ME591

Random Data
Analysis and Processing

Lecture 1

Introduction

ME591 Random Data  Prof. Yong-Hwa Park, KAIST

 ME591 Random Data: Analysis and Processing
Fall 2025

Lecture by Prof. Yong-Hwa Park

at yhpark@[Link]
in Rm.2202, Mechanical Eng. Bldg.
combination with KLMS
(The lecture material rely on the NOVIC+ professors’ experiences and lecture materials accumulated for
past 30+ years, being prepared by using various texts and sources emphasizing lecturer’s random data
analysis and processing experiences. )

ME591 Random Data 2  Prof. Yong-Hwa Park, KAIST

 Prof. Yong-Hwa Park
(1987~1999) BS, MS, Ph.D at Mechanical Eng., KAIST
Structural Vibration and Optimization
(2000~2003) Post Doc. of Univ. of Colorado at Boulder
Microelectromechanical System Design
(2003~2016) Research Master of SAIT, Samsung Electronics
미래 센서 연구 개발 (3D Lidar, Health Sensors,…)
Published 70+ US Patents

• NOVIC+ (Noise and Vibration Control Plus) 센터장 (2023~) Research Topics
• KAIST 기계공학과 산학협력위원장 (2017~2025)
• Mechanical Vibration, Condition Monitoring (PHM)
• KAIST 로봇학제 참여교수 (2022~) • Auditory Intelligence, Sound Event Detection
• KAIST 안보융합기술 대학원 겸임교수 (2023~) • Health Sensors with Human Vibration and Sound
• 대한기계학회, 소음진동공학회 부회장 (2023~) • 3D Measurements (Lidar) and AI-base Recognition
• KAIST Tennis Championship (1996, 2000)

ME591 Random Data 3  Prof. Yong-Hwa Park, KAIST

 Recent Achievement: Cough Detection Camera (기침인식 카메라)
[Link]
[Link]

ME591 Random Data 4  Prof. Yong-Hwa Park, KAIST

 Time Series / Spectral Analysis of Random Data
A practical example: time-frequency analysis (Spectrogram)

Time Series / Spectral Analysis Deep Neural Net

of experimental data (random) for classification

Time-domain Frequency-domain
Feature (input) Class (output)
Amplitude

Frequency

Frequency

Time (sec) Time (sec) Time (sec)

Frequency
Frequency

Frequency

Time (sec) Time (sec) Time (sec)

ME591 Random Data 5  Prof. Yong-Hwa Park, KAIST

 System Identification, Source Localization :
Source location (, r) ; Parameter identification (Pin); Transfer Function: H()= Pout ()/Pin()
Sound source
Input pressure, Pin
(Input-output Relation)
System Identification

 Output pressures, Pout

Microphone measurements
Source Localization

ME591 Random Data 6  Prof. Yong-Hwa Park, KAIST

 ME591 RANDOM DATA: ANALYSIS AND PROCESSING
Course Description

• This course masters both fundamentals and applications of signal processing techniques in mechanical systems
primarily based on the engineering knowledge of probability (ensemble), correlation (time), and spectral
analysis (frequency domain). It starts with the introduction of basic topics such as probability distribution,
ensemble averages, stochastic process, normal distribution, correlation, spectral density function, frequency
response function, and white noise for students with basic background on statistical analysis. Then, digital
signal processing techniques such as data acquisition, pre-processing, data qualification, discrete Fourier
transform, digital filter, decimation, zoom transform, windowing, overlapping and averaging on a random
process of interest are treated. Then, random vibration analysis with multi-input/multi-output systems is
extensively studied and applied to linear system identification, energy source identification, nonlinear system
analysis and so on. Finally, the essences of state-of-the-art techniques such as time-frequency analysis,
wavelets, cepstrum analysis and Hilbert transform are briefly discussed mainly focusing on the field
applications.

• This course is strongly recommended for those who expect to analyze and process the random data either taken
from experiments or computer simulations, and, in particular, those who are interested in identification of
input-output relations of the systems, linear or nonlinear, and source identification of interest, and
fundamentals of advanced state-of-the art signal processing as well.

ME591 Random Data 7  Prof. Yong-Hwa Park, KAIST

 ME 591 RANDOM DATA: ANALYSIS AND PROCESSING
Lecturer Professor Yong-Hwa Park, Human Lab, NOVIC+, Rm.4107 (Ext.3235; yhpark@[Link])
Office hour Friday morning 9:00-12:00 upon reservation, or CLASSUM at anytime.

Lectures Mon & Wed. 2:30-4:00 p.m.; Rm. 2202, ME, KAIST, Combination with pre-recorded
lectures via KLMS. Attendance of lectures will be checked by KLMS progress.
Class Format Education 4.0 Q: Pre-recorded lecture (50%) + Student participation in the class
(50%, student-led interactive Q&A and discussions)
Text J. S. Bendat and A. G. Piersol, Random Data: Analysis and Measurement
Procedures, 4th ed., John Wiley & Sons, 2010 :
yes24: [Link]
References J. S. Bendat and A. G. Piersol, Engineering Applications of Correlation and Spectral Analysis, 2nd
ed., John Wiley & Sons, 1993; J. S. Bendat, Nonlinear System Analysis and Identification from
Random Data, John Wiley & Sons, 1990; J. S. Bendat, Nonlinear Systems Techniques and
Applications, John Wiley & Sons, 1998; D. E. Newland, An Introduction to Random Vibrations,
Spectral & Wavelet Analysis, 3rd ed., Longman Scientific & Technical, 1993
Sal Khan, selected video lectures on Probability and Statistics from [Link].
Also available from [Link].
ME591 Random Data 8  Prof. Yong-Hwa Park, KAIST
 ME 591 RANDOM DATA: ANALYSIS AND PROCESSING

TA’s:
Head T.A. Junki Hong (N7) 4104, Ext.3075, wnsrlwns@[Link], 010-2065-8734
T.A. (HW#1) Jonghyeon Park (N7) 5212, Ext.3241, op585@[Link], 010-6439-3896
T.A. (HW#2) Sunho Park (N7) 7120, Ext.3285, psho2512@[Link], 010-9354-2520
T.A. (HW#3) Sol Han (N9) 4170, dream4future@[Link], 010-9016-3209
T.A. (HW#4) Minho Chae (N7) 2216, [Link]@[Link], 010-8280-7176
T.A. (HW#5) Jaejung Park (N7) 3105, parkfrank01@[Link], 010-5327-2484
T.A. Hyunsik Oh (N9) 4150, Ext.8901, [Link]@[Link],010-5254-9710
T.A. Seokjun An (N7) 4114, Ext.5235, asj0379@[Link],010-4499-0212
T.A. Yoonseop Lim (N7) 4114, Ext 5235, bu2050@[Link], 010-6756-2050
T.A. Junyo Boo (N7) 5212, Ext.3241, [Link]@[Link], 010-8869-9587
T.A. Jeong Hyun Lee (N9) 4150, Ext.8901, joshualee@[Link], 010-2999-7909

ME591 Random Data 9  Prof. Yong-Hwa Park, KAIST

 ME 591 RANDOM DATA: ANALYSIS AND PROCESSING
Homework Total 5 homework sets including matlab experiments will be given. The solutions and solution
video of homeworks will be uploaded on the due date. Late homeworks will not be accepted.

Tutorial Sessions by TAs: Total 5 in-person tutorials for 5 HW given by TAs:

1. TAs will offer office hour to give individual tutorial before the homework due date.
2. The video of solution procedure for each homework will be uploaded to KLMS on the due date
of each homework. Detailed tutoring method, contents, and date will be announced in KLMS.

Grading Homework: 30 %
Mid-term exam: 30 %
Final exam: 30 %
Class Participation + Attendance 10 % or more

1 Absence -0.5%, Late -0.2% (attendance will be checked by KLMS progress at 4 pm of the
lecture date). Others including questions, comments, any active participations in the class and
CALASSUM can be result in additional credit of 10% or more.

ME591 Random Data 10  Prof. Yong-Hwa Park, KAIST

 PART I : Theory
Probability,
Time,
Frequency
Analyses

PART II :
Data Acquisition,
Processing

Lecture Schedule PART III :

Applications,
System Identification,
Recent Signal
Processing

ME591 Random Data 11  Prof. Yong-Hwa Park, KAIST

 “Correlation” between
Lecture and Textbook

ME591 Random Data 12  Prof. Yong-Hwa Park, KAIST

 CLASSUM: open platform for interactive discussions!
CLASSUM 가입 및 수업 연동

1. KLMS 강의실 좌측 메뉴 중 CLASSUM 클릭 2. CLASSUM 이동 클릭 3. CLASSUM 회원가입 후 로그인

(kaist 메일이 아니어도 상관 없음)

ME591 Random Data 13  Prof. Yong-Hwa Park, KAIST

 CLASSUM: open platform for interactive discussions!
커뮤니티 기본 모습

ME591 Random Data 14  Prof. Yong-Hwa Park, KAIST

 KAKAO Link for each group
Group A: [Link]

오픈채팅방 Team Tutoring→

TA: 홍준기
Group B: [Link]
Group C: [Link]
Group D: [Link]
Group E: [Link]
Group F: [Link]

Get advices
TA: 박종현
from TA!

TA: 박선호

 교수님의 공지,  조별로 조원들

강의 보충 설명, 돌아가며 1개/1주
교수님과 직접 질의응답 질문 작성하고,
Link:
TA: Sol Han CLASSUM에
[Link] 질문 글 올리기

TA: 채민호

TA: 박재정

ME591 Random Data 15  Prof. Yong-Hwa Park, KAIST

 Korea Space Launch Vehicle II “the Nuri”
The 7th nation to launch a
space vehicle with satellite
over 1 ton.

Development history
Phase I :2010.3 ~ 2015.7
7 ton fluid engine
Phase II :2015.8 ~ 2019.2
75 ton fluid engine
Phase III :2018.4 ~ 2022.5
Engine clustering,
3-stage rocket structure

Launch history
1st : 2021. 10. 21 partial success
2nd : 2022. 6. 21 success
3rd : 2023. 5.25 success
4th : 2024.3 …
5th : 2026. 6
6th : 2027.9
ME591 Random Data 16  Prof. Yong-Hwa Park, KAIST
 Korea Space Launch Vehicle II “the Nuri”

Specifications
75 ton engine for the 1st, 2nd stage, 7 ton engine for the 3rd stage

ME591 Random Data 17  Prof. Yong-Hwa Park, KAIST

 Final firing test of Nuri's 75 ton engine (2021. 2. 25)

ME591 Random Data 18  Prof. Yong-Hwa Park, KAIST

 Example: Propellant rocket motor, firing test

load cell accelerometer

thrust
wall

• Mass : M (fuel full), 0.25 M (empty)

• Natural frequency : fn (full), 2fn Hz (empty)

ME591 Random Data 19  Prof. Yong-Hwa Park, KAIST

 Thrust Data (output of load cell) –
N
5
x 10
4 “Deterministic or Random”
transient fn Hz (full) steady state 2fn Hz (empty) transient
0

-5

-10
04 10 20 30 4
40 50 Time (s)
N x 10 N x 10
6 5

4 0

2
Ignition -5 tail-off
0
-10
N 0 4
0.2 0.4 0.6 0.8 1 52.6 52.8 53 53.2 53.4 53.6 Time (s)
x 10 Time (s)
4.8
steady state (random)
4.7

4.6

4.5
39 39.1 39.2 39.3 39.4 39.5 39.6 39.7 39.8 39.9 40 Time (s)

ME591 Random Data 20  Prof. Yong-Hwa Park, KAIST

 How to characterize random process x(t ) ?
p ( x(t1 )) p( x(t2 )) p( x(t3 ))
x1 (t )
t

x2 (t )
t

{xk (t )} x3 (t )
t
k = sample number
p ( x(t1 ), x(t2 )) p ( x(t2 ), x(t3 ))
…

p ( x(t1 ), x(t3 ))
xk (t )
t

t1 t2 t3

ME591 Random Data 21  Prof. Yong-Hwa Park, KAIST

 p ( x(t1 )) : Probability density function of random variable x(t ) at t = t1
p ( x(t1 ), x(t2 )) : Joint probability density function of random variable x(t ) at t = t1 and t = t2

p ( x(t1 )) p( x(t2 )) p( x(t3 ))

x1 (t )
t

x2 (t )
t

x3 (t )
t

p ( x(t1 ), x(t2 )) p ( x(t2 ), x(t3 ))

…

p ( x(t1 ), x(t3 ))
xk (t )
t

t1 t2 t3

ME591 Random Data 22  Prof. Yong-Hwa Park, KAIST

 2DOF
Swing Robot

[Link]

ME591 Random Data 23  Prof. Yong-Hwa Park, KAIST

 Robot Test Results ([Link])
Mass
center
alignment

Center

Left

Right

ME591 Random Data 24  Prof. Yong-Hwa Park, KAIST

 Who is an Unbiased Shooter?
Who is a Robust Shooter?
Whom you will choose in war time?
(The answer is not unique but multiple)

A B
In-situ Questions of lecture 1: send your answer/reason to CLASSUM! Any answer will be credited.
ME591 Random Data 25  Prof. Yong-Hwa Park, KAIST