, RF) module, which is used to communicate with the Internet wirelessly.

[0099] In addition, the above-mentioned electronic device further includes: a display 1208, which is used to display the above-mentioned abnormal monitoring results; and a connection bus 1210, which is used to connect each module component in the above-mentioned electronic device.

[0100] In other embodiments, the above-mentioned terminal device or server can be a node in a distributed system. Among them, the distributed system can be a blockchain system, and the blockchain system can be a distributed system formed by connecting the multiple nodes through network communication. Among them, the nodes can form a peer-to-peer network, and any form of computing device, such as an electronic device such as a server or a terminal, can become a node in the blockchain system by joining the peer-to-peer network.

[0101] According to an aspect of the present application, there is provided a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the electronic device executes an abnormal monitoring method for multi-channel lubricating oil abrasive particles provided in various optional implementations of the above aspect.

[0102] Optionally, in this embodiment, the above computer-readable storage medium may be configured to store instructions for executing the methods in the embodiments of the present application.

[0103] Optionally, in this embodiment, those of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by a program instructing the relevant hardware of the terminal device. The program can be stored in a computer-readable storage medium, and the storage medium may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disc, etc.

[0104] The serial numbers of the above embodiments of the present application are only for description and do not represent the superiority or inferiority of the embodiments.

[0105] If the integrated unit in the above embodiments is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in the above computer-readable storage medium. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing one or more electronic devices to execute all or part of the steps of the methods described in the various embodiments of the present application.

[0106] In the above embodiments of the present application, the descriptions of the various embodiments have their own emphases. For the parts not detailed in a certain embodiment, reference may be made to the relevant descriptions of other embodiments.

[0107] In several embodiments provided by the present application, it should be understood that the disclosed application program can be implemented in other ways. Among them, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, 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 displayed or discussed mutual coupling or direct coupling or communication connection can be through some interfaces. The indirect coupling or communication connection of units or modules can be in an electrical or other form.

[0108] 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.

[0109] In addition, each functional unit in each embodiment of the present application 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. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0110] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0111] In summary, it can be seen from the above description that the above embodiments of the present invention achieve the following technical effects:

[0112] The dual autoencoder combination model constructed by the present invention can make full use of historical health data, deeply explore the complex features hidden therein, effectively integrate multi-channel wear data, and reduce information conflicts. This model architecture not only re-endows a large amount of historical health data with value and application potential, but also further enhances the learning ability of the model through adversarial learning strategies, making it no longer overly sensitive to noise and changes in the input data, so as to better adapt to unknown data. Through adversarial learning training, the performance of the dual autoencoder model can be further improved, realizing high-performance multi-channel wear anomaly monitoring, and effectively improving the accuracy and reliability of monitoring.

[0113] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0114] It should be noted that in the description of this specification, the description referring to terms such as "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

Claims

1. An abnormal monitoring method for multi-channel lubricating oil abrasive particles, characterized in that, Including: Construct the first autoencoder and the second autoencoder; Obtain the historical health data of multi-channel lubricating oil abrasive particles; Use the historical health data to train the first autoencoder and the second autoencoder; Optimize the trained first autoencoder and second autoencoder based on the adversarial learning strategy; Use the combined model of the optimized first autoencoder and second autoencoder to perform anomaly monitoring on the multi-channel lubricating oil abrasive particles.

2. The method according to claim 1, characterized in that, Both the first autoencoder and the second autoencoder include: an encoder and a decoder, and the encoder and the decoder are used to extract features from multi-channel lubricating oil abrasive particle data.

3. The method according to claim 1, characterized in that, Both the first autoencoder and the second autoencoder adopt fully connected neural networks.

4. The method according to claim 1, wherein Using the historical health data to train the first autoencoder and the second autoencoder includes: Using the historical health data to optimize the internal parameters of the first autoencoder and the second autoencoder according to a preset loss function.

5. The method according to claim 4, wherein The loss function corresponding to the first autoencoder is as follows: The loss function corresponding to the second autoencoder is as follows: Where W represents historical health data, AE A (W) represents the data after W is reconstructed by the first autoencoder, AE B (W) represents the data after W is reconstructed by the second autoencoder, λ and γ are regularization term coefficients, ||AE A ||2, ||AE B ||2 are the regularization terms of the first encoder and the second encoder respectively.

6. The method according to claim 1, wherein Optimizing the trained first autoencoder and second autoencoder based on the adversarial learning strategy includes: Optimize the trained first autoencoder and second autoencoder according to the loss function of the adversarial learning strategy, where the loss function corresponding to the trained first autoencoder is The loss function corresponding to the trained second autoencoder is 7. The method according to any one of claims 1 to 6, characterized in that, Using the combined model of the optimized first autoencoder and second autoencoder to perform anomaly monitoring on the multi-channel lubricating oil abrasive particles includes: Obtain the real-time operation data of multi-channel lubricating oil abrasive particles; Using the combined model of the optimized first autoencoder and second autoencoder, the anomaly score of the real-time operation data is determined through Formula 1, and Formula 1 is where W′ represents the real-time operation data, AE A (W′) represents the data after W′ is reconstructed by the first autoencoder, AE B (AE A (W′)) represents the result after AE A (W′) is reconstructed by the second autoencoder, and α and β are coefficients.

8. An abnormal monitoring device for multi-channel lubricating oil abrasive particles, characterized in that, Including: A construction module for constructing the first autoencoder and the second autoencoder; An acquisition module for acquiring the historical health data of multi-channel lubricating oil abrasive particles; A training module for using the historical health data to train the first autoencoder and the second autoencoder; An optimization module for optimizing the trained first autoencoder and second autoencoder based on the adversarial learning strategy; An anomaly monitoring module for using the combined model of the optimized first autoencoder and second autoencoder to perform anomaly monitoring on the multi-channel lubricating oil abrasive particles.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein the computer program can be executed by an electronic device to perform the method described in any one of claims 1 to 7.

10. A computer program product comprising a computer program, characterized in that, When the computer program is executed by a processor, the steps of the method described in any one of claims 1 to 7 are implemented.

11. An electronic device, comprising a memory and a processor, characterized in that, A computer program is stored in the memory, and the processor is configured to execute the method described in any one of claims 1 to 7 through the computer program.

Images