CN114399064A - Equipment health index construction method based on multi-source sensor data fusion - Google Patents

Equipment health index construction method based on multi-source sensor data fusion Download PDF

Info

Publication number
CN114399064A
CN114399064A CN202111665963.XA CN202111665963A CN114399064A CN 114399064 A CN114399064 A CN 114399064A CN 202111665963 A CN202111665963 A CN 202111665963A CN 114399064 A CN114399064 A CN 114399064A
Authority
CN
China
Prior art keywords
equipment
self
matrix
sensor data
source sensor
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202111665963.XA
Other languages
Chinese (zh)
Inventor
姜伟
薛小明
张楠
陈中
王馨梓
孙娜
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huaiyin Institute of Technology
Original Assignee
Huaiyin Institute of Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Huaiyin Institute of Technology filed Critical Huaiyin Institute of Technology
Priority to CN202111665963.XA priority Critical patent/CN114399064A/en
Publication of CN114399064A publication Critical patent/CN114399064A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/20—Administration of product repair or maintenance
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G06F18/2155—Generating training patterns; Bootstrap methods, e.g. bagging or boosting characterised by the incorporation of unlabelled data, e.g. multiple instance learning [MIL], semi-supervised techniques using expectation-maximisation [EM] or naïve labelling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/25—Fusion techniques
    • G06F18/253—Fusion techniques of extracted features
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/088—Non-supervised learning, e.g. competitive learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/80—Management or planning

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Physics & Mathematics (AREA)
  • Human Resources & Organizations (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Strategic Management (AREA)
  • Molecular Biology (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computational Linguistics (AREA)
  • General Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Evolutionary Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Biophysics (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Development Economics (AREA)
  • Educational Administration (AREA)
  • Game Theory and Decision Science (AREA)
  • Testing Or Calibration Of Command Recording Devices (AREA)

Abstract

The invention relates to the field of health prediction and management of mechanical equipment, and discloses an equipment health index construction method based on multi-source sensor data fusion, which comprises the following steps: 1) reconstructing a multi-source sensor data sequence to obtain a multi-scale reconstruction matrix; 2) based on a stack type self-encoder model, performing self-adaptive feature extraction on the multi-scale reconstruction matrix to generate a health state feature set; 3) constructing a self-organizing mapping neural network and finishing training by taking the health state feature set as input; 4) and solving the minimum quantization error of the self-organizing mapping neural network, and taking the minimum quantization error as a mechanical equipment health index. The method provided by the invention can meet the requirements of evaluating, predicting and managing the comprehensive health state of mechanical equipment under the data-driven background, and based on the feature level data fusion thought, the health index which effectively represents the health evolution process of the equipment is obtained, so that necessary basis is provided for formulating a reasonable and feasible equipment maintenance plan.

Description

Equipment health index construction method based on multi-source sensor data fusion
Technical Field
The invention relates to the field of mechanical equipment health prediction and management, in particular to an equipment health index construction method based on multi-source sensor data fusion.
Background
The safe, stable and reliable running state of the mechanical equipment is ensured, and the method has extremely important significance on the modern industrial capacity and the economic benefits of enterprises. In recent years, along with the remarkable improvement of the structural complexity and the integration level of equipment, the sources of the monitoring signals of the sensors which can be acquired are wider, the types are increasingly rich, and a state analysis method based on multi-source sensor signal data becomes a main means for evaluating the state of the equipment and making maintenance decisions. In order to comprehensively evaluate the running state of equipment and improve the quantitative representation level of the health state of the equipment, an equipment health state index construction method based on multi-source sensor data becomes an important work to be researched urgently.
The existing equipment health state index construction mainly adopts a symbolic signal screening mode, namely, a signal capable of representing the equipment health state evolution process is selected from a large number of collectable sensor signals and is used as a health index. The method depends on rich expert experience and priori knowledge, and the difficulty of signal screening is obviously improved along with the increase of the types of the acquired signals. Therefore, the index construction method based on multi-source signal fusion can provide a new idea for solving the problems. Aiming at the construction of the health state index of the equipment, at present, no system-in-depth research is carried out, a health index construction theoretical system with self-adaptability and strong robustness is lacked, and the functional requirements on the evaluation, prediction and management of the comprehensive health state of the equipment cannot be met.
Disclosure of Invention
The purpose of the invention is as follows: aiming at the problems in the prior art, the invention provides an equipment health index construction method based on multi-source sensor data fusion, which establishes an index construction theoretical system taking a feature level data fusion thought as a core and solves the technical problems that the prior art is difficult to screen feature signals, single in state characterization scale, insufficient in adaptability and incapable of performing health prediction and management on equipment.
The technical scheme is as follows: the invention provides a method for constructing equipment health indexes based on multi-source sensor data fusion, which comprises the following steps:
step 1: reconstructing an acquired multi-source sensor data sequence on the equipment to obtain a multi-scale reconstruction matrix;
step 2: based on a stack type self-encoder model, performing self-adaptive feature extraction on the multi-scale reconstruction matrix obtained in the step 1 to generate a health state feature set;
and step 3: taking the health state feature set obtained in the step 2 as input, constructing a self-organizing mapping neural network and finishing training;
and 4, step 4: and solving the minimum quantization error of the self-organizing mapping neural network, and taking the minimum quantization error as a mechanical equipment health index.
Further, the step 1 comprises the following sub-steps:
(1-1) collecting multi-source sensor data sequence set { X) in the running process of target equipment1,X2,…XnIn which X isi＝xi1,xi2,…,xim；
(1-2) carrying out spatial reconstruction on the multi-source sensor data sequence, wherein the reconstruction method comprises the following steps: initializing a two-dimensional matrix, wherein the dimension of the matrix is nxm, taking the number of data sequences of sensors from different sources as the row of the matrix and the length of the data sequence of the sensor from a single source as the column of the matrix, assigning the two-dimensional matrix, and further obtaining a multi-scale reconstruction data matrix [ X [ [ X ]1；X2；…；Xn]n×m。
Further, the generating of the health status feature set in step 2 includes the following steps:
(2-1) reconstruction of data matrix [ X ] in multiple scales1；X2；…；Xn]n×mAs input data, training the stacked self-encoder to obtain a first hidden layer feature vector gamma1；
(2-2) using the first hidden layer feature vector Γ1As input data, continuing to train the stacked self-encoder to obtain a second hidden layer feature vector gamma2；
And (2-3) training the next single self-encoder by taking the above hidden layer feature vector as input until the training of the stacked self-encoder is completed, and acquiring a top hidden layer feature vector set gamma which is taken as the extracted equipment health state feature.
Further, the stacked self-encoder model includes 5 layers, which are an input layer, a first hidden layer, a second hidden layer, a third hidden layer and an output layer.
Further, the step 3 comprises the following sub-steps:
(3-1) constructing a self-organizing map neural network, wherein the network comprises a two-layer structure, namely an input layer and an output layer; the input layer nodes and the output layer topological structure are connected through a network weight matrix, and an optimal weight matrix is obtained through network training;
(3-2) initializing a network weight matrix;
(3-3) sequentially calculating the Euclidean distance E between the ith neuron and the equipment health state feature in the topological layeriTaking the neuron with the minimum Euclidean distance as a winning unit;
(3-4) correcting the connection weight between the winning unit and the adjacent neurons through training until the maximum training times N is reached to obtain the weight vector omega of the final winning unitw。
Further, the step 4 comprises the following sub-steps:
(4-1) solving a minimum quantized error value MQE between the health status feature of the equipment and the winning cell, wherein the calculation formula is as follows:
MQE＝||Γ-ωw||
wherein the content of the first and second substances,ωwa weight vector representing a winning unit;
(4-2) taking the minimum quantization error obtained above as a finally constructed equipment health index h, namely:
h＝MQE。
has the advantages that:
(1) the invention introduces the concept of feature level data fusion into the field of mechanical equipment health prediction and management for the first time, comprehensively considers the representation mechanism of multisource sensor signals on the evolution process of the equipment health state, overcomes the limitation that the screening of the feature signals excessively depends on expert experience and priori knowledge, avoids the problem of single representation scale of the equipment health state, fills the blank that the comprehensive representation of the equipment health state has no scientific theoretical basis and technical guidance, effectively meets the requirements of evaluating, predicting and managing the comprehensive health state of mechanical equipment under the data driving background, and provides necessary basis for formulating a reasonable and feasible equipment maintenance plan.
(2) According to the invention, the effective characteristic information deep extraction of a multi-source sensor data set is firstly considered to be completed by adopting a stacked self-encoder model, and effective fusion of the extracted characteristic information is realized based on the advantages of unsupervised self-organizing mapping neural network, strong adaptability and the like, so that a quantization index capable of comprehensively measuring the health state level of equipment is obtained, the characterization capability of the health state of the equipment is remarkably improved, the theoretical problem that the health state of the equipment is accurately depicted and difficultly realized is overcome, the evaluation and predictable level of the health state evolution of the equipment is effectively improved, and reliable theoretical guidance and technical support are provided for implementing intelligent state maintenance of the equipment.
Drawings
FIG. 1 is a flow chart of an apparatus health index construction method based on multi-source sensor data fusion according to the present invention;
FIG. 2 is a data sequence variation curve of a multi-source sensor of an aircraft engine apparatus in an embodiment of the invention;
FIG. 3 is a model schematic diagram of the extraction of the health condition feature set of the aircraft engine device in an embodiment of the invention;
FIG. 4 is a model schematic diagram of the fusion of state of health features of aircraft engine equipment in an embodiment of the invention;
FIG. 5 is a series of aeroengine equipment health indicators constructed in accordance with an embodiment of the present invention.
Detailed Description
The present invention will be described in detail with reference to the accompanying drawings.
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
The invention takes the construction of the health status index of certain aeroengine equipment as an analysis case, and fig. 1 shows a flow chart of the construction method of the health index of the equipment based on the multi-source sensor data fusion, which specifically comprises the following steps:
step 1: and reconstructing the data sequence of the multi-source sensor to obtain a multi-scale reconstruction matrix.
(1-1) acquiring a multi-source sensor data sequence set { X ] in the running process of target equipment by using a state monitoring system1,X2,…XnIn which X isi＝xi1,xi2,…,xim. The health state index construction of certain aircraft engine equipment is used as an analysis case, the included multi-source sensor signals are shown in table 1, and fig. 2 shows a data sequence change curve of the multi-source sensor of the aircraft engine equipment. The graphical results show that the single sensor signal is difficult to accurately represent the evolution process of the health state of the equipment, and new exploration and attempt are needed for the construction method of the health index of the equipment.
TABLE 1 Multi-Source sensor Signal description
Figure BDA0003448293010000041
(1-2) reconstructing a data sequence of the multi-source sensor, initializing a two-dimensional matrix with dimension of nxm, and sensing by different sourcesThe number of the data sequences of the device is used as the row of the matrix, the length of the data sequences of the single-source sensor is used as the column of the matrix, the two-dimensional matrix is assigned, and then a multi-scale reconstruction data matrix [ X ] is obtained1；X2；…；Xn]n×m. In this case, the following data matrix can be obtained after reconstructing the multi-source sensor signal: [ T ]24；T30；T50；P30；Ps30；Phi；BPR]The matrix dimension is 7 × 210.
Step 2: and (3) based on a stack type self-encoder model, performing self-adaptive feature extraction on the multi-scale reconstruction matrix obtained in the step (1) to generate a health state feature set.
(2-1) reconstruction of data matrix [ T ] in multiple scales24；T30；T50；P30；Ps30；Phi；BPR]As input data, training the stacked self-encoder to obtain a first hidden layer feature vector gamma1. As shown in fig. 3, the schematic diagram of a stacked self-encoder model extracted from the feature set of the health state of the aircraft engine equipment includes 5 layers, and the node structures of each layer are as follows: 7 (input layer) -50 (first hidden layer) -50 (second hidden layer) -50 (third hidden layer) -7 (output layer), the model detailed parameter settings are as follows: the learning rate is 0.1, the momentum is 0.5, and the maximum number of iterations is 300.
(2-2) using the first hidden layer feature vector Γ1As input data, continuing to train the stacked self-encoder to obtain a second hidden layer feature vector gamma2。
(2-3) using the second hidden layer feature vector Γ2And taking a third hidden layer feature vector set gamma as an input, namely the extracted equipment health state feature.
And step 3: and (3) taking the health state feature set obtained in the step (2) as input, constructing a self-organizing mapping neural network and finishing training.
And (3-1) initializing a self-organizing mapping neural network connection weight matrix. As shown in fig. 4, a schematic diagram of a self-organizing map neural network model with fusion of health status features of aircraft engine equipment is shown, and network parameters are set as follows: the number of nodes of the input layer is 50, the dimensionality of the output layer is 5 multiplied by 5, the learning rate is 0.85, and the maximum iteration number is 200.
(3-2) sequentially calculating the Euclidean distance E between the ith neuron and the equipment health state feature in the topological layeriAnd taking the neuron with the minimum Euclidean distance as a winning unit.
(3-3) correcting the connection weight between the winning unit and the adjacent neurons by training until the maximum training time N is 200, and obtaining the weight vector omega of the final winning unitw。
And 4, step 4: and solving the minimum quantization error of the self-organizing mapping neural network, and taking the minimum quantization error as a mechanical equipment health index.
(4-1) solving the minimum quantization error MQE between the health status characteristics of the equipment and the winning unit, wherein the calculation formula is as follows:
MQE＝||Γ-ωw||
(4-2) taking the minimum quantization error obtained above as a finally constructed equipment health index h, namely:
h＝MQE
fig. 5 shows a health index sequence curve of an aircraft engine device constructed in an embodiment of the present invention, that is, a linear representation of a health index h on a one-dimensional scale. According to the graphic results, the finally constructed equipment health index sequence curve can well represent the health state evolution process of the aircraft engine equipment, and the effectiveness of the equipment health index construction method based on the multi-source sensor data fusion is fully verified.
The above embodiments are merely illustrative of the technical concepts and features of the present invention, and the purpose of the embodiments is to enable those skilled in the art to understand the contents of the present invention and implement the present invention, and not to limit the protection scope of the present invention. All equivalent changes and modifications made according to the spirit of the present invention should be covered within the protection scope of the present invention.

Claims (6)

1. The equipment health index construction method based on multi-source sensor data fusion is characterized by comprising the following steps:
step 1: reconstructing an acquired multi-source sensor data sequence on the equipment to obtain a multi-scale reconstruction matrix;
step 2: based on a stack type self-encoder model, performing self-adaptive feature extraction on the multi-scale reconstruction matrix obtained in the step 1 to generate a health state feature set;
and step 3: taking the health state feature set obtained in the step 2 as input, constructing a self-organizing mapping neural network and finishing training;
and 4, step 4: and solving the minimum quantization error of the self-organizing mapping neural network, and taking the minimum quantization error as a mechanical equipment health index.
2. The device health index construction method based on multi-source sensor data fusion according to claim 1, wherein the step 1 comprises the following substeps:
(1-1) collecting multi-source sensor data sequence set { X) in the running process of target equipment1,X2,…XnIn which X isi＝xi1,xi2,…,xim；
(1-2) carrying out spatial reconstruction on the multi-source sensor data sequence, wherein the reconstruction method comprises the following steps: initializing a two-dimensional matrix, wherein the dimension of the matrix is nxm, taking the number of data sequences of sensors from different sources as the row of the matrix and the length of the data sequence of the sensor from a single source as the column of the matrix, assigning the two-dimensional matrix, and further obtaining a multi-scale reconstruction data matrix [ X [ [ X ]1；X2；…；Xn]n×m。
3. The method for constructing the health index of the equipment based on the multi-source sensor data fusion according to claim 1, wherein the step 2 of generating the health state feature set comprises the following steps:
(2-1) reconstruction of data matrix [ X ] in multiple scales1；X2；…；Xn]n×mAs input data, training the stacked self-encoder to obtain a first hidden layer feature vector gamma1；
(2-2) using the first hidden layer feature vector Γ1As an input numberAnd continuing to train the stacked self-encoder to obtain a second hidden layer feature vector gamma2；
And (2-3) training the next single self-encoder by taking the above hidden layer feature vector as input until the training of the stacked self-encoder is completed, and acquiring a top hidden layer feature vector set gamma which is taken as the extracted equipment health state feature.
4. The multi-source sensor data fusion-based equipment health index construction method according to claim 3, wherein the stacked self-encoder model comprises 5 layers, namely an input layer, a first hidden layer, a second hidden layer, a third hidden layer and an output layer.
5. The device health index construction method based on multi-source sensor data fusion of claim 1, wherein the step 3 comprises the following substeps:
(3-1) constructing a self-organizing map neural network, wherein the network comprises a two-layer structure, namely an input layer and an output layer; the input layer nodes and the output layer topological structure are connected through a network weight matrix, and an optimal weight matrix is obtained through network training;
(3-2) initializing a network weight matrix;
(3-3) sequentially calculating the Euclidean distance E between the ith neuron and the equipment health state feature in the topological layeriTaking the neuron with the minimum Euclidean distance as a winning unit;
(3-4) correcting the connection weight between the winning unit and the adjacent neurons through training until the maximum training times N is reached to obtain the weight vector omega of the final winning unitw。
6. The device health index construction method based on multi-source sensor data fusion of claim 1, wherein the step 4 comprises the following substeps:
(4-1) solving a minimum quantized error value MQE between the health status feature of the equipment and the winning cell, wherein the calculation formula is as follows:
MQE＝||Γ-ωw||
wherein, ω iswA weight vector representing a winning unit;
(4-2) taking the minimum quantization error obtained above as a finally constructed equipment health index h, namely:
h＝MQE。
CN202111665963.XA 2021-12-31 2021-12-31 Equipment health index construction method based on multi-source sensor data fusion Pending CN114399064A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111665963.XA CN114399064A (en) 2021-12-31 2021-12-31 Equipment health index construction method based on multi-source sensor data fusion

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111665963.XA CN114399064A (en) 2021-12-31 2021-12-31 Equipment health index construction method based on multi-source sensor data fusion

Publications (1)

Publication Number Publication Date
CN114399064A true CN114399064A (en) 2022-04-26

Family

ID=81228713

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111665963.XA Pending CN114399064A (en) 2021-12-31 2021-12-31 Equipment health index construction method based on multi-source sensor data fusion

Country Status (1)

Country Link
CN (1) CN114399064A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115328228A (en) * 2022-10-13 2022-11-11 新乡市合力鑫电源有限公司 High-frequency switching power supply
CN116911354A (en) * 2023-09-14 2023-10-20 首都信息发展股份有限公司 Encoder neural network model construction method and data processing method

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115328228A (en) * 2022-10-13 2022-11-11 新乡市合力鑫电源有限公司 High-frequency switching power supply
CN115328228B (en) * 2022-10-13 2023-04-07 新乡市合力鑫电源有限公司 High-frequency switching power supply
CN116911354A (en) * 2023-09-14 2023-10-20 首都信息发展股份有限公司 Encoder neural network model construction method and data processing method

Similar Documents

Publication Publication Date Title
CN112131760B (en) CBAM model-based prediction method for residual life of aircraft engine
CN106156003B (en) A kind of question sentence understanding method in question answering system
CN106779151B (en) A kind of line of high-speed railway wind speed multi-point multi-layer coupling prediction method
CN110321603A (en) A kind of depth calculation model for Fault Diagnosis of Aircraft Engine Gas Path
CN114399064A (en) Equipment health index construction method based on multi-source sensor data fusion
CN107066759A (en) A kind of Vibration Fault Diagnosis of Turbine Rotor method and device
CN111079977A (en) Heterogeneous federated learning mine electromagnetic radiation trend tracking method based on SVD algorithm
CN106408120B (en) Local area landslide prediction device and method
CN109947898B (en) Equipment fault testing method based on intellectualization
CN113743016B (en) Engine residual life prediction method based on self-encoder and echo state network
CN110177122A (en) A kind of method for establishing model and device identifying network security risk
CN114925536A (en) Airborne system PHM testability modeling and diagnosis strategy optimization method and device
Wang et al. Artificial intelligent fault diagnosis system of complex electronic equipment
CN113722997A (en) New well dynamic yield prediction method based on static oil and gas field data
CN114357372A (en) Aircraft fault diagnosis model generation method based on multi-sensor data driving
CN113486926B (en) Automatic change pier equipment anomaly detection system
CN107038450A (en) Unmanned plane policing system based on deep learning
CN114295967A (en) Analog circuit fault diagnosis method based on migration neural network
CN116432359A (en) Variable topology network tide calculation method based on meta transfer learning
CN113722308B (en) Acceleration response data complement method and device based on EEMD-MultiCNN-LSTM
CN115359197A (en) Geological curved surface reconstruction method based on spatial autocorrelation neural network
CN115099129A (en) Natural gas well yield prediction method based on input characteristic error correction
Li et al. CapsNet-Enabled Multimodal Data Fusion: A Case Study in RUL Estimation of Turbofan Engine
CN111104298A (en) LSTM-based power grid server running state prediction device
KR102590793B1 (en) Method and apparatus of self-balancing online dataset for machine learning

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination