0% found this document useful (0 votes)

6 views34 pages

Lecture 9 - Introduction To PHM

Lecture 9 introduces Prognostics and Health Management (PHM) and its significance in reliability analysis, contrasting classical approaches with unit-wise assessments. It includes case studies on model-based prognostics for lithium-ion batteries and data-driven prognostics for electric cooling fans, highlighting methodologies for predicting remaining useful life. The lecture emphasizes the importance of tailored reliability estimates based on actual usage conditions.

Uploaded by

Amirhossein Shokrani

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

6 views34 pages

Lecture 9 - Introduction To PHM

Uploaded by

Amirhossein Shokrani

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 34

Lecture 9: Introduction to Prognostics and Health

Management (PHM)
Lecture #9
Outline

• Recall: Time-Dependent Reliability Analysis

• Classical Reliability Engineering
• Prognostics and Health Management (PHM)

• Human PHM Process

• Engineering PHM Process

• Overview of PHM Process
• Case Study 1: Model-based Prognostics of Lithium-ion Battery
• Case Study 2: Data-Driven Prognostics of Electric Cooling Fan

2
Recall: Definition of Reliability

Approaches to Time-Dependent Reliability Analysis

 Classical approach: traditional reliability engineering
 To fit a probability distribution to TTF data (censored and uncensored) from
accelerated life testing (ALT), other lab testing, and/or field surveillance.
 Provides a population-wise, time-dependent reliability estimate that takes
the same value for the entire population of units.
 Emerging approach: prognostics and health management (PHM)
 Assesses the reliability of each individual unit under its actual use conditions
to determine the advent of a failure.
 Provides a unit-wise, time-dependent reliability estimate that is customized
for each individual unit

3
Recall: Definition of Reliability

Approaches to Time-Dependent Reliability Analysis

Population-wise (classical) Unit-wise (PHM)

Life PDF

Life PDF
1 Lifetime 1 Lifetime
2 2
3 Unit- 3 Unit-
4 independent 4 dependent
Unit ID Unit ID

4
Overview

Prognostics and Health Management (PHM)

Sensing Diagnostics Prognostics
Fault 2 Fault 1 Prediction at 120 cycles

+ ++ ++
Prediction at 80 cycles

Probability
Feature 2
Prediction at 40

density
cycles
+ ++ +
+

Healthy
20 40 60 80 100 120 140
Feature 1 Remaining Useful Life

Battery Prognostics Bearing Prognostics

Li-Ion Rechargeable Deep Brain Stimulators Bearings/gears Wind Turbine Gearbox

5
Case Study 1: Model-based Prognostics
of Lithium-ion Battery
Why Battery Prognostics?

Lithium-ion batteries are amazing.

They power our phones, computers, and electric vehicles (EV).
They last many years.
But, they eventually die.

Application Lifetime (yr) Comments

Phone ≈3 Highest energy density, high power capability

EV >8 High energy density, high power capability
Implantable ≈ 10 Moderate energy density, moderate power capability
Grid energy 20+ Not demonstrated yet

Slide courtesy of Dr. Jeff Dahn at Dalhousie University (Presented in 2015)

7
Lithium-ion Batteries in Implantable Medical Devices

Since 2004 Since 2010

Spinal Cord Stimulators Deep Brain Stimulators
Mild electrical stimulation in the Targeted electrical stimulation to part of
spinal cord to alleviate chronic pain. brain for mitigating movement disorder.

Targeted longevity of 10 years and 100s to 1000s cycles

8
Lithium-ion Batteries in Implantable Medical Devices

Battery technology
A few 100 mAh
LiCoO2 | Graphite chemistry
Prismatic, hermetic encasement

• Calendar life: ~10 years (potentially longer?)

• Cycle life: a few 100 to a few 1000 cycles

9
How Long Can These Batteries Last?

37 oC These are actual tests

which began in 2002 and
are still running.
Normalized Capacity (%)

The daily cells have about

5000 cycles as of 2018.
C/24 (daily) or C/150
(weekly) discharge rate
37 oC

Experimental Data

Time on Test (Years)

Data courtesy of Dr. Gaurav Jain – Medtronic

10
How Long Can These Batteries Last?

37 oC These are actual tests

which began in 2002 and
are still running.
Normalized Capacity (%)

The daily cells have about

5000 cycles as of 2018.
C/24 (daily) or C/150
(weekly) discharge rate
37 oC

Experimental Data

Time on Test (Years)

Data courtesy of Dr. Gaurav Jain – Medtronic

11
Model-Based Prognostics: How Does it Work?

Current
cycle
Capacity
Normalized capacity (%)

100 estimates
Capacity estimates
90 � �
𝑄𝑄(1), …, 𝑄𝑄(t)

80
End-of-life limit

70
0 100 200 300 400 500
Cycle number

Hu C., Jain G., Tamirisa P., and Gorka T., “Method for Estimating Capacity and Predicting Remaining Useful Life of Lithium-ion
Battery,” Applied Energy, v126, p182–189, 2014.

12
Model-Based Prognostics: How Does it Work?

Current
cycle
Capacity
Normalized capacity (%)

100 estimates
Capacity estimates
90 �
𝑄𝑄(1:t)

80
End-of-life limit

70
0 100 200 300 400 500
Cycle number

Hu C., Jain G., Tamirisa P., and Gorka T., “Method for Estimating Capacity and Predicting Remaining Useful Life of Lithium-ion
Battery,” Applied Energy, v126, p182–189, 2014.

13
Model-Based Prognostics: How Does it Work?

Current Capacity fade model

cycle e.g., Q(θ,t) = k1(1 – e-t/t0) + 1
Capacity θ = {k1, t0}
Normalized capacity (%)

100 estimates
Capacity estimates
90 �
𝑄𝑄(1:t)

80
End-of-life limit

70
0 100 200 300 400 500
Cycle number

Hu C., Jain G., Tamirisa P., and Gorka T., “Method for Estimating Capacity and Predicting Remaining Useful Life of Lithium-ion
Battery,” Applied Energy, v126, p182–189, 2014.

14
Model-Based Prognostics: How Does it Work?

Current Capacity fade model

cycle e.g., Q(θ,t) = k1(1 – e-t/t0) + 1
Capacity θ = {k1, t0}
Normalized capacity (%)

100 estimates
Capacity estimates Parameter
90 � �
estimate 𝛉𝛉(𝑡𝑡)
𝑄𝑄(1:t) Infer
parameters
80
End-of-life limit

70
0 100 200 300 400 500
Cycle number

Hu C., Jain G., Tamirisa P., and Gorka T., “Method for Estimating Capacity and Predicting Remaining Useful Life of Lithium-ion
Battery,” Applied Energy, v126, p182–189, 2014.

15
Model-Based Prognostics: How Does it Work?

Current Capacity fade model

cycle e.g., Q(θ,t) = k1(1 – e-t/t0) + 1
Capacity θ = {k1, t0}
Normalized capacity (%)

100 estimates
Capacity estimates Parameter
90 � �
estimate 𝛉𝛉(𝑡𝑡)
𝑄𝑄(1:t) Infer
parameters
80
End-of-life limit
�
𝑄𝑄(t+1:t EOL) Forecast
70 capacity
0 100 200 300 400 500
Cycle number

Hu C., Jain G., Tamirisa P., and Gorka T., “Method for Estimating Capacity and Predicting Remaining Useful Life of Lithium-ion
Battery,” Applied Energy, v126, p182–189, 2014.

16
Model-Based Prognostics: How Does it Work?

Current Capacity fade model

cycle e.g., Q(θ,t) = k1(1 – e-t/t0) + 1
Capacity θ = {k1, t0}
Normalized capacity (%)

Hu C., Jain G., Tamirisa P., and Gorka T., “Method for Estimating Capacity and Predicting Remaining Useful Life of Lithium-ion
Battery,” Applied Energy, v126, p182–189, 2014.

17
Model-Based Prognostics: How Does it Work?

Current True end of life Capacity fade model

cycle (unknown) e.g., Q(θ,t) = k1(1 – e-t/t0) + 1
Capacity θ = {k1, t0}
Normalized capacity (%)

Hu C., Jain G., Tamirisa P., and Gorka T., “Method for Estimating Capacity and Predicting Remaining Useful Life of Lithium-ion
Battery,” Applied Energy, v126, p182–189, 2014.

18
Applying Model-Based Prognostics

True Capacity
Data Available