TI 5003 METODOLOGI DAN PEMODELAN SISTEM

D INAMIKA SISTEM
LUCIA DIAWATI
Lab. Perencanaan & Optimasi Sistem Industri
Departemen Teknik Industri
Fakultas Teknologi Industri

INSTITUT TEKNOLOGI BANDUNG

 1. PENDAHULUAN: MATERI BAHASAN

A. Systems Thinking

B. System Dynamics

C. Continuous Deterministic Simulation

 VENSIM

D. Discrete Stochastic Simulation

 PROMODEL

MPS – LD SEM 1-2017/2018 2

 1. PENDAHULUAN: SASARAN PERKULIAHAN

1. Mahasiswa dapat mengaplikasikan metodologi Dinamika Sistem untuk memvisualisasikan sistem

relevan dalam bentuk struktur dan aspek yang menciptakan dinamika dan mengendalikan
performansinya.

2. Mengidentifikasi struktur sistem dan pola perilakunya.

3. Menterjemahkan struktur sistem relevan ke dalam CLD (causal loo diagram) dan formulasi model.

4. Menganalisis dan perilaku dan kinerja sistem.

5. Memformulasikan kebijakan organisasi untuk memperbaiki kinerja sistem.

MPS – LD SEM 1-2017/2018 3

 1. PENDAHULUAN: SASARAN SPESIFIK

Sasaran spesifik:

1. Mentransfer CLD menjadi SFD (Stock & Flow Diagrams)

2. Mengimplementasikan SFD dalam VENSIM

3. Mempelajari parameterisasi model VENSIM

4. Memvalidasi model VENSIM

5. Melakukan “what-if experiments”

6. Melakukan studi sensitivitas

MPS – LD SEM 1-2017/2018 4

 2. WHAT IS SYSTEM DYNAMICS

 Pemodelan simulasi komputer untuk mempelajari dan mengelola sistem

kompleks dan dinamis (complex feedback systems), seperti sistem bisnis dan
sistem sosial lainnya.
 Suatu cara untuk menggambarkan sistem sebagai stok (stock) dan aliran (flows) antar
stok.
 Stok: variabel yang mengakumulasikan dampak dari variabel lain.
 Rate: variabel yang mengendalikan aliran material masuk dan keluar dari stok.
 Auxiliary: variabel yang memodifikasi informasi setelah dilewatkan dari stok ke rate.

MPS – LD SEM 1-2017/2018 5

 2. WHAT IS SYSTEM DYNAMICS

 Sistem:
• Secara umum, kumpulan elemen yang saling berinteraksi yang berfungsi bersama untuk
mencapai tujuan tertentu.
• Dalam DS, feedback (umpan-balik) merupakan deskriptor pembeda.
 Karakteristik masalah dinamis:
• Mengandung kuantitas yang berubah sepanjang waktu
• Variabilitas dapat dideskripsikan secara kausal.
• Dampak kausal yang penting tercakup di dalam suatu sistem tertutup dari loop umpan
balik.

MPS – LD SEM 1-2017/2018 6

 3. METODOLOGI DINAMIKA SISTEM
Metodologi Dinamika Sistem (System Dynamics):
 Langkah-langkah sederhana:
A A. Kumpulkan informasi terkait
dengan masalah yang akan
A ditangani
H B. Identifikasi variabel-variabel
• Formulasi Masalah terkait
• Identifikasi Sistem
Relevan C. Gambarkan hubungan antar
variabel dengan CLD
G B–C D. Terjemahkan CLD ke dalam SFD
E. Masukkan (SFD dan formulasi
model) ke dalam VENSIM
• Model Konseptual F. Lakukan validasi dan analisis
D–E sensitivitas
F G. Desain skenario kebijakan dan
lakukan eksperiman “WHAT-IF”
• Pengujian Model
H. Tuliskan rekomendasi
MPS – LD SEM 1-2017/2018 7
 3. METODOLOGI DINAMIKA SISTEM

SD dalam Siklus Pengambilan Keputusan

Business System
• Symptom masalah
• Faktor2 terkait dengan masalah

Implement Strategy Data collection

• Eksperimen kebijakan (intervensi)
Simulate

Business Compare/
Model Model
• Model Konseptual Evaluate
• Model Matematis
• Pengujian Model

Analyze
• Formulasi Masalah
MPS – LD SEM 1-2017/2018 Analyze • Idetifikasi Sistem Relevan 8
 3. METODOLOGI DINAMIKA SISTEM
Sejarah Perkembangan Dinamika Sistem:

MPS – LD SEM 1-2017/2018 9

 3. METODOLOGI DINAMIKA SISTEM
Disiplin Dasar Dinamika Sistem
Manajemen Tradisional:
 Manajemen yang lazim digunakan dalam dunia nyata oleh para praktisi manajerial yang pada
dasarnya dibentuk oleh pengalaman dan penilaian dari manajer.
 Kekuatan utama dari manajemen tradisional: banyaknya informasi kualitatif yang diperoleh
dari pengamatan dan pengalaman langsung oleh manajer.
 Kelemahan: informasi tidak dapat digunakan secara efektif tanpa adanya suatu prinsip dalam
pemilihan informasi yang relevan dan strukturisasi informasi.
Cybernetics: ilmu mengenai komunikasi dan kontrol yang didasari oleh teori umpan balik. Peran
cybernetics: membantu manajer menyaring informasi yang ada sesuai dengan permasalahan yang
dihadapi, kemudian menghubungkan elemen-elemen informasi itu untuk menemukan hubungan
sebab akibat yang ada serta membangun struktur umpan balik sistem.
Simulasi Komputer: a way “to reproduce the conditions of a situation, as by means of a model, for study or
testing or training, etc.”
MPS – LD SEM 1-2017/2018 10
 3. METODOLOGI DINAMIKA SISTEM

Sejarah Perkembangan

 Cybernetics (Wiener, 1948): the study of how biological, engineering, social, and
economic systems are controlled and regulated.
 Industrial Dynamics (Forrester, 1961): applied principles of cybernetics to
industrial systems.
 System Dynamics: Forrester’s work has been broadened to include other social
and economic systems.
 Relying on computer, System Dynamics provides a framework in which to apply
the idea of systems theory to social and economic problems.

MPS – LD SEM 1-2017/2018 11

 3. METODOLOGI DINAMIKA SISTEM

Ciri Khas Dinamika Sistem:

Feedback system:
1. Sistem yang dijadikan model harus sistem tertutup, meski sistem tidak sungguh-sungguh
tertutup karena feedback loop tidak dapat melintasi batasan sistem. Sistem dapat
dipertimbangkan sebagai sistem tertutup.
2. Ada dua umpan balik dalam sistem, yakni positif dan negatif.
3. Adanya variabel STATE dan RATE (STATE: kondisi atau akumulasi dari sistem pada
waktu tertentu; RATE: aliran yang mengatur kuantitas dalam state).
4. RATE mengontrol kebijakan, artinya perilaku sistem hanya dikontrol oleh RATE.

MPS – LD SEM 1-2017/2018 12

 3. METODOLOGI DINAMIKA SISTEM
Ilustrasi Stok dan Rate dalam Model Dinamika Sistem:

Rate (Inflow)
Metafora Hidrolika:
Flow

Stok

Rate (Outflow)

Stock-Flow Diagram:

t
Persamaan Integral: Stock t = Inflow s −Outflow(s) ds+Stock (t0 )
t0
d(Stock)
Persamaan Diferensial: =Net Change Rate=Inflow t −Outflow(t)
dt
MPS – LD SEM 1-2017/2018 13
 3. METODOLOGI DINAMIKA SISTEM

Manfaat Dinamika Sistem (dan System Thinking):

 Tingkat belajar yang lebih dalam
 Jauh lebih baik dari deskripsi verbal

 Representasi struktur masalah atau proses yang jelas.

 Suatu cara untuk memahami impllikasi perilaku dari
struktur dan data sistem.
 Suatu perangkat praktis (hands on) untuk melakukan
eksperiman “WHAT IF”.

MPS – LD SEM 1-2017/2018 14

 4. CAUSAL LOOP DIAGRAM (CLD)
Causal Loop Diagram (CLD):
Alat bantu berupa diagram yang menggambarkan hubungan antar variabel yang merepresentasikan sistem relevan.

Contoh CLD:
Desired Water
Level Faucet
Position

Perceived
Gap
Water
Flow

Current
Water
Level

MPS – LD SEM 1-2017/2018 15

 4. CAUSAL LOOP DIAGRAM (CLD)
Polaritas (pola hubungan) hubungan kausal dalam CLD:
1. Hubungan Kausal Positif: hubungan dimana
peningkatan (penurunan) nilai variabel penyebab
akan mengakibatkan peningkatan (penurunan)
pada nilai variabel yang terkena dampak. Arah
perubahan dari nilai kedua variabel yang
berhubungan tersebut searah.
Contoh:
+ Tingkat
Tingkat Populasi
Kelahiran Penduduk
 Tingkat kelahiran meningkat, maka populasi penduduk
juga meningkat; sebaliknya
 Tingkat kelahiran menurun, maka populasi penduduk
juga menurun.
MPS – LD SEM 1-2017/2018
2. Hubungan Kausal Negatif: hubungan dimana
peningkatan (penurunan) nilai variabel penyebab
akan menyemengakibatkan babkan penurunan
(peningkatan) pada nilai variabel yang terkena
dampak. Arah perubahan dari nilai kedua variabel
yang berhubungan tersebut berlawanan arah.
Contoh:
 Volume
Inventori Pemesanan
Barang
 Tingkat inventori tinggi, maka volume
pemesanan barang rendah; sebaliknya
 Tingkat inventori rendah, maka volume
pemesanan barang tinggi.
16
 4. CAUSAL LOOP DIAGRAM (CLD)
Penentuan polaritas loop (CLD):
1. Signing (penetapan tanda hubungan antar variabel): tambahkan tanda ‘+’
atau ‘–’ pada setiap anak panah untuk menunjukkan polaritas hubungan
antar variabel yang dihubungkan dengan tanda panah tersebut.
2. Penentuan polaritas loop:
 Positive feedback loops
 Memiliki hubungan kausal dengan tanda ‘–’ berjumlah genap.
 Menciptakan perilaku growth, amplify, deviation, dan reinforce.
 Notasi: +

 Negative feedback loops

 Memiliki hubungan kausal dengan tanda ‘–’ berjumlah ganjil.
 Cenderung menghasilkan perilaku “stable”, “balance”, “equilibrium”
and “goal-seeking” behavior over time
 Notation: 
MPS – LD SEM 1-2017/2018 17
 5. NOTASI DALAM DINAMIKA SISTEM
Input / Parameter / Lookup:
Stok/Level: STOK
I1
Tidak ada aliran
(anah panah) yang
Rate: I2
masuk
(mempengaruhi)
input
I3
Auxilliary:

I1 O1 Output:
O1
I2 Auxilliary: O2 Tidak ada aliran
(anah panah) yang
O2
keluar dari output
I3 O3
O3
MPS – LD SEM 1-2017/2018 18
 5. NOTASI DALAM DINAMIKA SISTEM

Tipe hubungan kausal:

Ada 2 tipe hubungan kausal dalam model kausal,
Notasi aliran (flow) dan informasi:
yaitu:
1. Informasi: mengalir dari stok dan Informasi:
parameter/input menuju rate dimana A B
informasi akan digunakan untuk
mengendalikan aliran.
2. Aliran atau flow (material): aliran yang selalu Flow (of material):
mengalir dari rate ke stok dalam diagram
kausal.
X

MPS – LD SEM 1-2017/2018 19

 5. PERILAKU DASAR SISTEM

a. Exponential Growth c. Oscillation e. S-shaped Growth with Overshoot

1: Level 1: System Level 1 1: System Level
1: 1000.00 1: 350.00 1: 1400.00

1
1
1

1: 200.00 1 1: 700.00
1: 500.00

1 1
1

1 1

1
1: 50.00 1: 0.00 1
1: 0.00 0.00 10.00 20.00 30.00 40.00 0.00 7.50 15.00 22.50 30.00
0.00 3.00 6.00 9.00 12.00
Graph 1 (Untitled) Time 9:29 PM Tue, Nov 14, 2000 Graph 1 (Untitled) Time 10:18 PM Tue, Nov 14, 2000
Graph 1 (Untitled) Time 7:00 PM Tue, Nov 14, 2000

b. Goal Seeking d. S-shaped Growth f. Overshoot and Collapse

1: System Level 1 1: System Level 1: System Level 2: Carrying Capacity
1: 200.00 1 1: 1000.00 1: 55.00 2
1 1 2: 1000.00

1 1

1: 150.00 1: 500.00 1: 30.00

2: 500.00
1
2

1 1 1
1
1 1: 5.00
1: 100.00 1: 0.00 2: 0.00 2 2
0.00 3.00 6.00 9.00 12.00 0.00 5.00 10.00 15.00 20.00 0.00 1.88 3.75 5.62 7.50
Graph 1 (Untitled) Time 9:18 PM Tue, Nov 14, 2000 Graph 1 (Untitled) Time 10:05 PM Tue, Nov 14, 2000 Graph 1 (Untitled) Time 10:29 PM Tue, Nov 14, 2000

MPS – LD SEM 1-2017/2018 20

 5. PERILAKU DASAR SISTEM
a. Exponential Growth
Perilaku yang dihasilkan dari loop kausal dengan polaritas POSITIF. Loop ini membuat growth, amplify
deviations, and reinforce change.
Struktur sistem:

SFD:
System Level

Inflow

Exponential Family of Time Paths

Net increase rate
MPS – LD SEM 1-2017/2018 21
 5. PERILAKU DASAR SISTEM
a. Exponential Growth
Contoh: A causal loop diagram
about road congestion (Sterman
2000)

Sterman (2000). Business Dynamics:

Systems Thinking and Modeling for a
Complex World, McGraw Hill.

MPS – LD SEM 1-2017/2018 22

 5. PERILAKU DASAR SISTEM
b. Goal Seeking
+
Perilaku yang dihasilkan dari loop kausal dengan polaritas Target
NEGATIF. Loop ini mencari balance, equilibrium, and stasis. Inventori Inventori + Pesanan
+
 +
Struktur
sistem: + 
GAP

System Level
Corrective Action
SFD:

Discrepancy

Rate of change Goal

MPS – LD SEM 1-2017/2018

Goal Seeking Family of Time Paths 23
 5. PERILAKU DASAR SISTEM
b. Goal Seeking
Contoh: Coffee 100 0
90
cooling loop 80
-2

-4

Colling Rata (D/dT)

70

Coffee Temp (D)

60 -6

50 -8
40
-10
30
-12
20
10 -14

0 -16
Coffee cooling 0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 5 5,5 6 6,5 7 7,5 8 8,5 9 9,5 10
dT
stock flow diagram COFFEE (Coffee Temp) ROOM (Room Temp) CHNG (Cooling Rate)

MPS – LD SEM 1-2017/2018 24

 5. PERILAKU DASAR SISTEM
Change in
b. Goal Seeking TIME
COFFEE (Coffee ROOM (Room
ROOM-COFFEE
CHNG (Cooling
Coffee Temp per
Temp) Temp) Rate)
Komputasi temperatur kopi 0 0 0 0
0,5 Min
0
(Min) C C C C/Min C
0 90,0 20 -70,0 -14,0 -7,0
0,5 83,0 20 -63,0 -12,6 -6,3
1 76,7 20 -56,7 -11,3 -5,7
1,5 71,0 20 -51,0 -10,2 -5,1
2 65,9 20 -45,9 -9,2 -4,6
2,5 61,3 20 -41,3 -8,3 -4,1
3 57,2 20 -37,2 -7,4 -3,7
3,5 53,5 20 -33,5 -6,7 -3,3
4 50,1 20 -30,1 -6,0 -3,0
4,5 47,1 20 -27,1 -5,4 -2,7
5 44,4 20 -24,4 -4,9 -2,4
5,5 42,0 20 -22,0 -4,4 -2,2
6 39,8 20 -19,8 -4,0 -2,0
6,5 37,8 20 -17,8 -3,6 -1,8
7 36,0 20 -16,0 -3,2 -1,6
7,5 34,4 20 -14,4 -2,9 -1,4
8 33,0 20 -13,0 -2,6 -1,3
8,5 31,7 20 -11,7 -2,3 -1,2
9 30,5 20 -10,5 -2,1 -1,1
9,5 29,5 20 -9,5 -1,9 -0,9
MPS – LD SEM 1-2017/2018 10 28,5 20 -8,5 -1,7 -0,9 25
 5. PERILAKU DASAR SISTEM
c. Oscillation Perilaku yang disebabkan oleh adanya keterlambatan waktu (time delay) dalam merespon status sistem.
Time delays menyebabkan keadaan sistem secara konstan melebihi sasaran atau kondisi ekuilibrium,
atau sebaliknya lebih rendah, dan seterusnya  status sistem berosilasi di sekitar sasaran.
Struktur
Sistem:

SFD: Corrective Action

System Level

Perception

Discrepancy Perception Delay

MPS – LDRate
SEM 1-2017/2018 Oscillation Family of Time Paths 26
of change Goal
 5. PERILAKU DASAR SISTEM
d. S-shaped Growth Perilaku yang disebabkan adanya keterbatasan sumberdaya yang diperlukan untuk pertumbuhan.
Struktur Sistem:

1: System Level
1: 1000.00
1

1: 500.00

System Level
SFD: Net Increase Rate

1
1: 0.00
0.00 5.00 10.00 15.00 20.00

Graph 1 (Untitled) Time 10:05 PM Tue, Nov 14, 2000

Resource Adequacy
S-shaped Growth of Time Paths
Normal Growth Rate Carrying Capacity

MPS – LD SEM 1-2017/2018 27

 5. PERILAKU DASAR SISTEM
e. S-shaped Growth with Overshoot Time delays dalam loop kausal dengan polaritas positif menyebabkan kemungkinan status
sistem akan overshoot dan berosilasi di sekitas kapasitas sumber daya yang diperlukan.
Struktur
Sistem:

1: System Level
1: 1400.00

1
1

System Level
SFD: Net Increase Rate
1: 700.00

1: 0.00 1
Resource Adequacy 0.00 7.50 15.00 22.50 30.00
Graph 1 (Untitled) Time 10:18 PM Tue, Nov 14, 2000
Delayed Resource Effect

Normal Growth Rate

S-shaped Growth and Overshoot of Time Paths
Carrying Capacity

MPS – LD SEM 1-2017/2018 28

 5. PERILAKU DASAR SISTEM
f. Overshoot and Collapse Perilaku ini disebabkan oleh daya dukung sumber daya yang diperlukan untuk pertumbuhan
menurun dan habis.
Struktur Sistem:

SFD:
1: System Level 2: Carrying Capacity
Consumption rate
1: 55.00 2
2: 1000.00

1
Carrying Capacity

Erosion of Carrying Capacity

1: 30.00
2: 500.00
1
2
System Level
Net Increase Rate

1
1
1: 5.00
2: 0.00 2 2
0.00 1.88 3.75 5.62 7.50

Graph 1 (Untitled) Time 10:29 PM Tue, Nov 14, 2000

Normal Growth Rate Resource Adequacy

Overshoot and Collapse of Time Paths
MPS – LD SEM 1-2017/2018 29
 5. PERILAKU DASAR SISTEM
Perhatikan masalah Crime Spiral berikut.
a. Fenomena apa yang diperlihatkan oleh spiral tersebut? Berikan gambaran numerik dari fenomena tersebut.
b. Apakah fenomena tersebut telah sesuai dengan keadaan riel?
c. Bagaimana menangani masalah Crime Spiral tersebut?

MPS – LD SEM 1-2017/2018 30

 5. PERILAKU DASAR SISTEM

Tugas
Diskusikan dan atasi
masalah berikut.
a. Fenomena apa yang
diperlihatkan oleh
diagram loop ini?
b. Apakah fenomena
tersebut telah sesuai
dengan keadaan riel?
c. Berdasarkan loop
tersebut, apa saran
saudara untuk
mengurangi volume lalu
lintas? Berikan gambaran
dampak saran saudara
secara numerik.

MPS – LD SEM 1-2017/2018 31

 6. DELAY
Building block dalam System Thinking:
1. Reinforcing (amplifying) feedback (loop kausal positif)
2. Balancing (stabilizing) feedback (loop kausal negative)
3. Delay

DELAY

MPS – LD SEM 1-2017/2018 32

 6. DELAY
Delay:
 Perbedaan waktu antara mulainya suatu aliran input dan mulainya aliran output yang dihasilkan oleh
suatu proses konversi.
 Pada dasarnya terjadi pada setiap aliran.
 Merupakan faktor penting yang menciptakan dinamika perilaku sistem.

Dua Karakteristik Delay:

 Panjang waktu  menentukan dampak "steady-state" delay.
"Steady state":
 Tingkat inflow, outflow dan level konstan;
 Tingkat inflow = outflow;
 Jumlah dalam transit = (Flow rate) x rata-rata Delay
 "Transient response":
 Menggambarkan bagaimana bentuk waktu dari outflow berhubungan dengan bentuk waktu dari
inflow jika bentuk inflow berubah menurut waktu.

MPS – LD SEM 1-2017/2018 33

 6. DELAY
Delay Eksponensial:
 Digunakan karena bentuknya yang sederhana.
 Delay eksponensial orde-1:

 Persamaan outflow:

MPS – LD SEM 1-2017/2018 34

 6. DELAY
 Delay eksponensial orde-3:

MPS – LD SEM 1-2017/2018 35

 6. DELAY

 Delay eksponensial orde-3:

Atau:

MPS – LD SEM 1-2017/2018 36

 6. DELAY
 Response Waktu terhadap Delay Eksponensial
a. Delay eksponensial Orde – 1:

Rate Level

b. Delay eksponensial Orde – 2:

Rate Level

MPS – LD SEM 1-2017/2018 37

 6. DELAY
 Response Waktu terhadap Delay Eksponensial
c. Delay eksponensial Orde – 3:

Rate Level

d. Delay diskrit atau pipeline

Rate Level

MPS – LD SEM 1-2017/2018 38

 6. DELAY

 Response Delay Eksponensial terhadap "unit impulse"

MPS – LD SEM 1-2017/2018 39

 6. DELAY

 Response Delay Eksponensial terhadap "unit-step rate"

MPS – LD SEM 1-2017/2018 40

 6. DELAY

Interval Solusi dt:

 Interval solusi dt hanya ada pada persamaan level.

 Interval solusi menjadi pengali atas variabel rate dan memainkan
peranan penting dalam menentukan akumulasi (integrasi) dari fungsi
yang membentuk persamaan level.
 Nilai dt dipilih untuk memastikan stabilitas dalam proses komputasi.
 Aturan umum:

0,2 * Delay terendah  dt  0,5 * Delay (Tingkat I)

MPS – LD SEM 1-2017/2018 41

 7. PENGUJIAN

a. Analisis input
b. Verifikasi dan validasi model
 Termasuk analisis hubungan variabel (faktor)  banyak model tidak
standar / baku.
c. Analisis Output (kinerja / perilaku)
d. Perancangan Kebijakan

MPS – LD SEM 1-2017/2018 42

 7. PENGUJIAN
a. Analisis input
Analisis statistic input (data) nnumerik
a) Analisis karakteristik data: rata-rata (mean), median, mode, deviasi standar,
koefisien variasi, skewness, kurtosis, range.
b) Analisis:
 Analisis independensi: scatter plot, autocorrelation plot, runs test.
 Analisis distribusi data (goodness of fit): distribusi kontinyu (Uniform,
Exponential, Erlang, Gamma, Weibull, Normal, Lognormal, Beta, Triangular),
distribusi diskrit (Bernoulli, Binomial, Negative Binomial, Geometric, Poisson).

MPS – LD SEM 1-2017/2018 43

 7. PENGUJIAN
b. Verifikasi dan validasi
2 tipe dasar validitas DS:
• Validitas struktur: penilaian terhadap keserupaan struktur model dengan sistem nyata  VERIFIKASI.
 Berkaitan dengan: batasan sistem, variabel pembentuk sistem, dan asumsi mengenai interaksi yang
terjadi dalam sistem (Forrester, 1961).
 Merupakan hal yang kritis mengingat pendekatan model dinamika sistem pada dasarnya
berorientasi pada proses.
 Dapat dilihat dari segi kesesuaian model, konsistensi model, utilitas, dan efektivitas struktur model.
• Validitas kinerja (perilaku / output): penilaian terhadap hasil output kinerja (perilaku) dari suatu model
 VALIDASI.
 Diukur dari kemampuan model untuk menghasilkan ulang perilaku atau karakteristik suatu sistem
yang menjadi referensi.
 Salah satu pengujian model adalah kemampuan untuk menghasilkan ulang pola referensi pada hasil
keluarannya.
MPS – LD SEM 1-2017/2018 44
 7. PENGUJIAN
Validasi perilaku
b-1. Uji Reproduksi Perilaku: Theil Inequality

• Bias: U M

X m Xd  2

MSE
U M U S U C 1
• Unequal Variation:
s m  s d  2
U  S

MSE

21  r s m s d
• Unequal Covariation: UC 
MSE

MPS – LD SEM 1-2017/2018 45

 7. PENGUJIAN

Koefisien Korelasi :
1
r 
 
Xd  X d Xm  X m 
n sd sm

1
Rata - rata : X  X
n

Deviasi Standar : s
1
n
 X X 
2

1
Mean Square of Error : MSE  
 m d
X  X 2

n
MPS – LD SEM 1-2017/2018 46
 7. PENGUJIAN
b. Uji Model pada Kondisi Ekstrem
 Mengevaluasi perilaku model dalam kondisi ekstrem.
 Dalam kondisi ekstrim, model harus mampu menunjukkan robustness-nya, dimana perilaku
model mengikuti perilaku alami suatu kejadian.

c. Uji Kesalahan Integrasi

 Digunakan: untuk memeriksa sensitivitas model terhadap pemilihan time step dan metode
integrasi dalam simulasi.
 Model harus disimulasikan dengan beberapa time step dan metode integrasi yang berbeda.
 Dua metode integrasi: Euler dan Runge-Kutta.
 Idealnya, output simulasi tidak sensitif terhadap pemilihan time step, yang berarti model
memberikan perilaku yang sama untuk time step yang berbeda. Pemilihan metode Euler dalam
integrasi dapat diterima jika model tidak menunjukkan perbedaan yang signifikan jika
disimulasikan dengan Runge-Kutta.
MPS – LD SEM 1-2017/2018 47
 7. PENGUJIAN
d. Uji Prediksi Perilaku
 Dilakukan dengan tes prediksi kejadian (event prediction test) yang memfokuskan pada dinamika alami
suatu kejadian. Uji ini digunakan untuk melihat respon perilaku model ketika dilakukan perubahan
secara tiba-tiba pada suatu peubah.

80000 140,000

70000 120,000

60000
100,000

50000
80,000
40000
60,000
30000

40,000
20000 ACTUAL
Actual
SIMULATION
10000 Simulation 20,000

0 0
2000 2001 2002 2003 2004 2005 2006 2007 2008 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

MPS – LD SEM 1-2017/2018 48

 8. SYSTEM ARCHETYPES
ARCHETYPES  pola perilaku system yang berulang

System Thinking:

ESSENCES  Holism
 Intercon-
nectiveness

PRINCIPLES  Structure influence behavior

 Policy resistance
 Leverage

PRACTICE  System Archetypes

 Simulation

SYSTEM THINKING

MPS – LD SEM 1-2017/2018 49

 8. SYSTEM ARCHETYPES

ARCHETYPES
Archetype: pola perilaku yang muncul berulang kali; terjadi dalam berbagai disiplin – biologi, psikologi, terapi
keluarga, ekonomi, bidang politik, ekologi dan manajemen

• Memberikan titik pengungkit, titik waktu intervensi dimana perubahan subtansial dapat dilakukan.
• Mencakup satu atau lebih struktur dasar (reinforcing and balancing feedback loop).
• Sejumlah ARCHETYPES sistem telah diidentifikasi.
• Building blocks semua archetypes adalah: reinforcing processes, balancing processes, delays.
Permasalahannya:
• Bagaimana kita mengenalinya?
• Bagaimana kita menggambarkannya?
• Bagaimana kita memformulasikan cara mengatasinya?

MPS – LD SEM 1-2017/2018 50

 8. SYSTEM ARCHETYPES

Tipe Archetypes 1. Balancing with delay

2. Limits to Growth (aka Limits to Success)
3. Shifting the Burden
4. Eroding Goals
5. Escalation
6. Success to the Successful
7. Tragedy of the Commons
8. Fixes that Fail
9. Growth and Underinvestment
10. Accidental Adversaries
11. Attractiveness Principle
MPS – LD SEM 1-2017/2018 51
 6. ARCHETYPES (POLA DASAR) PERILAKU SISTEM

ARCHETYPE 1: Balancing Process with Delay

Struktur dasar: Perilaku sistem: Early Warning Symptom: "We

thought we were in balance, but
then we overshot the mark." (Later,
you may overshoot in the other
direction again.)
Management Principle: In a sluggish
system, aggressiveness produces
instability. Either be patient or make
Contoh: the system more responsive.
Business Story: Real estate
developers keep building new
properties until the market has gone
soft—but, by then, there are already
enough additional properties still
under construction to guarantee a
glut.
MPS – LD SEM 1-2017/2018 52
 6. ARCHETYPES (POLA DASAR) PERILAKU SISTEM
ARCHETYPE 2: Limits to Growth (limits to success)

Struktur dasar: Phenomenon: A reinforcing (amplifying)

Perilaku sistem: process is set in motion to produce a
desired result. It creates a spiral of success
but also creates inadvertent secondary
effects (manifested in a balancing process)
which eventually slow down the success.
Early Warning Symptom: "Why should we
worry about problems we don't have?
Contoh: We're growing tremendously." (A little
later, "Sure there are some problems, but
all we have to do is go back to what was
working before." Still later, "The harder
we run, the more we seem to stay in the
same place.")
Management Principle: Don't push on the
reinforcing (growth) process, remove (or
weaken) the source of limitation.

MPS – LD SEM 1-2017/2018 53

 6. ARCHETYPES (POLA DASAR) PERILAKU SISTEM
ARCHETYPE 3: Shifting The Burden Description: A short-term "solution" is
used to correct a problem, with
Contoh: seemingly positive immediate results. As
Struktur Dasar:
this correction is used more and more,
more fundamental long-term corrective
measures are used less and less. Over
time, the capabilities for the
fundamental solution may atrophy or
become disabled, leading to even
greater reliance on the symptomatic
solution.
Early Warning Symptom: "Look here,
Perilaku sistem: this solution has worked so far! What do
you mean, there's trouble down that
road?"
Management Principle: Focus on the
fundamental solution. If symptomatic
solution is imperative (because of delays
in fundamental solution), use it to gain
time while working on the fundamental
solution.
MPS – LD SEM 1-2017/2018 54
 6. ARCHETYPES (POLA DASAR) PERILAKU SISTEM
ARCHETYPE 3: Shifting The Burden

Contoh:
Perilaku sistem:

MPS – LD SEM 1-2017/2018 55

 6. ARCHETYPES (POLA DASAR) PERILAKU SISTEM
ARCHETYPE 4: Eroding Goals (special case of shifting the burden) Description: A shifting the burden type of structure in
which the short-term solution involves letting a long-
term, fundamental goal decline.
Struktur Dasar: Contoh:
Eariy Warning Symptom: "It's okay if our performance
standards slide a little, just until the crisis is over."
Management Principle: Hold the vision.
Business Story: A high-tech manufacturer finds itself
losing market share, despite a terrific product and
ongoing improvements. But the firm, oriented toward
its design "geniuses," had never gotten production
scheduling under control. An outside investigator
discovered that customers were increasingly
Perilaku sistem: dissatisfied with late schedules, and were turning to
competitors instead. The company stood on its record:
"We've maintained a consistent 90 percent success in
meeting the delivery time quoted to the customer." It
therefore looked elsewhere for the problem. However,
every time the company begin to slip its schedules, it
responded by making the quoted Delivery time a little
longer. Thus, the quoted delivery time to customers
was getting lengthier, and lengthier, and lengthier . . .
MPS – LD SEM 1-2017/2018 56
 6. ARCHETYPES (POLA DASAR) PERILAKU SISTEM
ARCHETYPE 5: Escalation (result oh fierce competition) Description: Two people or organizations each
see their welfare as depending on a relative
Struktur Dasar: Contoh: advantage over the other. Whenever one side
gets ahead, the other is more threatened, leading
it to act more aggressively to reestablish its
advantage, which threatens the first, increasing
its aggressiveness, and so on. Often each side
sees its own aggressive behavior as a defensive
response to the other's aggression; but each side
acting "in defense" results in a buildup that goes
far beyond either side's desires.
Perilaku sistem: Early Warning Symptom: "If our opponent
would only slow down, then we could stop
fighting this battle and get some other things
done."
Management Principle: Look for a way for both
sides to "win," or to achieve their objectives. In
many instances, one side can unilaterally reverse
the vicious spiral by taking overtly aggressive
"peaceful" actions that cause the other to feel
MPS – LD SEM 1-2017/2018 less threatened. 57
 6. ARCHETYPES (POLA DASAR) PERILAKU SISTEM
ARCHETYPE 6: Success to the successful (of the gap between the successful and the failed)
Description: Two activities compete for
Struktur Dasar: Contoh: limited support or resources. The more
successful one becomes, the more support
it gains, thereby starving the other.
Early Warning Symptom: One of the two
interrelated activities, groups, or
individuals is beginning to do very well and
the other is
struggling.
Management Principle: Look for the
Perilaku sistem: overarching goal for balanced achievement
of both choices. In some cases, break or
weaken the coupling between the two, so
that they do not compete for the same
limited resource (this is desirable in cases
where the coupling is inadvertent and
creates an unhealthy competition for
resources).

MPS – LD SEM 1-2017/2018 58

 6. ARCHETYPES (POLA DASAR) PERILAKU SISTEM
ARCHETYPE 7: Tragedy of the Commons Description: A fix, effective in the short term, has
unforeseen long-term consequences which may
Struktur Dasar: Contoh: Tabacco Industry require even more use of the same fix.
Early Warning Symptom: "It always seemed to
work before; why isn't it working now?“
Management Principle: Maintain focus on the long
term. Disregard short-term "fix," if feasible, or use it
only to "buy time" while working on long-term
remedy.
Business Story: A manufacturing company
launched a new set of high-performance parts,
Perilaku sistem: which were wildly successful at first. However, the
CEO was driven by maximizing his ROI, so he
deferred ordering expensive, new production
machines.
Manufacturing quality suffered, which led to a
reputation for low quality. Customer demand fell off
dramatically over the ensuing year, which depressed
returns and made the CEO even more unwilling to
MPS – LD SEM 1-2017/2018 invest in new production equipment. 59
 6. ARCHETYPES (POLA DASAR) PERILAKU SISTEM
ARCHETYPE 8: Fixes that fail (results of deminishing of returns) Description: Individuals use a commonly
available but limited resource solely on the
Struktur Dasar: Contoh: resource allocation basis of individual need. At first they are
rewarded for using it; eventually, they get
diminishing returns, which causes them to
intensify their efforts. Eventually, the
resource is either significantly depleted,
eroded, or entirely used up.
Early Warning Symptom: "There used to
be plenty for everyone. Now things are
getting tough. If I'm going to get any profit
out of it this year, I'll have to work harder."
Management Principle: Manage the
commons," either through educating
Perilaku sistem: everyone and creating forms of self-
regulation and peer pressure, or through an
official regulating mechanism, ideally
designed by participants.

MPS – LD SEM 1-2017/2018 60

 6. ARCHETYPES (POLA DASAR) PERILAKU SISTEM
ARCHETYPE 9: Growth and Underinvestment (result from late development of capacity, Description: Growth approaches a limit which
fail to forecast the new for capacity development) can be eliminated or pushed into the future if
Contoh: Capital planning the firm, or individual, invests in additional
Struktur Dasar: "capacity." But the investment must be
aggressive and sufficiently rapid to forestall
reduced growth, or else it will never get
made. Oftentimes, key goals or performance
standards are lowered to justify
underinvestment. When this happens, there
is a self-fulfilling prophecy where lower
goals lead to lower expectations, which are
then borne out by poor performance caused
by underinvestment.
Early Warning Symptom: "Well, we used to be
Perilaku sistem: the best, and we'll be the best again, but
right now we have to conserve our resources
and not over-invest."
Management Principle: If there is a genuine
potential for growth, build capacity in
advance of demand, as a strategy for creating
demand. Hold the vision, especially as
regards assessing key performance standards
and evaluating whether capacity to meet
potential demand is adequate.
MPS – LD SEM 1-2017/2018 61
 6. ARCHETYPES (POLA DASAR) PERILAKU SISTEM
ARCHETYPE 10: Accidental Adversaries (similar to escalation) Description: Accidental Adversaries is similar to the
Contoh: PPMC and Group Practice Escalation archetype in terms of the pattern of
Struktur Dasar: behavior that develops over time. It is different from it
insofar as the intent of the parties is concerned.
Accidental Adversaries begin their relationship with
win-win goals and objectives in mind, generally taking
advantage of their respective strengths, minimizing
their respective weaknesses, with the objective of
accomplishing together what cannot be achieved
separately.
Early Warning Symptom: This archetype states that when
teams or parties in a working relationship misinterpret
the actions of each other because of
misunderstandings, unrealistic expectations or
performance problems, suspicion and mistrust erode
the relationship. If mental models fueling the
deteriorating relationship are not challenged, all parties
may lose the benefits of their synergy.
Perilaku sistem:
Management Principle: • Revisit the original opportunity
that brought the parties together into a collaborative
relationship. • Use the archetype to identify the origins
of adversarial attitudes. • Renew the Shared Vision of
the collaborative effort and commit to Team Learning.
MPS – LD SEM 1-2017/2018 62
 6. ARCHETYPES (POLA DASAR) PERILAKU SISTEM
ARCHETYPE 10: Accidental Adversaries Struktur Dasar:
(similar to escalation, ‘musuh kebetulan’)

Perilaku sistem:

MPS – LD SEM 1-2017/2018 63

 6. ARCHETYPES (POLA DASAR) PERILAKU SISTEM
ARCHETYPE 11: Attractiveness Principle (similar to Limits to growth with many obstacles) Description: This archetype bears strong
resemblance to Limits to Growth with the
Struktur Dasar: Contoh: Consulting Firm addition of multiple slowing actions. Each
slowing action is a challenge the firm faces
and which must be addressed if the firm is
to overcome the aggregate limits to
growth.
Early Warning Symptom: The archetype
takes its name from the dilemma of
deciding which of the limits to address
first, that is, which is more attractive in
terms of the future benefit to the desired
Perilaku sistem: results that are being pushed by the effort
(or growing action).
Management Principle: • Focus on
identifying interconnected and
interdependent limits. • Use the archetype
to identify potential synergistic tactics to
remove the balancing processes before
they begin to affect growth. • Establish
priorities; carefully match available
MPS – LD SEM 1-2017/2018 64
resources with specific slowing actions.
 6. ARCHETYPES (POLA
DASAR) PERILAKU SISTEM
Relasi antar Archetypes

MPS – LD SEM 1-2017/2018 65

 7. REFERENSI

 Sterman, John D. (2000). Business Dynamics: Systems thinking and modeling for a complex world,
McGraw-Hill, 2000.
 Fishwick (1995). Simulation Model Design and Execution, Prentice-Hall.
 Senge, Peter M. (1990). The Fifth Discipline, The Art and Practice of The Learning Organization,,
Currency Doubleday, New York.
 Berbagai artikel.

MPS – LD SEM 1-2017/2018 66