CN111915101A - Complex equipment fault prediction method and system based on LPP-HMM method - Google Patents
Complex equipment fault prediction method and system based on LPP-HMM method Download PDFInfo
- Publication number
- CN111915101A CN111915101A CN202010827424.0A CN202010827424A CN111915101A CN 111915101 A CN111915101 A CN 111915101A CN 202010827424 A CN202010827424 A CN 202010827424A CN 111915101 A CN111915101 A CN 111915101A
- Authority
- CN
- China
- Prior art keywords
- hmm
- state
- lpp
- fault
- dimensional
- 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
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
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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
-
- 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
Abstract
The invention discloses a fault prediction method and a fault prediction system for complex equipment based on an LPP-HMM method, belongs to the field of fault prediction of complex equipment, introduces nonlinear manifold learning into fault diagnosis of the complex equipment, and provides a fault diagnosis method using the LPP and the HMM. Simulation experiments show that the method can better retain the overall structure information in the fault data and is beneficial to fault identification. The LPP algorithm can be used for reducing nonlinear fault characteristics of an analog circuit, and the effect is superior to that of linear dimension reduction methods such as PCA (principal component analysis).
Description
Technical Field
The invention relates to the field of complex equipment fault prediction, in particular to a complex equipment fault prediction method and system based on an LPP-HMM method.
Background
The fault diagnosis technology is the core content of a PHM technology of a complex equipment system, and due to the characteristics of nonlinearity, time-varying property and the like of the complex equipment system, a fault signal of the complex equipment system has the characteristics of high dimension, nonlinearity, instability and the like, and has external noise or interference, so that the conventional signal processing method is difficult to effectively acquire the fault characteristic. In the diagnosis process, the traditional fault diagnosis method only divides the system state into a normal state and a fault state, and the system state is difficult to be reflected comprehensively. Intermittent faults and random faults cannot be identified, the intermediate state of the system cannot be monitored, and gradual fault monitoring is more difficult.
Disclosure of Invention
The invention provides a complex equipment fault prediction method and a complex equipment fault prediction system based on an LPP-HMM method to solve the problems.
In order to achieve the purpose, the invention adopts the following technical scheme:
the complex equipment fault prediction method and system based on the LPP-HMM method comprise the following steps;
s1, data processing and feature extraction: collecting sample data, and preprocessing the sample data by adopting an LPP algorithm;
s2, determining the number L of hidden states of an HMM model according to the actual condition of equipment, setting initial parameters of the model, establishing an initial HMM model lambda (pi, A and B), carrying out maximum likelihood estimation, and selecting k feature vectors with any side and any intention to form a group of state sequences to train the HMM model of each state type until a convergence condition is met, wherein the state types are hidden states, and each state type needs to train one HMM model;
s3, inputting the original characteristic information in the preprocessed data in the S1 into a fuzzy neural network, and performing network calculation to select n HMM models with the strongest relevance;
s4, selecting N original signals in S1 according to the fuzzy neural network state prediction information, subjecting the N original signals to the same feature dimension reduction method to obtain N low-dimensional feature vectors, constructing the N feature vectors into a group of feature sequences, inputting the N HMM models selected in the trained S3, and calculating likelihood probability values p (o | lambda), p (o | lambda)1)K p(o|λn) And the HMM model corresponding to the maximum probability value is the state of the current system.
Preferably, the LPP algorithm in S1 preprocess the sample data includes the following steps:
a1, constructing a sample point x according to the sample dataiThere are two methods for selecting the neighborhood of the sample point: a spherical approximation method and a k nearest neighbor method;
a2, calculating a weight matrix W, and constructing the weight matrix by a thermonuclear method or a reduction method;
a3 by formula XLXTK＝λXDXTAnd K, acquiring a mapping matrix, reducing the dimension of the original high-dimensional data space according to the mapping matrix, acquiring the inherent low-dimensional manifold characteristics, and keeping the original characteristic information.
Preferably, the specific steps of training the HMM model in S2 are as follows:
b1, setting L system state modes aiming at the characteristics of the monitored object, such as: carrying out m times of signal acquisition on the measuring point signal of each state mode in a normal state, an intermediate state, an intermittent state, a fault state and the like to obtain L multiplied by m original characteristic signals;
b2, sampling each original characteristic signal to construct a pi multiplied by n-dimensional high-dimensional original signal characteristic space, and forming Lxm m multiplied by n-dimensional high-dimensional spaces;
b3, learning the training sample by adopting an LPP algorithm, selecting an optimal parameter neighborhood factor K and an optimal embedding dimension d, and converting a complex high-dimensional data space into a low mxd-dimensional feature space, wherein d is more than or equal to 1 and less than or equal to n;
and B4, for each system state, randomly selecting C feature vectors to form a group of state sequences to train the HMM models of the system state, wherein C is less than or equal to m, and training L HMM models.
Preferably, the network calculation performed by the fuzzy neural network in S3 includes the following specific steps:
c1, determining a fuzzy relation matrix, and determining the membership degree of a single symptom to a single reason by adopting a weighted statistical method;
rij＝max(γ(k)+μij (k)-γ(k)μij (k))k＝0,1,2
in the formula: in the formula, gamma(k)The weight values of three terms of empirical statistics (k is 0), mechanism analysis (k is 1) and field analysis (k is 2), and gamma is(k)>0 and
the correction can be made according to different stages of system operation and the mastery of various factors, and then, muij (k)Is the value of the credit of each factor;
c2, selecting a fuzzy comprehensive evaluation model;
and C3, outputting a result vector as a group of membership values according to the fuzzy comprehensive evaluation model, and judging which category the fault belongs to and the fault reason thereof according to the group of membership values.
Compared with the prior art, the invention provides a complex equipment fault prediction method and a complex equipment fault prediction system based on an LPP-HMM method, and the method and the system have the following beneficial effects:
1. the invention introduces nonlinear manifold learning into the fault diagnosis of complex equipment, and provides a fault diagnosis method by using LPP and HMM. Simulation experiments show that the method can better retain the overall structure information in the fault data and is beneficial to fault identification. The LPP algorithm can be used for nonlinear fault feature reduction of an analog circuit, the effect is superior to linear dimension reduction methods such as PCA, a mixed HMM classifier reflects the real state of each stage of the system, the better recognition effect on the state of the analog circuit system is demonstrated through example verification, the early fault and the intermediate state of the system can be effectively recognized, the LPP method is combined with the HMM model, the respective advantages of the two methods can be fully utilized, the improvement of fault recognition rate is facilitated, the better recognition effect on the soft fault of the analog circuit is achieved, and the method is different from the conventional method combining the LPP method with the HMM model, a fuzzy neural network is introduced for screening, and the operation process of the HMM model is effectively reduced.
Drawings
Fig. 1 is a system flow diagram of a complex equipment fault prediction method and system based on the LPP-HMM method according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments.
In the description of the present invention, it is to be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", and the like, indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, are merely for convenience in describing the present invention and simplifying the description, and do not indicate or imply that the device or element being referred to must have a particular orientation, be constructed and operated in a particular orientation, and thus, should not be construed as limiting the present invention.
Example 1:
the complex equipment fault prediction method and system based on the LPP-HMM method comprise the following steps;
s1, data processing and feature extraction: collecting sample data, and preprocessing the sample data by adopting an LPP algorithm;
s2, determining the number L of hidden states of an HMM model according to the actual condition of equipment, setting initial parameters of the model, establishing an initial HMM model lambda (pi, A and B), carrying out maximum likelihood estimation, and selecting k feature vectors with any side and any intention to form a group of state sequences to train the HMM model of each state type until a convergence condition is met, wherein the state types are hidden states, and each state type needs to train one HMM model;
s3, inputting the original characteristic information in the preprocessed data in the S1 into a fuzzy neural network, and performing network calculation to select n HMM models with the strongest relevance;
s4, selecting N original signals in S1 according to the fuzzy neural network state prediction information, subjecting the N original signals to the same feature dimension reduction method to obtain N low-dimensional feature vectors, constructing the N feature vectors into a group of feature sequences, inputting the N HMM models selected in the trained S3, and calculating likelihood probability values p (o | lambda), p (o | lambda)1)K p(o|λn) And the HMM model corresponding to the maximum probability value is the state of the current system.
Further, preferably, the LPP algorithm in S1 pre-processing the sample data includes the following steps:
a1, constructing a sample point x according to the sample dataiThere are two methods for selecting the neighborhood of the sample point: a spherical approximation method and a k nearest neighbor method;
a2, calculating a weight matrix W, and constructing the weight matrix by a thermonuclear method or a reduction method;
a3 by formula XLXTK＝λXDXTAnd K, acquiring a mapping matrix, reducing the dimension of the original high-dimensional data space according to the mapping matrix, acquiring the inherent low-dimensional manifold characteristics, and keeping the original characteristic information.
Further, preferably, the specific steps of training the HMM model in S2 are as follows:
b1, setting L system state modes aiming at the characteristics of the monitored object, such as: carrying out m times of signal acquisition on the measuring point signal of each state mode in a normal state, an intermediate state, an intermittent state, a fault state and the like to obtain L multiplied by m original characteristic signals;
b2, sampling each original characteristic signal to construct a pi multiplied by n-dimensional high-dimensional original signal characteristic space, and forming Lxm m multiplied by n-dimensional high-dimensional spaces;
b3, learning the training sample by adopting an LPP algorithm, selecting an optimal parameter neighborhood factor K and an optimal embedding dimension d, and converting a complex high-dimensional data space into a low mxd-dimensional feature space, wherein d is more than or equal to 1 and less than or equal to n;
and B4, for each system state, randomly selecting C feature vectors to form a group of state sequences to train the HMM models of the system state, wherein C is less than or equal to m, and training L HMM models.
Further, preferably, the network calculation performed by the fuzzy neural network in S3 includes the following specific steps:
c1, determining a fuzzy relation matrix, and determining the membership degree of a single symptom to a single reason by adopting a weighted statistical method;
rij＝max(γ(k)+μij (k)-γ(k)μij (k))k＝0,1,2
in the formula: in the formula, gamma(k)The weight values of three terms of empirical statistics (k is 0), mechanism analysis (k is 1) and field analysis (k is 2), and gamma is(k)>0 and
the correction can be made according to different stages of system operation and the mastery of various factors, and then, muij (k)Is the value of the credit of each factor;
c2, selecting a fuzzy comprehensive evaluation model;
and C3, outputting a result vector as a group of membership values according to the fuzzy comprehensive evaluation model, and judging which category the fault belongs to and the fault reason thereof according to the group of membership values.
Example 2:
the invention introduces nonlinear manifold learning into the fault diagnosis of complex equipment, and provides a fault diagnosis method by using LPP and HMM. Simulation experiments show that the method can better retain the overall structure information in the fault data and is beneficial to fault identification. The LPP algorithm can be used for nonlinear fault feature reduction of an analog circuit, the effect is superior to linear dimension reduction methods such as PCA, a mixed HMM classifier reflects the real state of each stage of the system, the better recognition effect on the state of the analog circuit system is demonstrated through example verification, the early fault and the intermediate state of the system can be effectively recognized, the LPP method is combined with the HMM model, the respective advantages of the two methods can be fully utilized, the improvement of fault recognition rate is facilitated, the better recognition effect on the soft fault of the analog circuit is achieved, and the method is different from the conventional method combining the LPP method with the HMM model, a fuzzy neural network is introduced for screening, and the operation process of the HMM model is effectively reduced.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art should be considered to be within the technical scope of the present invention, and the technical solutions and the inventive concepts thereof according to the present invention should be equivalent or changed within the scope of the present invention.
Claims (4)
1. The method and the system for predicting the fault of the complex equipment based on the LPP-HMM method are characterized in that: comprises the following steps;
s1, data processing and feature extraction: collecting sample data, and preprocessing the sample data by adopting an LPP algorithm;
s2, determining the number L of hidden states of an HMM model according to the actual condition of equipment, setting initial parameters of the model, establishing an initial HMM model lambda (pi, A and B), carrying out maximum likelihood estimation, and selecting k feature vectors with any side and any intention to form a group of state sequences to train the HMM model of each state type until a convergence condition is met, wherein the state types are hidden states, and each state type needs to train one HMM model;
s3, inputting the original characteristic information in the preprocessed data in the S1 into a fuzzy neural network, and performing network calculation to select n HMM models with the strongest relevance;
s4, selecting N original signals in S1 according to the state prediction information of the fuzzy neural network, and enabling the N original signals to pass through the sameThe feature dimension reduction method comprises the steps of obtaining N low-dimensional feature vectors, constructing the N feature vectors into a group of feature sequences, inputting the N HMM models selected in the trained S3, and calculating likelihood probability values p (o | lambda), p (o | lambda)1)K p(o|λn) And the HMM model corresponding to the maximum probability value is the state of the current system.
2. The LPP-HMM method based complex equipment fault prediction method and system of claim 1, wherein: the LPP algorithm in S1 preprocesses the sample data, including the following steps:
a1, constructing a sample point x according to the sample dataiThere are two methods for selecting the neighborhood of the sample point: a spherical approximation method and a k nearest neighbor method;
a2, calculating a weight matrix W, and constructing the weight matrix by a thermonuclear method or a reduction method;
a3 by formula XLXTK＝λXDXTAnd K, acquiring a mapping matrix, reducing the dimension of the original high-dimensional data space according to the mapping matrix, acquiring the inherent low-dimensional manifold characteristics, and keeping the original characteristic information.
3. The LPP-HMM method based complex equipment fault prediction method and system of claim 1, wherein: the specific steps of training the HMM model in the S2 are as follows:
b1, setting L system state modes aiming at the characteristics of the monitored object, such as: carrying out m times of signal acquisition on the measuring point signal of each state mode in a normal state, an intermediate state, an intermittent state, a fault state and the like to obtain L multiplied by m original characteristic signals;
b2, sampling each original characteristic signal to construct a pi multiplied by n-dimensional high-dimensional original signal characteristic space, and forming Lxm m multiplied by n-dimensional high-dimensional spaces;
b3, learning the training sample by adopting an LPP algorithm, selecting an optimal parameter neighborhood factor K and an optimal embedding dimension d, and converting a complex high-dimensional data space into a low mxd-dimensional feature space, wherein d is more than or equal to 1 and less than or equal to n;
and B4, for each system state, randomly selecting C feature vectors to form a group of state sequences to train the HMM models of the system state, wherein C is less than or equal to m, and training L HMM models.
4. The LPP-HMM method based complex equipment fault prediction method and system of claim 1, wherein: the network calculation through the fuzzy neural network in S3 includes the following specific steps:
c1, determining a fuzzy relation matrix, and determining the membership degree of a single symptom to a single reason by adopting a weighted statistical method;
rij＝max(γ(k)+μij (k)-γ(k)μij (k))k＝0,1,2
in the formula: in the formula, gamma(k)The weight values of three terms of empirical statistics (k is 0), mechanism analysis (k is 1) and field analysis (k is 2), and gamma is(k)>0 and
the correction can be made according to different stages of system operation and the mastery of various factors, and then, muij (k)Is the value of the credit of each factor;
c2, selecting a fuzzy comprehensive evaluation model;
and C3, outputting a result vector as a group of membership values according to the fuzzy comprehensive evaluation model, and judging which category the fault belongs to and the fault reason thereof according to the group of membership values.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010827424.0A CN111915101A (en) | 2020-08-17 | 2020-08-17 | Complex equipment fault prediction method and system based on LPP-HMM method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010827424.0A CN111915101A (en) | 2020-08-17 | 2020-08-17 | Complex equipment fault prediction method and system based on LPP-HMM method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111915101A true CN111915101A (en) | 2020-11-10 |
Family
ID=73278153
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010827424.0A Pending CN111915101A (en) | 2020-08-17 | 2020-08-17 | Complex equipment fault prediction method and system based on LPP-HMM method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111915101A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112925292A (en) * | 2021-01-24 | 2021-06-08 | 国网辽宁省电力有限公司电力科学研究院 | Generator set process monitoring and fault diagnosis method based on layered partitioning |
| CN113033078A (en) * | 2021-03-05 | 2021-06-25 | 国网安徽省电力有限公司 | Construction method, system and early warning method of fault early warning model of relay protection equipment |
| CN114662371A (en) * | 2022-05-24 | 2022-06-24 | 深圳市信润富联数字科技有限公司 | Knowledge distillation-based light PHM system implementation method, device and system |
| CN115599079A (en) * | 2022-12-15 | 2023-01-13 | 中国航空工业集团公司西安飞机设计研究所（Cn） | Airborne PHM system fault diagnosis function test method |
| CN112925292B (en) * | 2021-01-24 | 2024-05-14 | 国网辽宁省电力有限公司电力科学研究院 | Generator set process monitoring and fault diagnosis method based on layered and segmented |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103234767A (en) * | 2013-04-21 | 2013-08-07 | 蒋全胜 | Nonlinear fault detection method based on semi-supervised manifold learning |
| CN103544963A (en) * | 2013-11-07 | 2014-01-29 | 东南大学 | Voice emotion recognition method based on core semi-supervised discrimination and analysis |
| CN110705812A (en) * | 2019-04-15 | 2020-01-17 | 中国石油大学(华东) | Industrial fault analysis system based on fuzzy neural network |
-
2020
- 2020-08-17 CN CN202010827424.0A patent/CN111915101A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103234767A (en) * | 2013-04-21 | 2013-08-07 | 蒋全胜 | Nonlinear fault detection method based on semi-supervised manifold learning |
| CN103544963A (en) * | 2013-11-07 | 2014-01-29 | 东南大学 | Voice emotion recognition method based on core semi-supervised discrimination and analysis |
| CN110705812A (en) * | 2019-04-15 | 2020-01-17 | 中国石油大学(华东) | Industrial fault analysis system based on fuzzy neural network |
Non-Patent Citations (2)
| Title |
|---|
| 李向前: "复杂装备故障预测与健康管理关键技术研究", 《中国博士学位论文全文数据库（电子期刊）》 * |
| 王玲: "复杂装备故障智能诊断技术研究", 《中国硕士学位论文全文数据库（电子期刊）》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112925292A (en) * | 2021-01-24 | 2021-06-08 | 国网辽宁省电力有限公司电力科学研究院 | Generator set process monitoring and fault diagnosis method based on layered partitioning |
| CN112925292B (en) * | 2021-01-24 | 2024-05-14 | 国网辽宁省电力有限公司电力科学研究院 | Generator set process monitoring and fault diagnosis method based on layered and segmented |
| CN113033078A (en) * | 2021-03-05 | 2021-06-25 | 国网安徽省电力有限公司 | Construction method, system and early warning method of fault early warning model of relay protection equipment |
| CN113033078B (en) * | 2021-03-05 | 2022-06-03 | 国网安徽省电力有限公司 | Construction method, system and early warning method of fault early warning model of relay protection equipment |
| CN114662371A (en) * | 2022-05-24 | 2022-06-24 | 深圳市信润富联数字科技有限公司 | Knowledge distillation-based light PHM system implementation method, device and system |
| CN115599079A (en) * | 2022-12-15 | 2023-01-13 | 中国航空工业集团公司西安飞机设计研究所（Cn） | Airborne PHM system fault diagnosis function test method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107977932B (en) | Face image super-resolution reconstruction method based on discriminable attribute constraint generation countermeasure network | |
| CN112784881B (en) | Network abnormal flow detection method, model and system | |
| CN111915101A (en) | Complex equipment fault prediction method and system based on LPP-HMM method | |
| CN109522857B (en) | People number estimation method based on generation type confrontation network model | |
| CN111353153B (en) | GEP-CNN-based power grid malicious data injection detection method | |
| CN110287777B (en) | Golden monkey body segmentation algorithm in natural scene | |
| CN110726898B (en) | Power distribution network fault type identification method | |
| CN111397902B (en) | Rolling bearing fault diagnosis method based on feature alignment convolutional neural network | |
| CN111222457B (en) | Detection method for identifying authenticity of video based on depth separable convolution | |
| CN114676742A (en) | Power grid abnormal electricity utilization detection method based on attention mechanism and residual error network | |
| CN116524200A (en) | High-voltage circuit breaker fault diagnosis method based on image recognition | |
| CN114511912A (en) | Cross-library micro-expression recognition method and device based on double-current convolutional neural network | |
| CN114821155A (en) | Multi-label classification method and system based on deformable NTS-NET neural network | |
| WO2022188425A1 (en) | Deep learning fault diagnosis method integrating prior knowledge | |
| CN108694375B (en) | Imaging white spirit identification method applicable to multi-electronic nose platform | |
| CN114037001A (en) | Mechanical pump small sample fault diagnosis method based on WGAN-GP-C and metric learning | |
| CN112633301A (en) | Traditional Chinese medicine tongue image greasy feature classification method based on depth metric learning | |
| CN114387524B (en) | Image identification method and system for small sample learning based on multilevel second-order representation | |
| CN115840116A (en) | Switch cabinet partial discharge fault mode identification method based on GAN-CNN | |
| CN115908842A (en) | Transformer partial discharge data enhancement and identification method | |
| CN115204272A (en) | Industrial system fault diagnosis method and equipment based on multi-sampling rate data | |
| CN115578325A (en) | Image anomaly detection method based on channel attention registration network | |
| CN113673451A (en) | Graph volume module for extracting image features of tissue cytology pathology pieces | |
| CN113011370A (en) | Multi-state face recognition method based on deep learning | |
| CN111414286A (en) | Fault diagnosis method for unbalanced hard disk data based on deep 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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20201110 |