🖍️
gitbook_docs
  • Introduction
  • Machine Learning
    • Recommended Courses
      • For Undergrad Research
      • Math for Machine Learning
    • ML Notes
      • Covariance Correlation
      • Feature Selection
      • Linear Regression
      • Entropy, Cross-Entropy, KL Divergence
      • Bayesian Classifier
        • Terminology Review
        • Bayesian Classifier for Normally Distributed classes
      • Linear Discriminant Analysis
      • Logistic Regression
        • Logistic Regression Math
      • Logistic Regression-MaximumLikelihood
      • SVM
        • SVM concept
        • SVM math
      • Cross Validation
      • Parameter, Density Estimation
        • MAP, MLE
        • Gaussian Mixture Model
      • E-M
      • Density Estimation(non-parametric)
      • Unsupervised Learning
      • Clustering
      • kNN
      • WaveletTransform
      • Decision Tree
    • Probability and Statistics for Machine Learning
      • Introduction
      • Basics of Data Analysis
      • Probability for Discrete Random Variable
      • Poisson Distribution
      • Chi-Square Distribution
      • P-value and Statistical Hypothesis
      • Power and Sample Size
      • Hypothesis Test Old
      • Hypothesis Test
      • Multi Armed Bandit
      • Bayesian Inference
      • Bayesian Updating with Continuous Priors
      • Discrete Distribution
      • Comparison of Bayesian and frequentist inference
      • Confidence Intervals for Normal Data
      • Frequenist Methods
      • Null Hypothesis Significance Testing
      • Confidence Intervals: Three Views
      • Confidence Intervals for the Mean of Non-normal Data
      • Probabilistic Prediction
  • Industrial AI
    • PHM Dataset
    • BearingFault_Journal
      • Support Vector Machine based
      • Autoregressive(AR) model based
      • Envelope Extraction based
      • Wavelet Decomposition based
      • Prediction of RUL with Deep Convolution Nueral Network
      • Prediction of RUL with Information Entropy
      • Feature Model and Feature Selection
    • TempCore Journal
      • Machine learning of mechanical properties of steels
      • Online prediction of mechanical properties of hot rolled steel plate using machine learning
      • Prediction and Analysis of Tensile Properties of Austenitic Stainless Steel Using Artificial Neural
      • Tempcore, new process for the production of high quality reinforcing
      • TEMPCORE, the most convenient process to produce low cost high strength rebars from 8 to 75 mm
      • Experimental investigation and simulation of structure and tensile properties of Tempcore treated re
    • Notes
  • LiDAR
    • Processing of Point Cloud
    • Intro. 3D Object Detection
    • PointNet
    • PointNet++
    • Frustrum-PointNet
    • VoxelNet
    • Point RCNN
    • PointPillars
    • LaserNet
  • Simulator
    • Simulator List
    • CARLA
    • Airsim
      • Setup
      • Tutorial
        • T#1
        • T#2
        • T#3: Opencv CPP
        • T#4: Opencv Py
        • Untitled
        • T#5: End2End Driving
  • Resources
    • Useful Resources
    • Github
    • Jekyll
  • Reinforcement Learning
    • RL Overview
      • RL Bootcamp
      • MIT Deep RL
    • Textbook
    • Basics
    • Continuous Space RL
  • Unsupervised Learning
    • Introduction
  • Unclassified
    • Ethics
    • Conference Guideline
  • FPGA
    • Untitled
  • Numerical Method
    • NM API reference
Powered by GitBook
On this page
  • Methodology
  • Framework
  • Feature Extraction
  • Model Construction
  • Deep Neural Network
  • Smoothing
  • Experiment and Analysis
  • Data Description
  • Experiment
  • Comparison and analysis

Was this helpful?

  1. Industrial AI
  2. BearingFault_Journal

Prediction of RUL with Deep Convolution Nueral Network

PreviousWavelet Decomposition basedNextPrediction of RUL with Information Entropy

Last updated 3 years ago

Was this helpful?

Lei Ren, Yaqiang Sun, Hao Wang, Lin Zhang, IEEE Access, 2018

Methodology

Framework

  1. 스펙트럼 차원(fft)과 진동 신호 차원은 2560 차원

  2. 스펙트럼을 64개의 블록으로 나눔

  3. 각 블록의 주파수 대역에서 maximum amplitude를 해당 주파수 대역의 eigenvalue로 선택

  4. 64개의 Spectrum-Principal-Energy-Vector가 얻어짐

  5. 각각의 time point t에서는 원하는 t-63, t-62, ..., t data point를 사용(총 64개)

    (time에서 특정 주파수 신호의 변화를 사용)

  6. feature map 크기가 64*64가 됨

  7. CNN 결과 360차원 벡터 얻음

  8. RUL 단계에서는 360차원 벡터를 DNN의 input으로 사용

Feature Extraction

  • 기존의 time domain feature, freq. domain feature는 정보 손실이 큼

  • 하지만 2560 차원의 진동 신호를 전부 input으로 쓰기엔 너무 큼

  • 진동 신호에는 noise도 많음

-> Spectrum-Principal-Energy-Vector 제안

where k = 1, 2, 3, ..., K

K = 64

X_SPEV is a 64-dimensional column vector

Model Construction

여러 번 실험해서 찾은 가장 적합한 structure

Relu 사용

RUL 예측에 필요한 정보만 남김

fully connected layer에서 얻은 feature vector와 eigenvalue로 regression prediction 수행

Deep Neural Network

Smoothing

RUL이 discrete하게 나오는 경우 필요함

running time과 RUL은 linear한 관계

-> 10개의 point씩 linear regression함

Experiment and Analysis

Data Description

Experiment

Feature map construction

spectrum 64등분 하고 local maximum 찾기

CNN의 input으로 사용

어두운 부분은 에너지가 낮고 밝은 부분은 높음

가로는 freq., 세로는 time

Model prediction

training loss: mean squared error

epoch 100

Smoothing

Comparison and analysis

The validity of feature map

Analysis of feature extraction methods

  • Spectral Mean Energy Vector와 비교

    • 주파수 영역에서 64등분 할 때 평균값 사용

Contrast with traditional features

  • 시간-주파수 영역의 feature와 비교

테스트하기 위해 SVM과 DNN 사용

The validity of the CNN Model

동일한 feature로 비교

  • DNN, SVM과 비교

다른 feature로 비교

  • wavelet 변환으로 얻은 time-freq. feature 사용