Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Unbalanced hard disk data fault diagnosis method based on deep learning

A technology of deep learning and fault diagnosis, applied in faulty hardware testing methods, machine learning, electrical digital data processing, etc., can solve problems such as unbalanced data sets, achieve high robustness, low environmental requirements, and strong noise resistance Effect

Active Publication Date: 2020-07-14
TONGJI UNIV
View PDF5 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] It is a challenging task to balance the data set by oversampling and use the deep learning model to predict. Although there have been a lot of research work to solve the problem of unbalanced data sets, they have not been able to achieve better accuracy. The accuracy of fault detection

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Unbalanced hard disk data fault diagnosis method based on deep learning
  • Unbalanced hard disk data fault diagnosis method based on deep learning
  • Unbalanced hard disk data fault diagnosis method based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0043] Since the hard disk model is specified by the manufacturer, the present invention trains different models according to different models. Capacity_bytes is the capacity of each hard disk, and S.M.A.R.T. numbers are some self-monitoring analysis and reporting techniques, which can represent the characteristics of the hard disk. The hard drive data is distributed in the form of dates, and more importantly, each hard drive has nearly 120 features, and in each feature, the missing values ​​are half of the total. On this basis, the present invention adopts the three-dimensional reconstruction of the original data, and preprocesses the data features with the feature engineering Auto-encoder and the missing value processing method. The present invention adopts the sample division method for extremely unbalanced samples , the processed training set is processed by SMOTE to balance the number of unbalanced samples, and finally a model with higher accuracy is obtained by combining ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention relates to an unbalanced hard disk data fault diagnosis method based on deep learning. The method comprises the following steps: 1) carrying out preprocessing and three-dimensional reconstruction on original hard disk data; 2) performing data category balance processing on the hard disk data after three-dimensional reconstruction, and performing division to obtain a training set anda test set of a deep learning network; and 3) constructing a deep learning network and performing training through the training set, and finally performing fault prediction by adopting the trained deep learning network to judge whether the hard disk data is faulty or not. Compared with the prior art, the method has the advantages of high accuracy, high application generalization, low environmentalrequirements, risk prediction and the like.

Description

technical field [0001] The invention relates to the field of computer fault diagnosis, in particular to a deep learning-based fault diagnosis method for unbalanced hard disk data. Background technique [0002] When the equipment is down, it may delay the production line of the whole factory. What's more, it will waste a lot of time if experts check the machine. Therefore, the machine needs to be equipped with sensors and log functions to collect a large amount of historical data of the machine over a period of time, and Use machine learning to predict downtime. [0003] Many industrial machines are now IoT enabled, which means that the sensors and log functions on each machine can be transmitted to a central hub for analysis, which enables the powerful use of machine learning algorithms Imaging technology and deep learning technology can achieve more than 80% early fault detection. [0004] The data imbalance problem is common in various fields, usually, when the ratio of ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06F11/22G06N20/00
CPCG06F11/2205G06F11/2263G06F11/2273G06N20/00
Inventor 李莉刘宇广林国义
Owner TONGJI UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com