CN115563464A - Nuclear valve abnormal state identification method based on quantum convolution neural network - Google Patents

Nuclear valve abnormal state identification method based on quantum convolution neural network Download PDF

Info

Publication number
CN115563464A
CN115563464A CN202211135581.0A CN202211135581A CN115563464A CN 115563464 A CN115563464 A CN 115563464A CN 202211135581 A CN202211135581 A CN 202211135581A CN 115563464 A CN115563464 A CN 115563464A
Authority
CN
China
Prior art keywords
quantum
neural network
layer
state
valve
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
CN202211135581.0A
Other languages
Chinese (zh)
Inventor
唐樟春
岳涧洲
湛力
夏艳君
刘志龙
潘阳红
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.)
University of Electronic Science and Technology of China
Original Assignee
University of Electronic Science and Technology of China
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 University of Electronic Science and Technology of China filed Critical University of Electronic Science and Technology of China
Priority to CN202211135581.0A priority Critical patent/CN115563464A/en
Publication of CN115563464A publication Critical patent/CN115563464A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F16—ENGINEERING ELEMENTS AND UNITS; GENERAL MEASURES FOR PRODUCING AND MAINTAINING EFFECTIVE FUNCTIONING OF MACHINES OR INSTALLATIONS; THERMAL INSULATION IN GENERAL
    • F16K—VALVES; TAPS; COCKS; ACTUATING-FLOATS; DEVICES FOR VENTING OR AERATING
    • F16K37/00—Special means in or on valves or other cut-off apparatus for indicating or recording operation thereof, or for enabling an alarm to be given
    • F16K37/0075—For recording or indicating the functioning of a valve in combination with test equipment
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N10/00—Quantum computing, i.e. information processing based on quantum-mechanical phenomena
    • G06N10/20—Models of quantum computing, e.g. quantum circuits or universal quantum computers
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N10/00—Quantum computing, i.e. information processing based on quantum-mechanical phenomena
    • G06N10/60—Quantum algorithms, e.g. based on quantum optimisation, quantum Fourier or Hadamard transforms
    • 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

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Computing Systems (AREA)
  • Artificial Intelligence (AREA)
  • Software Systems (AREA)
  • Mathematical Physics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Computational Mathematics (AREA)
  • Condensed Matter Physics & Semiconductors (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Analysis (AREA)
  • Biophysics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Molecular Biology (AREA)
  • Computational Linguistics (AREA)
  • General Health & Medical Sciences (AREA)
  • Mechanical Engineering (AREA)
  • Monitoring And Testing Of Nuclear Reactors (AREA)

Abstract

The invention discloses a quantum convolution neural network-based nuclear valve abnormal state identification method, which comprises the following steps: carrying out Fourier transform on the vibration data, extracting time-frequency domain characteristics and converting the time-frequency domain characteristics into quantum states; constructing a quantum convolutional neural network model with a quantum analog circuit layer, a convolutional neural network layer, a quantum bit observation layer, a full connection layer and a quantum network optimization updating layer; and (3) iteratively training the model for multiple times and optimizing quantum gate parameters in the model, so that the output result reaches the target output as far as possible, and the multi-classification task of machine learning is realized. The invention can extract the quantum characteristics from more aspects and can more accurately identify the abnormal state of the valve.

Description

Nuclear valve abnormal state identification method based on quantum convolution neural network
Technical Field
The invention relates to a method for realizing the abnormal state identification of a nuclear valve by using a quantum convolution neural network and simulating on a classical computer, and belongs to the technical field of machine learning and quantum computing.
Background
The nuclear power device has a complex structure and a plurality of systems, is a complex nonlinear system, and can generate adverse effects on the device, personnel and environment once major accidents occur, so that higher requirements are provided for the abnormal state identification technology. In order to ensure safe and reliable operation, it is necessary to develop a system capable of accurately identifying the abnormal state of the nuclear power plant.
Due to the fact that the nuclear power device is quite complex in structure and special in environment, meanwhile, the traditional abnormal state identification technology enters a bottleneck that knowledge acquisition is difficult and identification efficiency is low, and the requirement for identifying the abnormal state of the complex nuclear valve cannot be met. Therefore, aiming at the complex abnormal state of typical equipment of a nuclear power device, the comprehensive abnormal state identification technology is researched, a quantum convolution neural network model which can fully mine quantum computation potential and efficiently extract time-frequency domain data characteristics based on quantum computation is provided, and the technical problem to be solved is to be solved urgently in processing of machine learning tasks.
Disclosure of Invention
Aiming at the problems, the invention provides a method for processing vibration signal data by depending on the parallel characteristics of quantum behaviors, which can efficiently and accurately extract effective characteristic data in vibration signal big data and judge the abnormal state of a nuclear valve.
The technical scheme of the invention is a method for identifying the abnormal state of a nuclear valve based on a quantum convolution neural network, which comprises the following steps:
step 1: acquiring a nuclear valve vibration signal and corresponding normal and fault state types thereof, wherein the fault states comprise valve rod leakage, valve seat leakage, sealing surface damage caused by over-large actuator selection and over-high closing torque and leakage outside a sealing surface, and then establishing a data set which is divided into a training set and a testing set;
step 2: fourier transform is carried out on vibration signals in a data set, then characteristic data in time-frequency domain signals are extracted and converted into quantum states, and the method specifically comprises the following steps:
step 2.1: extracting peak value x of characteristic data of time domain signal max Peak to peak value x P-P Mean value x MV Effective value x RMS Square root amplitude x SRA Average amplitude x MA And standard deviation SD, the formulas are respectively shown as follows;
x Max ＝max(x(i))
x P-P ＝max(x(i))-min(x(i))
Figure BDA0003851379300000011
Figure BDA0003851379300000021
Figure BDA0003851379300000022
Figure BDA0003851379300000023
Figure BDA0003851379300000024
wherein x (i) is represented as a signal of a time domain signal subjected to sampling discretization, and N is the number of sampling points;
step 2.2: extracting mean square frequency MSF and mean frequency F of characteristic data of frequency domain signals 1 And the standard deviation frequency F 2 The formulas are respectively shown as follows;
Figure BDA0003851379300000025
Figure BDA0003851379300000026
Figure BDA0003851379300000027
wherein p is i For the frequency spectrum of the vibration signal x (i), i =1,2, …, M, M is the spectral line bus, f i The frequency value of the ith spectral line;
step 2.3: and (3) carrying out time-frequency domain characteristic data: peak value x max Peak to peak value x P-P Mean value x MV Effective value x RMS Square root amplitude x SRA Average amplitude x MA Standard deviation SD, frequency MSF, mean frequency F 1 And the standard deviation frequency F 2 Converting into quantum state with the following formula;
Figure BDA0003851379300000031
wherein x is i Representing the ith characteristic data;
step 2.4: converting the quantum state characteristic data into a quantum state by adopting an amplitude coding mode, designing a quantum gate sequence according to the target quantum state, and converting the quantum bit into the target quantum state;
and step 3: constructing a quantum convolution neural network model, inputting the target quantum state obtained in the step 2.4, and outputting the target quantum state as the state type of the nuclear valve: the method comprises the following steps of normality, valve rod leakage, valve seat leakage, sealing surface damage caused by too large actuator matching and too high closing torque, and leakage outside a sealing surface, wherein the sealing surface damage and the sealing surface leakage comprise a quantum simulation circuit layer, a convolutional neural network layer, a quantum bit observation layer, a full connection layer and a quantum network optimization updating layer; the quantum analog circuit layer is used for constructing an analog quantum circuit, applying a plurality of controlled quantum gate operations to the prepared quantum bit according to the structural characteristics of input data, changing the quantum state of the quantum bit, and carrying out quantum angle coding and quantum amplitude coding on the input data so as to realize characteristic extraction; the convolutional neural network layer comprises a plurality of convolutional layers and pooling layers and is used for extracting quantum state features; the quantum bit observation layer is used for applying measurement operation to the quantum bits processed by the quantum analog circuit layer to obtain an expected value; the full connection layer is used for mapping the expected value obtained by the quantum bit observation layer to neurons for multi-classification output, and the output result is used as an output result of a machine learning task; the quantum network optimization updating layer is used for analyzing the error of the output result of the quantum convolutional neural network model and the target output corresponding to the input data and updating the parameters of the controlled quantum gate in the quantum convolutional neural network model by utilizing an optimization algorithm according to the error analysis result;
and 4, step 4: carrying out iterative training by using the training set target quantum state obtained in the step 2 and the quantum convolution neural network obtained in the step 3 until a threshold condition set by a machine learning task is met, and terminating the training to obtain a trained quantum convolution neural network model;
and 5: verifying the trained quantum convolution neural network model by using the quantum state data of the test set obtained in the step 2; and after verification is completed, the method is used for identifying the abnormal state of the nuclear valve in real time.
Compared with the prior art, the invention has the following beneficial effects:
1. the invention adopts the quantum angle coding and the quantum amplitude coding to process the time-frequency domain characteristic signals, extracts the quantum characteristics from more aspects and can more accurately identify the abnormal state of the valve.
2. According to the method, quantum behaviors are introduced in the abnormal state identification of the valve, and the abnormal state of the nuclear valve can be identified more quickly by utilizing the parallelism of quantum computation.
3. After quantum observation, the invention uses the full-connection layer to carry out multi-classification, carries out error analysis and automatically updates adjustable parameters in the model.
Drawings
FIG. 1 is a flow chart of quantum angle encoding
FIG. 2 is a flow chart of operation of a quantum neural network module
Fig. 3 is a flow chart of the whole training method of the invention.
Detailed Description
The present invention will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the invention, but are not intended to limit the invention in any way. It should be noted that it would be obvious to those skilled in the art that various changes and modifications can be made without departing from the spirit of the invention. All falling within the scope of the invention.
The embodiment of the invention provides a quantum neural network method for identifying the abnormal state of a nuclear valve, which is shown in figure 3 and comprises the following steps:
step 1: acquiring a vibration signal and a state type corresponding to the vibration signal, establishing a data set, and dividing the data set into a training set and a test set;
step 2: fourier transform is carried out on vibration signals in a data set, then characteristic data in time-frequency domain signals are extracted and converted into quantum states, and the method specifically comprises the following steps:
step 2.1: extracting peak value x of characteristic data of time domain signal max Peak to peak value x P-P Mean value x MV Effective value x RMS Square root amplitude x SRA Average amplitude x MA And standard deviation SD, the formulas are shown below, respectively.
x Max ＝max(x(i))
x P-P ＝max(x(i))-min(x(i))
Figure BDA0003851379300000041
Figure BDA0003851379300000042
Figure BDA0003851379300000043
Figure BDA0003851379300000044
Figure BDA0003851379300000045
Wherein x (i) is represented by a signal obtained by discretization of a time domain signal through sampling, and N is the number of sampling points.
Step 2.2: extracting mean square frequency MSF and mean frequency F of characteristic data of frequency domain signals 1 And the standard deviation frequency F 2 The formulas are shown below.
Figure BDA0003851379300000051
Figure BDA0003851379300000052
Figure BDA0003851379300000053
Step 2.3: the time-frequency domain characteristic data are converted into quantum states, and a conversion formula is shown as follows.
Figure BDA0003851379300000054
Wherein x is i The ith feature data is represented.
And step 3: constructing a quantum convolutional neural network model, as shown in fig. 2, including a quantum analog circuit layer, a convolutional neural network layer, a qubit observation layer, a convolutional layer, and a quantum network optimization update layer; the following layers are specifically described:
(1) Quantum analog circuit layer
The analog circuit layer is used for constructing an analog quantum circuit, and applying a plurality of controlled quantum gate operations to the prepared quantum bit according to input data to change the quantum state of the quantum bit and realize the feature extraction of the input data.
The quantum analog circuit layer is composed of a plurality of quantum gates, the quantum gates are used for operating the connected quantum bits, the quantum states of the quantum bits are changed, and the extraction of quantum state features is realized, wherein the quantum gates comprise adjustable parameters theta, the operation of the corresponding quantum gates on the quantum bits can be changed by changing the adjustable parameters theta, and the quantum gates with the adjustable parameters comprise R x (theta) rotation quantum gate and R z (θ) the process of quantum gate rotation and quantum angle encoding is shown in fig. 1, and the formulas are respectively as follows:
Figure BDA0003851379300000061
Figure BDA0003851379300000062
where i is a complex number unit.
This results in each encoded initial qubit quantum state being
Figure BDA0003851379300000063
Then, quantum amplitude encoding is performed on the quantum bits, and the formula of the quantum amplitude encoding is as follows:
Figure BDA0003851379300000064
(2) Convolutional neural network layer
The convolutional neural network layer is used for training a quantum circuit with parameters, the extraction of quantum data features is realized, and as the data generated by the quantum analog circuit layer is one-dimensional, a plurality of one-dimensional convolutional layers and pooling layers are arranged to extract the feature data.
(3) Quantum bit observation layer
After the QB measurement layer acts on the QB simulation circuit layer and the convolutional neural network layer, an observation layer is constructed by using PQC function provided by Tensorflow Quantum, and Y-direction measurement operation is applied to one or more QBs to obtain expected value (which must be different from the original rotation angle)
(4) Full connection layer:
only one output layer is set, the number of the neurons is set to be 4, the activation function is set to be a softmax function, and the weight value in each neural network is set to be a random number which obeys even distribution between 0 and 1 and is used for classifying the observation results.
(5) The design process of the quantum network optimization updating layer is as follows:
the quantum network optimization layer is used for carrying out error analysis on the output result of the full connection layer and the corresponding target output result in the training set, updating adjustable parameters of quantum gates in the constructed quantum convolutional neural network model by using an optimization algorithm according to the error analysis result, and improving the performance of the quantum convolutional neural network model for processing machine learning tasks.
And 4, step 4: and (3) performing iterative training by using the quantum state data of the training set obtained in the step (2) and the quantum convolution neural network obtained in the step (3), and terminating the training until a threshold condition set by the machine learning task is met to obtain a trained quantum convolution neural network model.
And 5: and (3) verifying the trained quantum convolution neural network model by using the quantum state data of the test set obtained in the step (2).
The embodiment of the invention provides a quantum neural network method for identifying the abnormal state of a nuclear valve. By identifying the vibration data of the nuclear valve, the identification accuracy rate of the abnormal state of the nuclear valve can reach 93%, and the accuracy rate is higher compared with that of a traditional neural network.
The foregoing description of specific embodiments of the present invention has been presented. It is to be understood that the present invention is not limited to the specific embodiments described above, and that various changes or modifications may be made by one skilled in the art within the scope of the appended claims without departing from the spirit of the invention. The embodiments and features of the embodiments of the present application may be combined with each other arbitrarily without conflict.

Claims (5)

1. A nuclear valve abnormal state identification method of a quantum convolution neural network comprises the following steps:
step 1: acquiring a nuclear valve vibration signal and corresponding normal and fault state types thereof, wherein the fault states comprise valve rod leakage, valve seat leakage, sealing surface damage caused by over-large selection of an actuating mechanism and over-high closing torque and leakage outside a sealing surface, establishing a data set, and dividing the data set into a training set and a testing set;
and 2, step: carrying out Fourier transform on the vibration signals in the data set, then extracting characteristic data in the time-frequency domain signals and converting the characteristic data into quantum states;
and 3, step 3: constructing a quantum convolution neural network model, inputting the target quantum state obtained in the step 2.4, and then outputting the state type of the nuclear valve: the method comprises the following steps of normality, valve rod leakage, valve seat leakage, sealing surface damage caused by too large actuator matching and too high closing torque, and leakage outside a sealing surface, wherein the sealing surface damage and the sealing surface leakage comprise a quantum simulation circuit layer, a convolutional neural network layer, a quantum bit observation layer, a full connection layer and a quantum network optimization updating layer; the quantum analog circuit layer is used for constructing an analog quantum circuit, applying a plurality of controlled quantum gate operations to the prepared quantum bit according to the structural characteristics of input data, changing the quantum state of the quantum bit, and carrying out quantum angle coding and quantum amplitude coding on the input data so as to realize the characteristic extraction; the convolutional neural network layer comprises a plurality of convolutional layers and pooling layers and is used for extracting quantum state features; the quantum bit observation layer is used for applying measurement operation to the quantum bit processed by the quantum analog circuit layer to obtain an expected value; the full connection layer is used for mapping the expected value obtained by the quantum bit observation layer to neurons for multi-classification output and taking the multi-classification output as an output result of a machine learning task; the quantum network optimization updating layer is used for analyzing the error of the output result of the quantum convolutional neural network model and the target output corresponding to the input data and updating the parameters of the controlled quantum gate in the quantum convolutional neural network model by utilizing an optimization algorithm according to the error analysis result;
and 4, step 4: carrying out iterative training by using the quantum state data of the training set obtained in the step (2) and the quantum convolution neural network obtained in the step (3) until a threshold condition set by a machine learning task is met, and terminating the training to obtain a trained quantum convolution neural network model;
and 5: verifying the trained quantum convolution neural network model by using the test set quantum state data obtained in the step 2, and after verification, using the trained quantum convolution neural network model for identifying the abnormal state of the nuclear valve in real time;
the specific process of the step 2 is as follows:
step 2.1: extracting peak value x of characteristic data of time domain signal max Peak to peak value x P-P Mean value x MV Effective value x RMS Square root amplitude x SRA Average amplitude x MA And standard deviation SD;
step 2.2: extracting mean square frequency MSF and mean frequency F of characteristic data of frequency domain signals 1 And the standard deviation frequency F 2 ；
Step 2.3: and converting the time-frequency domain characteristic data into quantum states.
2. The method for identifying the abnormal state of the nuclear valve of the quantum convolution neural network according to claim 1, wherein the quantum simulation circuit layer is specifically:
the analog circuit layer is used for constructing an analog quantum circuit, and applying a plurality of controlled quantum gate operations to the prepared quantum bit according to the structural characteristics of the input data to change the quantum state of the quantum bit and realize the characteristic extraction of the input data;
the quantum analog circuit layer is composed of a plurality of quantum gates, the quantum gates are used for operating the connected quantum bits, the quantum states of the quantum bits are changed, and the extraction of quantum state features is realized, wherein the quantum gates comprise adjustable parameters theta, the operation of the corresponding quantum gates on the quantum bits can be changed by changing the adjustable parameters theta, and the quantum gates with the adjustable parameters comprise R x (theta) rotation quantum gate and R z (θ) the rotary quantum gates are respectively expressed as follows:
Figure FDA0003851379290000021
Figure FDA0003851379290000022
wherein i is a complex unit;
this results in each encoded initial qubit quantum state being R z (θ i z )R x (θ i x )|0>(ii) a Then, quantum amplitude encoding is performed on the quantum bits, and the formula of the quantum amplitude encoding is as follows:
Figure FDA0003851379290000023
3. the method as claimed in claim 1, wherein the qubit observation layer constructs an observation layer by using a PQC function provided by a tensoflow Quantum, and a Y-direction measurement operation is applied to one or more qubits to obtain an expected value.
4. The method for identifying the abnormal state of the nuclear valve of the quantum convolutional neural network as claimed in claim 1, wherein the fully connected layer has only one output layer for classifying the results obtained by the qubit observation layer.
5. The method for identifying the abnormal state of the nuclear valve of the quantum convolution neural network as claimed in claim 1, wherein the conversion into the quantum state in step 2.3 is as follows;
Figure FDA0003851379290000024
wherein x i The ith feature data is represented.
CN202211135581.0A 2022-09-19 2022-09-19 Nuclear valve abnormal state identification method based on quantum convolution neural network Pending CN115563464A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211135581.0A CN115563464A (en) 2022-09-19 2022-09-19 Nuclear valve abnormal state identification method based on quantum convolution neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211135581.0A CN115563464A (en) 2022-09-19 2022-09-19 Nuclear valve abnormal state identification method based on quantum convolution neural network

Publications (1)

Publication Number Publication Date
CN115563464A true CN115563464A (en) 2023-01-03

Family

ID=84740466

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211135581.0A Pending CN115563464A (en) 2022-09-19 2022-09-19 Nuclear valve abnormal state identification method based on quantum convolution neural network

Country Status (1)

Country Link
CN (1) CN115563464A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116738376A (en) * 2023-07-06 2023-09-12 广东筠诚建筑科技有限公司 Signal acquisition and recognition method and system based on vibration or magnetic field awakening
CN117649668A (en) * 2023-12-22 2024-03-05 南京天溯自动化控制***有限公司 Medical equipment metering certificate identification and analysis method

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116738376A (en) * 2023-07-06 2023-09-12 广东筠诚建筑科技有限公司 Signal acquisition and recognition method and system based on vibration or magnetic field awakening
CN116738376B (en) * 2023-07-06 2024-01-05 广东筠诚建筑科技有限公司 Signal acquisition and recognition method and system based on vibration or magnetic field awakening
CN117649668A (en) * 2023-12-22 2024-03-05 南京天溯自动化控制***有限公司 Medical equipment metering certificate identification and analysis method

Similar Documents

Publication Publication Date Title
CN107884706B (en) Analog circuit fault diagnosis method based on vector value regular kernel function approximation
Gou et al. Aeroengine control system sensor fault diagnosis based on CWT and CNN
CN115563464A (en) Nuclear valve abnormal state identification method based on quantum convolution neural network
Yang et al. Rotating machinery fault diagnosis using long-short-term memory recurrent neural network
CN112200244B (en) Intelligent detection method for anomaly of aerospace engine based on hierarchical countermeasure training
WO2017128455A1 (en) Analogue circuit fault diagnosis method based on generalized multiple kernel learning-support vector machine
CN108197648A (en) A kind of Fault Diagnosis Method of Hydro-generating Unit and system based on LSTM deep learning models
CN111259532B (en) Fault diagnosis method of aeroengine control system sensor based on 3DCNN-JTFA
CN101614787B (en) Analogical electronic circuit fault diagnostic method based on M-ary-structure classifier
CN107909118A (en) A kind of power distribution network operating mode recording sorting technique based on deep neural network
CN110705456A (en) Micro motor abnormity detection method based on transfer learning
Huang et al. Machine fault detection for intelligent self-driving networks
CN110067696B (en) Wind turbine generator load determining method, device, equipment and readable storage medium
Gai et al. A Parameter‐Optimized DBN Using GOA and Its Application in Fault Diagnosis of Gearbox
CN105606914A (en) IWO-ELM-based Aviation power converter fault diagnosis method
CN112507479B (en) Oil drilling machine health state assessment method based on manifold learning and softmax
Huang et al. Deep residual networks-based intelligent fault diagnosis method of planetary gearboxes in cloud environments
CN108805195A (en) A kind of motor group method for diagnosing faults based on two-value deep-neural-network
CN112257862A (en) Semi-supervised identification method based on relational network marker sample expansion
Li et al. Intelligent fault diagnosis of aeroengine sensors using improved pattern gradient spectrum entropy
CN115329986A (en) Wind turbine generator anomaly detection and positioning method based on interpretable graph neural network
CN111122811A (en) Sewage treatment process fault monitoring method of OICA and RNN fusion model
CN115163424A (en) Wind turbine generator gearbox oil temperature fault detection method and system based on neural network
CN111160393A (en) Data-driven modularized modeling method for carrier rocket health assessment model
Wang et al. Intermittent fault diagnosis of analog circuit based on enhanced one-dimensional vision transformer and transfer learning strategy

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