CN113257329A - Memory fault diagnosis method based on machine learning - Google Patents

Memory fault diagnosis method based on machine learning Download PDF

Info

Publication number
CN113257329A
CN113257329A CN202110390638.0A CN202110390638A CN113257329A CN 113257329 A CN113257329 A CN 113257329A CN 202110390638 A CN202110390638 A CN 202110390638A CN 113257329 A CN113257329 A CN 113257329A
Authority
CN
China
Prior art keywords
model
fault diagnosis
optimized
training
som
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
CN202110390638.0A
Other languages
Chinese (zh)
Inventor
怀娜娜
肖爱斌
魏志超
张洪伟
刘辉
张竞择
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.)
China Academy of Space Technology CAST
Original Assignee
China Academy of Space Technology CAST
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 China Academy of Space Technology CAST filed Critical China Academy of Space Technology CAST
Priority to CN202110390638.0A priority Critical patent/CN113257329A/en
Publication of CN113257329A publication Critical patent/CN113257329A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G11—INFORMATION STORAGE
    • G11C—STATIC STORES
    • G11C29/00—Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation
    • G11C29/04—Detection or location of defective memory elements, e.g. cell constructio details, timing of test signals
    • G11C29/08—Functional testing, e.g. testing during refresh, power-on self testing [POST] or distributed testing
    • G11C29/12—Built-in arrangements for testing, e.g. built-in self testing [BIST] or interconnection details
    • 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
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
    • 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
    • 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)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • General Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Software Systems (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Biomedical Technology (AREA)
  • Evolutionary Biology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Health & Medical Sciences (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Medical Informatics (AREA)
  • Debugging And Monitoring (AREA)

Abstract

The application discloses a memory fault diagnosis method based on machine learning, which comprises the following steps: determining a characteristic data set corresponding to at least one fault characteristic of a memory to be detected, constructing a fault diagnosis model, and training the fault diagnosis model according to the characteristic data set to obtain an optimized fault diagnosis model, wherein the fault diagnosis model is a neural network model; and acquiring the characteristic data of the memory to be detected in real time, and performing fault diagnosis according to the characteristic data and the optimized fault diagnosis model to obtain an analysis result. The method and the device solve the technical problems of low fault diagnosis efficiency, and low system operation safety and reliability in the prior art.

Description

Memory fault diagnosis method based on machine learning
Technical Field
The application relates to the technical field of memory fault diagnosis, in particular to a memory fault diagnosis method based on machine learning.
Background
In the aerospace field, a memory is one of important components of an on-orbit system, the memory serving as an electronic device is very prone to failure, and once the memory fails, huge losses are caused in some important defense and military fields, so that severe challenges are brought to the safety and reliability of the operation of the whole system. Therefore, the on-orbit fault diagnosis capability of the memory is increasingly emphasized, and it is very important and urgent to grasp the on-orbit health state information of the memory in real time and quickly diagnose the fault type, and the on-orbit fault diagnosis of the memory becomes an important content of the stability research of the modern electronic device.
At present, a common fault diagnosis method is mainly used for a device-level screening test before a memory is installed for use, and the fault diagnosis method comprises the following steps: whether the device has a fault is judged by monitoring parameters such as voltage and current of the device and whether read-write data are wrong or not in the using process of the memory; when the memory is determined to have a fault, fault location is mainly carried out by manually analyzing the memory and even system data, particularly in the on-orbit operation of a spacecraft, delay exists in downloading fault data, the data volume is small, fault recurrence cannot be carried out, the analysis of fault reasons by ground operators is difficult, the fault location is inaccurate, and the analysis efficiency is low.
Disclosure of Invention
The technical problem that this application was solved is: the method comprises the steps of determining a feature data set corresponding to at least one fault feature of a memory to be detected and constructing a fault diagnosis model, then training the fault diagnosis model according to the feature data set based on the machine learning method to obtain an optimized fault diagnosis model, and then rapidly and accurately carrying out fault analysis and fault identification on the memory according to the optimized fault diagnosis model, so that the efficiency of fault diagnosis and the safety and reliability of system operation are improved.
In a first aspect, an embodiment of the present application provides a memory fault diagnosis method based on machine learning, where the method includes:
determining a characteristic data set corresponding to at least one fault characteristic of a memory to be detected, constructing a fault diagnosis model, and training the fault diagnosis model according to the characteristic data set to obtain an optimized fault diagnosis model, wherein the fault diagnosis model is a neural network model;
and acquiring the characteristic data of the memory to be detected in real time, and performing fault diagnosis according to the characteristic data and the optimized fault diagnosis model to obtain an analysis result.
In the scheme provided by the embodiment of the application, the characteristic data set corresponding to at least one fault characteristic of the memory to be detected is determined, the fault diagnosis model is constructed, the optimized fault diagnosis model is obtained by training the fault diagnosis model according to the characteristic data set based on a machine learning method, and then fault analysis and fault recognition are rapidly and accurately performed on the memory according to the optimized fault diagnosis model, so that the efficiency of fault diagnosis and the safety and reliability of system operation are improved beneficially
Optionally, the fault diagnosis model includes a self-organizing feature mapping (SOM) model, a Support Vector Machine (SVM) model or a hybrid model of the SOM model and the SVM model.
Optionally, before training the fault diagnosis model according to the feature data set to obtain the optimized fault diagnosis model, the method further includes:
dividing the characteristic data set into a training set and a test set according to a preset proportion;
and compressing the number of the data samples in the training set through undersampling to obtain a compressed training set.
In the scheme provided by the embodiment of the application, the number of the data samples in the training set is compressed through undersampling treatment to obtain the compressed training set, namely, the set with more data samples is undersampled to compress the number of the data samples, so that the number of the samples is less than the number of the samples, and the phenomenon that the classification effect is deteriorated due to unbalance of small samples is avoided.
Optionally, before training the fault diagnosis model according to the feature data set to obtain the optimized fault diagnosis model, the method further includes:
and carrying out normalization processing on the characteristic data set to obtain a normalized characteristic data set.
According to the scheme provided by the embodiment of the application, the characteristic data set is subjected to normalization processing to obtain the normalized characteristic data set, so that the influence of different types of data on different dimension units is eliminated, and the convergence rate of the data in the model is accelerated.
Optionally, if the fault diagnosis model is an SOM model, training the fault diagnosis model according to the feature data set to obtain an optimized fault diagnosis model, including:
initializing the SOM model, setting the number of input neurons and weight vectors, and inputting the training sample to an input layer of the SOM model to determine a winning neuron;
and determining a winning domain by taking the winning neuron as a center, and adjusting the weight vector according to the winning domain until the output result of the SOM meets a preset requirement to obtain an optimized SOM.
Optionally, if the fault diagnosis model is an SVM model, training the fault diagnosis model according to the feature data set to obtain an optimized fault diagnosis model, including:
determining a kernel function, an optimal penalty factor and kernel function parameters of the SVM model;
and training the SVM model according to the feature data set, the kernel function, the optimal penalty factor and the kernel function parameter to obtain an optimized SVM model.
Optionally, if the fault diagnosis model is a mixed model of an SOM model and an SVM model, training the fault diagnosis model according to the feature data set to obtain an optimized fault diagnosis model, including:
training the SOM according to the training set to obtain an optimized SOM;
compressing the training set to obtain a compressed training set, and optimizing the SVM model according to the compressed training set to obtain an optimized SVM model;
and obtaining an optimized mixed model according to the optimized SOM model and the optimized SVM model.
Drawings
Fig. 1 is a schematic flowchart of a method for diagnosing a memory fault based on machine learning according to an embodiment of the present disclosure;
FIG. 2 is a schematic diagram illustrating an optimization procedure of an SOM model according to an embodiment of the present disclosure;
FIG. 3 is a schematic diagram illustrating an optimization procedure of an SVM model according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of a hybrid model optimization procedure of an SOM model and an SVM model provided in the embodiments of the present application;
FIG. 5 is a flowchart illustrating a method for diagnosing memory failure based on machine learning according to an embodiment of the present disclosure;
fig. 6 is a schematic structural diagram of a memory fault diagnosis system based on machine learning according to an embodiment of the present disclosure.
Detailed Description
In the solutions provided in the embodiments of the present application, the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
The following describes in further detail a machine learning-based memory failure diagnosis method provided by an embodiment of the present application with reference to the drawings of the specification, and a specific implementation manner of the method may include the following steps (a method flow is shown in fig. 1):
step 101, determining a feature data set corresponding to at least one fault feature of a memory to be detected, constructing a fault diagnosis model, and training the fault diagnosis model according to the feature data set to obtain an optimized fault diagnosis model, wherein the fault diagnosis model is a neural network model.
Specifically, in the solution provided in the embodiment of the present application, the memory to be detected includes, but is not limited to, a random access memory, a read only memory, and an external memory, and the fault characteristics include, but is not limited to, a fault type and fault data, where the fault data includes static leakage current data, dynamic current data, and high-speed pulse rising edge and falling edge data when a fault is determined.
Further, after determining the feature data set corresponding to at least one fault feature of the memory, a fault diagnosis model is constructed, wherein the types of the fault diagnosis models are various, and several of the fault diagnosis models are described as examples below.
In one possible implementation, the fault diagnosis model includes a self-organizing feature mapping (SOM) model, a Support Vector Machine (SVM) model, or a hybrid model of the SOM model and the SVM model.
Specifically, in the solution provided in the embodiment of the present application, the fault diagnosis model includes a Self-organizing feature mapping (SOM) model, a Support Vector Machine (SVM) model, or a hybrid model of an SOM model and an SVM model; the SOM model is a self-organizing competitive artificial neural network, and the SVM model is a two-classification model.
And 102, acquiring the characteristic data of the memory to be detected in real time, and performing fault diagnosis according to the characteristic data and the optimized fault diagnosis model to obtain an analysis result.
Further, in order to prevent the occurrence of a phenomenon that the classification effect is deteriorated due to imbalance of the small samples, in a possible implementation manner, before training the fault diagnosis model according to the feature data set to obtain the optimized fault diagnosis model, the method further includes: dividing the characteristic data set into a training set and a test set according to a preset proportion; and compressing the number of the data samples in the training set through undersampling to obtain a compressed training set.
In the solution provided in the embodiment of the present application, the preset ratio may be 7:3, or may be other ratios, which is not limited herein. Specifically, firstly, the feature data set is divided into a training set and a test set according to a preset proportion, then the number of data samples in the training set is compressed through undersampling to obtain a compressed training set, that is, the set with more data samples is undersampled to compress the number of the data samples, so that the number of the samples is less than the number of the samples.
Further, in order to eliminate the influence of different types of data not in the same dimension unit and to accelerate the convergence rate of data in the model, in a possible implementation manner, before training the fault diagnosis model according to the feature data set to obtain the optimized fault diagnosis model, the method further includes: and carrying out normalization processing on the characteristic data set to obtain a normalized characteristic data set.
In the solution provided in the embodiment of the present application, the fault diagnosis model includes an SOM model, or a mixed model of an SOM model and an SVM model, different fault diagnosis models have different training and optimization manners, and for convenience of understanding, an optimization process of each type of fault diagnosis model is briefly described below.
Firstly, the fault diagnosis model is an SOM model
In a possible implementation manner, if the fault diagnosis model is an SOM model, training the fault diagnosis model according to the feature data set to obtain an optimized fault diagnosis model includes:
initializing the SOM model, setting the number of input neurons and weight vectors, and inputting the training sample to an input layer of the SOM model to determine a winning neuron;
and determining a winning domain by taking the winning neuron as a center, and adjusting the weight vector according to the winning domain until the output result of the SOM meets a preset requirement to obtain an optimized SOM.
Specifically, in the solution provided in the embodiment of the present application, referring to fig. 2, the steps of training the SOM model are as follows:
a) initializing the network, and setting the number of input neurons and weight vectors.
b) And transmitting the sample to be trained into an input layer of the network.
c) Looking for winning neurons.
d) And the SOM network determines the area of win by taking the winning neuron as the center.
e) And adjusting the weight value by using a formula 1:
wij(t+1)＝wij(t)+η(t)(xi(t)-wij(t)) (1)
where η ∈ (0,1), η (t) can be obtained from equation 2:
Figure BDA0003016592200000061
where T represents the step size.
f) And calculating the output result by using the formula 3.
Figure BDA0003016592200000062
Wherein f () is a function of 0 ~ 1.
g) And c, outputting whether the output meets the requirements, if so, finishing the model training, marking the standard fault, otherwise, returning to the step b, and carrying out the next round of learning.
Secondly, the fault diagnosis model is an SVM model
In a possible implementation manner, if the fault diagnosis model is an SVM model, training the fault diagnosis model according to the feature data set to obtain an optimized fault diagnosis model includes: determining a kernel function, an optimal penalty factor and kernel function parameters of the SVM model; and training the SVM model according to the feature data set, the kernel function, the optimal penalty factor and the kernel function parameter to obtain an optimized SVM model.
Specifically, in the solution provided in the embodiment of the present application, referring to fig. 3, the training steps of the SVM model are as follows:
1) and selecting a kernel function.
2) And seeking an optimal penalty factor and a kernel function parameter in the model.
3) And training the model to obtain the optimal SVM model.
Thirdly, the fault diagnosis model is a mixed model of an SOM model and an SVM model
In a possible implementation manner, if the fault diagnosis model is a mixed model of an SOM model and an SVM model, training the fault diagnosis model according to the feature data set to obtain an optimized fault diagnosis model, including: training the SOM according to the training set to obtain an optimized SOM; compressing the training set to obtain a compressed training set, and optimizing the SVM model according to the compressed training set to obtain an optimized SVM model; and obtaining an optimized mixed model according to the optimized SOM model and the optimized SVM model.
Specifically, in the solution provided in the embodiment of the present application, referring to fig. 4, the training steps of the hybrid model of the SOM model and the SVM model are as follows:
(a) and using the training set in the data set for the SOM model, and training to obtain the optimal SOM model.
(b) And compressing the training set according to the clustering result.
(c) And applying the compressed training set to an SVM model, and training to obtain an optimal SVM model, namely an optimal mixed model.
(d) And selecting one piece of data in the test set as the input of the optimal mixed model.
(e) And outputting the fault type and degree through diagnosis of the hybrid model.
Specifically, in the solution provided in the embodiment of the present application, the optimized SOM model, the optimized SVM model, and the hybrid model of the optimized SOM model and the SVM model all include an optimal fault diagnosis multi-classifier, and each fault diagnosis classifier can be used to separate out fault feature data of a corresponding type.
Further, in order to facilitate understanding of the above fault diagnosis process, a brief description is provided below, and reference is made to fig. 5, which is a schematic flow chart of a memory fault diagnosis method provided in an embodiment of the present application; in fig. 5, a fault feature is selected from a memory to be detected, a feature data set is obtained according to the fault feature, the feature data set is normalized and a fault diagnosis model is constructed, the fault diagnosis model is trained according to the feature data set to obtain an optimal fault diagnosis multi-classifier, and then the fault is analyzed and diagnosed according to the optimal fault diagnosis multi-classifier to obtain an analysis and diagnosis result.
Further, in order to execute the above-described failure diagnosis algorithm, a brief description of a failure diagnosis system is provided below. Referring to fig. 6, a schematic structural diagram of a memory fault diagnosis system based on machine learning according to an embodiment of the present application is provided. In fig. 6, the system includes a core board 61, an interface 62, a memory adaptation board 63, a memory under test 64, a static leakage current monitor 65, a dynamic current monitor 66, and a high-speed pulse counting circuit 67.
Specifically, the core board 61 is a Field Programmable Gate Array (FPGA), for example, the core board 61 is a ZYNQ XAZU3 FPGA. After the optimized fault diagnosis model is obtained through the fault diagnosis model training, the optimized fault diagnosis model is burnt into the core board 61, and the core board 61 is used for operating the fault diagnosis model in the fault diagnosis process; the interface 62 includes, but is not limited to, an IO interface, an ethernet interface, a CAN interface, and a USB interface, and the interface 62 is connected to the core board 61; the memory adapter board 63 comprises various types of interfaces for connecting the memory 64 to be detected with the core board 61, the static leakage current monitor 65, the dynamic current monitor 66 and the high-speed pulse counting circuit 67; the memory to be detected 64 is connected with the memory adapting board 63; the static leakage current monitor 65 is used for monitoring the static current of the memory 64 to be detected in the working process; the dynamic current monitor 66 is used for monitoring the dynamic current of the memory 64 to be detected during operation; the high-speed pulse counting circuit 67 is used to monitor rising and falling edges, and the high-speed pulse counting circuit 67 includes a delay time monitor 671 and a bit flip monitor 672.
According to the scheme provided by the embodiment of the application, the characteristic data set corresponding to at least one fault characteristic of the memory to be detected is determined, the fault diagnosis model is constructed, then the optimized fault diagnosis model is obtained by training the fault diagnosis model according to the characteristic data set based on a machine learning method, and then fault analysis and fault recognition are rapidly and accurately carried out on the memory according to the optimized fault diagnosis model, so that the efficiency of fault diagnosis and the safety and reliability of system operation are improved.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present application without departing from the spirit and scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is intended to include such modifications and variations as well.

Claims (7)

1. A memory fault diagnosis method based on machine learning is characterized by comprising the following steps:
determining a characteristic data set corresponding to at least one fault characteristic of a memory to be detected, constructing a fault diagnosis model, and training the fault diagnosis model according to the characteristic data set to obtain an optimized fault diagnosis model, wherein the fault diagnosis model is a neural network model;
and acquiring the characteristic data of the memory to be detected in real time, and performing fault diagnosis according to the characteristic data and the optimized fault diagnosis model to obtain an analysis result.
2. The method of claim 1, wherein the fault diagnosis model comprises a self-organizing feature mapping (SOM) model, a Support Vector Machine (SVM) model, or a hybrid model of SOM and SVM models.
3. The method of claim 2, wherein prior to training the fault diagnosis model based on the feature data set to obtain the optimized fault diagnosis model, further comprising:
dividing the characteristic data set into a training set and a test set according to a preset proportion;
and compressing the number of the data samples in the training set through undersampling to obtain a compressed training set.
4. The method according to any one of claims 1 to 3, wherein before training the fault diagnosis model according to the feature data set to obtain the optimized fault diagnosis model, the method further comprises:
and carrying out normalization processing on the characteristic data set to obtain a normalized characteristic data set.
5. The method of claim 4, wherein training the fault diagnosis model according to the feature data set to obtain an optimized fault diagnosis model if the fault diagnosis model is an SOM model comprises:
initializing the SOM model, setting the number of input neurons and weight vectors, and inputting the training sample to an input layer of the SOM model to determine a winning neuron;
and determining a winning domain by taking the winning neuron as a center, and adjusting the weight vector according to the winning domain until the output result of the SOM meets a preset requirement to obtain an optimized SOM.
6. The method of claim 5, wherein training the fault diagnosis model according to the feature data set to obtain an optimized fault diagnosis model if the fault diagnosis model is an SVM model comprises:
determining a kernel function, an optimal penalty factor and kernel function parameters of the SVM model;
and training the SVM model according to the feature data set, the kernel function, the optimal penalty factor and the kernel function parameter to obtain an optimized SVM model.
7. The method of claim 6, wherein if the fault diagnosis model is a hybrid model of an SOM model and an SVM model, training the fault diagnosis model according to the feature data set to obtain an optimized fault diagnosis model comprises:
training the SOM according to the training set to obtain an optimized SOM;
compressing the training set to obtain a compressed training set, and optimizing the SVM model according to the compressed training set to obtain an optimized SVM model;
and obtaining an optimized mixed model according to the optimized SOM model and the optimized SVM model.
CN202110390638.0A 2021-04-12 2021-04-12 Memory fault diagnosis method based on machine learning Pending CN113257329A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110390638.0A CN113257329A (en) 2021-04-12 2021-04-12 Memory fault diagnosis method based on machine learning

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110390638.0A CN113257329A (en) 2021-04-12 2021-04-12 Memory fault diagnosis method based on machine learning

Publications (1)

Publication Number Publication Date
CN113257329A true CN113257329A (en) 2021-08-13

Family

ID=77220632

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110390638.0A Pending CN113257329A (en) 2021-04-12 2021-04-12 Memory fault diagnosis method based on machine learning

Country Status (1)

Country Link
CN (1) CN113257329A (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104462846A (en) * 2014-12-22 2015-03-25 山东鲁能软件技术有限公司 Intelligent device failure diagnosis method based on support vector machine
CN111190349A (en) * 2019-12-30 2020-05-22 中国船舶重工集团公司第七一一研究所 Method, system and medium for monitoring state and diagnosing fault of ship engine room equipment
CN111580506A (en) * 2020-06-03 2020-08-25 南京理工大学 Industrial process fault diagnosis method based on information fusion

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104462846A (en) * 2014-12-22 2015-03-25 山东鲁能软件技术有限公司 Intelligent device failure diagnosis method based on support vector machine
CN111190349A (en) * 2019-12-30 2020-05-22 中国船舶重工集团公司第七一一研究所 Method, system and medium for monitoring state and diagnosing fault of ship engine room equipment
CN111580506A (en) * 2020-06-03 2020-08-25 南京理工大学 Industrial process fault diagnosis method based on information fusion

Similar Documents

Publication Publication Date Title
US10706332B2 (en) Analog circuit fault mode classification method
CN113792825B (en) Fault classification model training method and device for electricity information acquisition equipment
WO2019214309A1 (en) Model test method and device
US11971959B2 (en) Data test method, electronic device and storage medium
CN110688288A (en) Automatic testing method, device, equipment and storage medium based on artificial intelligence
CN107705819A (en) A kind of storage chip sorting technique, sorter and categorizing system
CN108717496B (en) Radar antenna array surface fault detection method and system
CN113541985B (en) Internet of things fault diagnosis method, model training method and related devices
CN111290967A (en) Software defect prediction method, device, equipment and storage medium
CN111860568B (en) Method and device for balanced distribution of data samples and storage medium
CN108595884A (en) Power system transient stability appraisal procedure and device
CN109635864B (en) Fault-tolerant control method and device based on data
CN112419268A (en) Method, device, equipment and medium for detecting image defects of power transmission line
CN114355094B (en) Product reliability weak link comprehensive evaluation method and device based on multi-source information
DE102013114558B4 (en) Cut-at-Diagnosis (CID) - A method to improve the throughput of the process for increasing yield
CN110020637A (en) A kind of analog circuit intermittent fault diagnostic method based on more granularities cascade forest
CN113257329A (en) Memory fault diagnosis method based on machine learning
CN115629930A (en) Fault detection method, device and equipment based on DSP system and storage medium
CN113033845B (en) Construction method and device for power transmission resource co-construction and sharing
CN113656354A (en) Log classification method, system, computer device and readable storage medium
CN113449778A (en) Model training method for quantum data classification and quantum data classification method
CN113434408B (en) Unit test case sequencing method based on test prediction
CN109474445B (en) Distributed system root fault positioning method and device
CN113946506A (en) Software embedded point quality evaluation method and device, computing equipment and storage medium
CN117077736A (en) Fault diagnosis model training method, power grid fault diagnosis method and storage medium

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