2012phm.zip_phm2012轴承数据集,phm2012资源-CSDN文库

共3个文件

pdf：2个

docx：1个

数据集

滚动轴承

需积分: 42 76 浏览量 2020-04-10 10:34:10 上传评论 6 收藏 4.45MB ZIP 举报

资源推荐

资源详情

资源评论

收起资源包目录

2012phm.zip （3个子文件）

2012phm

数据下载链接.docx 12KB

PHM a review.pdf 2.83MB

IEEEPHM2012-Challenge-Details.pdf 2.17MB

Review Article

Prognostics and Health Management: A Review on

Data Driven Approaches

Kwok L. Tsui,

Nan Chen,

Qiang Zhou,

Yizhen Hai,

and Wenbin Wang

3,4

Department of System Engineering and Engineering Management, City University of Hong Kong, Hong Kong

Department of Industrial & Systems Engineering, National University of Singapore, Singapore 117576

Donlinks School of Economics and Management, University of Science and Technology Beijing, Beijing 100083, China

Faculty of Business and Law, Manchester Metropolitan University, Manchester M15 6BH, UK

Correspondence should be addressed to Wenbin Wang; wangwb@.com

Received  July ; Revised  October ; Accepted  October 

Academic Editor: Shaomin Wu

which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Prognostics and health management (PHM) is a framework that oers comprehensive yet individualized solutions for managing

system health. In recent years, PHM has emerged as an essential approach for achieving competitive advantages in the global

market by improving reliability, maintainability, safety, and aordability. Concepts and components in PHM have been developed

separately in many areas such as mechanical engineering, electrical engineering, and statistical science, under varied names. In

this paper, we provide a concise review of mainstream methods in major aspects of the PHM framework, including the updated

research from both statistical science and engineering, with a focus on data-driven approaches. Real world examples have been

provided to illustrate the implementation of PHM in practice.

1. Introduction

To fulll the increasing demand on functionality and quality,

modern systems are oen built with overwhelming complex-

ities. ese systems are oen featured rich electronics and

intricate interactions among subsystems/components. For

example, a typical car consists of about , functional com-

ponents, , parts, and  million lines of soware code

[].

Additionally, extremely high requirements of system

reliability are essential since a single failure can result in catas-

trophicconsequences.Despiteeveryeortmadeinthepast,

disasters keep occurring with profound implications. In June

, the Metro rail crash in Washington D.C. killed nine

people and injured dozens more, suspiciously due to sensor

circuit “anomalies” under the rail track []. Brazil blackouts

in November  aected more than  million people

andshutdowneverythingfromsubwaytolightbulbs[].

Despite the explanation attributed to lightning, wind, and

rain, it was still believed that “there was obviously some fail-

ure, either technical or human.” Other examples include the

failure of LED lighting system in Xiamen, China, two months

aer installation although the manufacturer promised ve

years’ lifespan of their products, not to mention the infamous

sudden acceleration failures of Toyota automobiles which

have signicantly damaged the company’s prot and reputa-

tion [].

In view of the high impact and extreme costs usually

associated with system failures, methods that can predict and

prevent such catastrophes have long been investigated. Appli-

cationsofdevelopedmethodsarenotrareindomainssuchas

electronics-rich systems, aerospace industries, or even public

health environment [, ]. In general, these methodologies

can all be grouped under the framework of prognostics and

health management (PHM). Particularly, prognostics is the

process of predicting the future reliability of a product by

assessing the extent of deviation or degradation of the prod-

uct from its expected normal operating conditions; health

management is the process of real time measuring, recording,

and monitoring the extent of deviation and degradation from

normal operation condition [, ]. Dierent from traditional

handbook based reliability prediction methods (e.g., U.S.

Hindawi Publishing Corporation

Mathematical Problems in Engineering

Volume 2015, Article ID 793161, 17 pages

http://dx.doi.org/10.1155/2015/793161

 Mathematical Problems in Engineering

Department of Defense Mil-Hdbk- and Telcordia SR-

(formerly [])), which assumed that constant hazard rate of

each component can be tailored by independent “modiers”

to account for various quality, operating, and environmental

conditions, PHM methodologies instead monitor the health

stateinrealtimeanddynamicallyupdatethereliability

function (hazard rates) based on in situ measurements and

tailored evolution models obtained from historical data. Due

to the success of existing PHM methodologies, there is no

need to exhibit the growing interests of studying new PHM

techniques and applying PHM to underdeveloped domains.

Nevertheless, the increasingly complex modern systems

pose new challenges on PHM. One of the most prominent

problems is called No Fault Found (NFF) problem (related

terminologies include “cannot duplicate,” “re-test OK,” “trou-

ble not identied,” and “intermittent malfunctions”) [–],

particularly in electronic-rich systems. As the name sug-

gests, it refers to the situation that no failure/fault can be

detected/replicated during laboratory tests even when the

failure has been reported in the eld []. NFF issues not only

make the prognosis and diagnosis extremely dicult, but

also can cause skyrocketing maintenance cost. As reported

by Williams et al. [], NFF failures account for more than

% of all led failures and % of overall maintenance costs

in avionics; it is estimated that NFF related activities cost the

U.S. Department of Defense ∼ billion U.S. dollars per year

[]. Evidently, NFF contributes a lot in operational cost in

many dierent application areas. On top of the maintenance

costs, potential safety hazards related to NFFs are even more

striking. For example, both Toyota and National Highway

Trac Safety Administration (NHTSA) spent quite a long

time to investigate the root causes of sudden acceleration

failures in some car models, a problem that might be linked

to  deaths in  crashes since  according to NHTSA

[]. Unfortunately, no conclusive nding has been reached

despite the eorts in trying to repeat the failures in a variety

of laboratory conditions. ese intermittent faults are also

suspected to be the main reason for other catastrophes, like

Washington Metro crashes and Brazil blackouts.

Intermittent faults or NFF problems pose signicant bar-

riers to apply traditional methods for reliability prediction,

which are oen empirical and population based. From the

aforementioned examples, we can nd that intermittent faults

are oen tightly related to the environmental conditions and

operation histories of the particular individual system. ey

can hardly be repeated due to unknown random disturbances

involved. erefore, laboratory testing and assessment can

only provide a reference on the “average” characteristics of

the whole population and are insucient to provide accurate

modeling and prediction for each individual. To reduce the

maintenance cost and eliminate safety hazards caused by NFF,

the paradigm of PHM needs to shi from empirical to data

supported and from population based to individual based.

Companioned with the challenges, the fast development

of information and sensing technology has enabled the

collection of many in situ measurements during operations

andprovidedthecapabilityofrealtimedatamanagementand

processing for each individual. ese advancements provide

us great opportunity to develop sophisticated models with

increasing accuracy of prognostics for individual items. For

instance, many dierent types of data during the whole life

cycle of the products can be easily retrieved, especially in crit-

ical applications. ese data could include production pro-

cess information, quality records, operation logs, and sensor

measurements. Moreover, unlike manually entered data used

before which are slow, costly, and error-prone, most of

current records are automatic, accurate, and timely entries

due to the advancements of technology. e use of Radio Fre-

quency Identication (RFID) technology, for example, is not

rare in supply chain distribution network, healthcare, and

even military applications, to provide reliable and timely

tracking or surveillance of products/components. Advanced

sensor technologies also enable abundant measurements at

both macro and micro scale, such as vibration, frequency

response,magneticelds,andthecurrent/voltage,toname

afew.

Inresponsetotheemergingchallengesaswellasopportu-

nities, this paper tries to review recent advancements of PHM

methodologies, with a focus on data-driven approaches, and

their applications in practice, and identify research problems

that may lead to further improvement of PHM in both theory

and practice. Before we move on to the next section, we would

like to use an example from real practice to better illustrate the

ideas in PHM.

AMotivatingExample.Bearingsarefoundinmostmechan-

ical systems with rotational components. ey provide nec-

essary support as well as constraining the moving parts

to desired motion mode. It is hard to overemphasize the

importance of keeping bearings under normal working con-

ditions in engineering applications. Breaking down of a single

bearing may cause failure of an entire system. For example,

on August th , a Qantas Boeing  aircra departing

from San Francisco International Airport encountered an

accidental engine shutdown, which has later been conrmed

to be caused by a fractured turbine blade and a failed bearing

[].

As a bearing tends to exhibit larger vibration as it

degrades, its health condition can thus be assessed via the

vibration signal collected by sensors. Such signal is oen

referred to as degradation signal intheliteratureofPHM.

When the amplitude of a bearing’s vibration exceeds certain

threshold, the bearing can be considered as no longer suitable

for further operations. Figure  demonstrates the degradation

paths of three dierent bearings, where the -axisisthework-

ing time of bearings (in minutes), the -axis is the average

amplitude of the vibration at dierent harmonic frequencies,

and the horizontal line is the vibration threshold considered

as indicator of bearing failures [, ].

Unlike conventional reliability analysis which mostly

provides population-based assessment, individualized pre-

diction results are possible by taking advantage of the degra-

dation signal. Based on the vibration data collected up to the

current time, we can build models to predict the evolution

path of vibration signal in the future and consequently

predict the remaining useful life (RUL) of an in-service

bearing by assuming its failure as the vibration signal hits

the threshold for the rst time. Unfortunately, many factors

Mathematical Problems in Engineering 

0 100 200 300 400 500 600

0.005

0.01

0.015

0.02

0.025

0.03

0.035

Time

Degradation

F : Evolution of vibration signals of three bearings.

in the degradation make the exciting task very challenging.

Figure  demonstrates several important features of the

degradation signals. For example, the degradation of bearings

exhibits two distinctive phases. At the initial stage, the vibra-

tion signals emitted are small and stable. However, aer a

change point (oen a crack appears), the magnitude of vibra-

tions increases dramatically and features large variability.

Despite similar shapes of degradation paths, the location of

changing point, increasing rate of vibration magnitude, and

soforthvaryfromonebearingtoanother.Wewillreturnto

this example in Section  for more technical details.

e rest of the paper is organized as follows. Section 

reviews advancements in PHM. Section  uses three exam-

ples to illustrate the procedures and strategies of PHM.

Section  concludes the paper with summing up comments

and future works that drive PHM further in academia and

industries.

2. Overview of Data-Driven PHM Approaches

In general, typical workows in a PHM system can be con-

ceptually illustrated, as shown in Figure .reemajortasks

can be identied in the owchart: fault diagnostics, prognos-

tics, and condition-based maintenance. e rst task is to

diagnose and identify the root causes of system failures. e

root causes identied can provide useful information for

prognostic models as well as feedback for system design

improvement. e second task takes the processed data and

existing system models or failure mode analysis as inputs

and employs the developed library of prognosis algorithms to

online update degradation models and predict failure times of

the system. e third task makes use of the prognosis results

(e.g., the distribution of remaining useful life) and considers

the cost versus benets for dierent maintenance actions to

determine when and how the preventive maintenance will be

conducted to achieve minimal operating costs and risks. All

of these three tasks need to be executed dynamically and in

real time.

Other than these three major tasks, there are also some

other important components listed in Figure . Nevertheless,

they are oen prepared oine and only timely updating may

be needed during the system operations. For example, signal

processing/feature extraction is the procedure to preprocess

the signals using rules or methods developed according to

engineering knowledge, expert experience, or statistical nd-

ings from historical data. ey serve the purpose to eliminate

noise, reduce data dimensions (complexities), and transform

the data into proper space for future analysis. Similarly,

prognosis and diagnosis algorithms can also be developed

oinetocaterthespecialcharactersofthesignalsandsystem

properties. Upon new arrival of sensing signals, appropriate

algorithms can be selected to compute distribution of RUL,

determine maintenance actions, or nd root causes of abnor-

malities.

Duetoitsundeniableimportance,recentyearshaveseen

prosperous development in dierent aspects of PHM. e

reviews on statistical data-driven approaches by Si et al. []

and Sikorska et al. []havecoveredmostofthemodels

used in RUL estimation with a statistical orientation, but

our work focuses on a wide range of the varied models

in PHM methodologies from diagnosis to prognosis with

their motivations. e subsequent sections are devoted to

discussion of research progress and open issues of these tasks/

components in PHM to provide us with an overview for

further advancement.

2.1. Signal Processing and Feature Extraction. In current

data rich environment, huge amounts of data are oen

automatically collected in a short time period. Dierent from

the problem that very limited data was available decades

ago, this overwhelming data poses new challenges in data

management, analysis, and interpretation. Consequently,

data preprocessing and feature extraction procedures become

standard in many complex systems to improve data quality,

reduce data redundancy, and boost eciency of analysis. Due

to its importance, many researchers have investigated this

problem in the literature, as summarized in some of the

review papers in dierent application areas (e.g., [–]).

Instead of giving a comprehensive review on dierent

techniques in the literature, in this section we will list some

of the commonly used methods in the context of PHM. ese

techniques can be roughly classied as statistical methods

and engineering knowledge based methods. In the rst

category, data can be transformed to optimize certain prede-

termined criteria without the input of the domain knowledge.

For example, principle component analysis (PCA) and inde-

pendent component analysis (ICA) have been used widely

to reduce data dimensions. Similar to analysis of variance

(ANOVA), distance evaluation technique (DET) []isalso

preferred and its varied versions are applied, such as a two-

stage feature selection and weighting technique (TFSWT)

via Euclidean distance evaluation technique (EDET) [], a

modied version of ANOVA, which takes both the dierence

between the variances in each group and the maximum

versus the minimum dierences between the mean of each

group into consideration. However, certain useful informa-

tion cannot be retained when dealing with highly nonlinear

 Mathematical Problems in Engineering

Condition monitoring

and data collection

Data preprocessing/

signal processing

Feature selection

Statistical modeling

Prognostics

Condition-based

maintenance

Fault

diagnostics

F : An illustration of typical ows of PHM systems.

data, as reported in []. Other techniques include mutual

information (MI) based method [], self-organizing map

(SOM) [], and density based methods []. ese tech-

niques work well on nonlinear data and hence are employed

broadly in many applications.

Methods in the second category, on the other hand,

utilize some of the domain knowledge in the process of

feature extractions. Based on the procedure of conditional-

based maintenance, the data type could be summarized into

three: value type (e.g., temperature, pressure, humidity, etc.),

waveform type (e.g., vibration data), and multidimensional

type (e.g., image data, X-ray images, etc.) [].

Waveform data analysis is among the most common

methods in diagnostics of mechanical systems, due to the

popularity of waveform data collected from sensors, particu-

larly in vibration signal analysis of rotating elements [, ].

Dierent kinds of techniques and algorithms are developed

in this eld. ey can be categorized into time-domain

analysis, frequency-domain analysis, and the combination

of both. ese methods oen create features that have clear

physical meanings or interpretations. For example, Table 

summarizes ten commonly used time-domain features [].

Instead of using the raw waveform data, these ten summary

statistics provide us with extracted information of the sig-

nals. ese ten features can be generally applied to many

applications. However, if the mechanism of how the abnor-

malities inuence the measured signal is known, extracted

featuresbasedondomainknowledgemaybemoreeective.

For example, Lei and Zuo []summarizedstatistical

features specically developed for gear damage detection.

Another time-domain analysis approach is time synchronous

average (TSA), popularly used in fault detection of rotating

equipment [, ]. e idea is to use the average over a

number of evolutions of raw signal in order to remove/reduce

noise. Time series models are naturally applied here as well

[], in an attempt to extract features based on parametric

models. For example, coecients in a tted autoregressive

moving average model (ARMA) can be indicative of the

health condition [].

T : Commonly used time-domain features.

Feature Denition

Peak value max = max

𝑗=1,...,𝑁



𝑗

Mean =



𝑁

∑

𝑗=1



𝑗

Standard deviation =



−1

𝑁

∑

𝑗=1

(

𝑗

−)

Root mean square RMS =





𝑁

∑

𝑗=1

(

𝑗

)

Skewness SK =

∑

𝑁

𝑗=1

(

𝑗

−)

(−1)

Kurtosis KU =

∑

𝑁

𝑗=1

(

𝑗

−)

(−1)

Crest indicator CI =

max





(1/)∑

𝑁

𝑗=1

(

𝑗

)

Clearance indicator CLI =

max



(1/)

∑

𝑁

𝑗=1







𝑗





Shape indicator SI =



(1/)

∑

𝑁

𝑗=1

(

𝑗

)

(1/)

∑

𝑁

𝑗=1





𝑗



Impulse indicator MI =

max



(1/)∑

𝑁

𝑗=1





𝑗



Meanwhile, it is believed that some faults will show

certain characters in frequency domain.Fouriertransformis

the most common form of further signal processing, which

decomposes a time waveform into its constituent frequencies.

Fast Fourier transform (FFT) is usually used to generate the

frequency spectrum from time series signals. A high vibra-

tion level at a particular frequency may be the signature of

a particular fault type. Besides FFT spectrum, other methods

such as cepstrum [], high-order spectra [], and holospec-

trum analysis [] are also developed for fault diagnostics in

the frequency domain.

Mathematical Problems in Engineering 

One limitation of frequency domain analysis is its inabil-

ity to handle nonstationary waveform signals, commonly

observed during machine faults. A combination of both time

and frequency domains, time-frequency analysis, has been

developed to solve the problem [, ]. A typical method

is called short-time Fourier transform (STFT) [], which

divides the whole waveform signal into segments with short-

time window and then applies Fourier transform to each

segment. Wavelet transform is another popular method with

a similar idea. Wavelet analysis has been successfully applied

to feature extraction and fault diagnostics in various applica-

tions (e.g., [–]). A review on the application of wavelet

analysis in machine fault diagnosis and fault feature extrac-

tion can be found in Peng and Chu []. Other methods of

time-frequency analysis include spectrogram [], Wigner-

Ville distribution [–], and Choi-Williams distribution

[].

2.2. Fault Diagnostics and Classication. Fault diagnostics

is designed to eciently and accurately identify the root

cause of the faults. Eective diagnosis can not only reduce

downtime and repair cost, but also provide useful informa-

tion for prognostics to improve its accuracy. Fault detection

is dened to be the task of determining if a system is

experiencing problems. Fault diagnostics, then, is the task

oflocatingthesourceofafaultonceitisdetected.Because

of its importance, researchers from dierent elds have

investigated the issue of fault diagnostics extensively, such as

in manufacturing processes [, ], discrete event systems

[, ], and communication systems and networks [, ].

We do not attempt to give a comprehensive review but focus

on the approaches that are commonly used in PHM practices.

In general, methodologies in fault diagnostics can be clas-

sied into two categories: model based approaches and model

free approaches. In the model based category, some forms of

underlying models linking failure modes and observations

are proposed. ese models are oen derived according

to rst principles and physical mechanisms. Based on the

model structure and parameters, observations can be used

to infer the root causes or the failure modes using dierent

algorithms. In contrast, model free methods oen do not

assume the knowledge of underlying processes. Although

in many cases implicit or statistical/surrogate models are

used in the fault diagnostics, we use the name to emphasize

that approaches in this category are purely data-driven

without additional assumptions on the systems’ operation

mechanisms.

In fault diagnostics, we would like to know the exact time

when a fault appears, its location, and its severity. erefore,

the diagnostics consists of three aspects: (1)anomaly detec-

tion, as we rst identify any potential performance deviation

from normal operation; (2)fault localization, which localizes

the problem to the specic component or subsystem; (3)

fault classication, which discriminates known and unknown

faults and identies the type of the fault if it is previously

known []. Many popular methods from machine learning

and articial intelligence are applied in this context, such as

support vector machine [, ], -nearest neighbors [, ],

anddecisiontrees[, ]. Among them, articial neural

network has been preferred by many engineers and widely

applied to fault diagnostics of various engineering systems

[, –]. Unsupervised machine learning algorithms such

as fuzzy -means and self-organizing maps have been applied

when no response is provided [–]. It is worth mentioning

that the sensitivity to a given fault is oen a function of oper-

ating conditions and the nature of the anomaly. erefore,

the environmental conditions need careful consideration. For

example, self-organizing maps are used for regionalization of

the system operating conditions [

]. Excellent reference for

fundamentals of these methods can be found in Bishop [],

Hastie et al. [], and Kotsiantis et al. [].

2.3. Data-Driven Prognostics Method. As mentioned in

the introduction, prognostics algorithms predict the future

reliability of a product considering current and past health

information collected. rough constant inspection, the

observed health information is oen referred to as condition

monitoring (CM) data. CM data may be directly or indirectly

related with the system health status and hence can be

viewed as system health indicators. Examples of CM data

are amount of tire wear, chemical concentration, size of a

fatigue crack, power output of an amplier, and the light

intensity of LED. As a system degrades inevitably through

usage, its health status deteriorates and is manifested through

the observed CM data (e.g., the light intensity decreases as

the LED degrades). Hence, CM is normally viewed as the

system degradation signal. Failures are oen dened as

the degradation reaches a predetermined threshold set by

experts. us, by modeling the evolution of degradation and

calculating the time it rst hits the failure threshold, we will

be able to predict the system remaining useful life (RUL).

Due to randomness in the evolution paths of the degradation,

thecalculatedRULwillbeintheformofsomeprobability

distribution. Two excellent comprehensive review papers in

RUL research can be found in Si et al. [] and Sikorska et al.

[]. e main dierence between our paper and Si et al.

[] could be understood through the illustration in Figure .

Our paper describes the entire process of data-driven

approach related to PHM, addresses the three PHM

objectives (fault diagnostics, prognostics, and conditional-

based maintenance), and discusses additional prognostics

approaches developed from other areas and relationships

among dierent approaches. On the other hand, Si et al. []

focus mainly on various modeling approaches for prognostics

and do not address the tasks before the three PHM

objectives. Sikorska et al. [] focus on evaluating the various

prognostics approaches from the industry point of view

without details on methodologies.

As a core ingredient of PHM, we summarize the major

categories of data-driven prognostics in this section.

2.3.1. Independent Increment Process Based Model. Generally

speaking, the stochastic process models (Table ) consist of

two basic components: (1) astochasticprocess{(), ∈

T, ∈ }with initial value (0) = 

,whereT is the

time space and is the state space of the process, and (2)a

boundary set ,where⊂.Taking(0) =

outside the

boundary set , the rst hitting time (FHT) is the random

评论收藏

内容反馈

晴天娃娃素晗

粉丝: 0
资源: 1

2012phm.zip

phm2012寿命预测.zip

行业文档-设计装置-PHM仿真验证中的一种进程调度方法.zip

教育科研-学习工具-PHM仿真验证中的一种进程调度方法.zip

基于PHM的STM32F4平衡力式继电器性能检测系统的设计与实现.zip

整理好的IEEE2012挑战轴承数据集

轴承matlab代码-PHM-ToolBox:这是我的本科生项目。由我和Yongyu在AIMS（中国杭州）开发

IEEE_PHM2009年竞赛数据集；含视频和说明

IEEE-PHM2009数据集

projectRUL:根据2012 PHM数据预测轴承的剩余使用寿命

基于PHM10的刀具磨损振动信号转wav数据信号测试（完整数据）.zip

2018-phm-data-challenge-master_PHM_2018_PHO_propergnf_zip_源码.zip

2018-phm-data-challenge-master_PHM_2018_PHO_propergnf_zip.zip

PHM-master.zip_故障诊断

PHM2011-数据竞赛数据打算自己用的

PHM2008 挑战赛数据集

整理好的IEEE2012挑战轴承数据集2和3

基于深度学习的刀具磨损状态监测技术的研究

基于机器学习的故障预测与健康管理（PHM）方法研究

python轴承寿命预测（附带GUI界面）源码.zip

Thermal Overstress Aging with DC at gate.zip_IGBTAging 数据_NASA I

NASA锂离子电池数据集.zip

HealthMonitoring.zip

设备剩余寿命预测学习，CMAPSS发动机数据集

NASA铣削数据集( Milling Data Set)

NASA PCoE实验室锂电池数据集

NASA MDP数据集

辛辛那提大学轴承数据集，解压后6.1GB

最新资源