Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar...

61
Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008 [email protected] King Saud University College of Engineering Industrial Engineering Department م ي ح ر ل ا ن م ح ر ل له ا ل م ا س ب

Transcript of Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar...

Page 1: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

Solving Stochastic Project Scheduling Problems Using Simulation/Optimization

Approach

By: Omar Al-ShehriSupervised by: Prof. A. M. Al-Ahmari

Winter 1429/2008

[email protected]

King Saud UniversityCollege of Engineering

Industrial Engineering Department

بسم الله الرحمن الرحيم

Page 2: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

Contents

1 -Introduction.

2 -Stochsticity (Problem definition).

3 -Project Objectives.

4 -Solution Methodology.

5 -Converting the AON Network into Simulation

Model.

6 -Future Work (In IE 499) .

Page 3: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

1 -Introduction

-10 Trillion dollar are invested in projects world wide.

-60 Million professional person

involved.

-More profitable projects, means more GDP and more growth.

Page 4: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

The Project Life Cycle

Initiation Planning Scheduling

ExecutingMonitoring and control

Closing

Page 5: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

Problems with Projects

1 -Unexpected necessary activities.

2 -Tracking the plan.

3 -Actual vs. Planned makespan .Time

Num

ber o

f ac

tiviti

es

The actual path of the project.

The planned path of the project.

Page 6: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

The project manager.

!This is the

Stochasticity Zone!

Page 7: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

2 -Stochasticity

- For scheduling real world problems, there are:

1 -Many uncertainties.

2 -Complex relations

between factors .

3 -Many constraints.

4 -Many non-linearities .

?????????

Page 8: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

Is it possible to build the

model?

NoWhat to do?

Will we get the optimal

solution?

No

UsingSimulation

UsingOptimization

Very long time.

How long it takes?

Yes

Page 9: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

Simulation + Optimization

Simulation Optimization overcome that, where:

1 -The simulation model this

stochasticity.

2 -The optimization manage it.

Page 10: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

Trad

ition

alop

timiz

atio

nSimulation optimization

Many realistic problems

Size

Time

Page 11: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

Arena & OptQuest

-We used Arena software for simulation

modeling.

-And we will use OptQuest for optimization.

Arena software

OptQuest

Performance estimates

Candidate results

Page 12: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

3 -The Project Objectives

1 -To suggest a proper scheme for converting jjffproject network into Arena model.

2 -To determine the optimum number or the llllresources required by the project, as well as llllthe makespan.

Page 13: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

4 -The Solution Methodology1 -Identifying a set of rules for converting the

network into Arena model.

2 -Modeling the stochastic resource constrained project when the resources are subjected to

break downs, using Arena.

3 -Linking the developed model into OptQuest.

Page 14: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

The Methodology (Continued)

4 -Verifying and validating the simulation using

optimization model using simulation

experiment.

5 -Interpreting and analyzing the results.

Page 15: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

5 -Converting of the Network

-We are basically dealing with the activity on node networks (AON).

-Based on the Arena modules, we can divide the AON network into four basic elements:

1 -Starting node (source).

2 -Activities nodes.

3 -Arrows.

4 -Finishing node .Start Finish84

3

2

5 6

7

1

Page 16: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

5.1 converting of Starting Node

Start

Page 17: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

5.2 converting of Activities

1

Page 18: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

5.2 converting of Activities

32

Page 19: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

5.2 converting of Activities

4

5

6

Page 20: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

5.2 converting of Activities

5

6

7

Page 21: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

5.3 converting of Finishing Node

7 Finish

Page 22: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

5.4 Complete Network & Model

8

4

3

2

5 6

7

Start Finish1

Page 23: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

The selected project has the following network:

First Case study

Page 24: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

The stochastic project data are as follows:

Activity Act. TimeResources Needed

Worker Machine

1 Norm(10,2) 1 1

2 Norm(12,3) 2 -

3 Unif(5,8) 2 2

4 5 1 -

5 Tria(4,5,6) 2 1

6 Expo(12) 1 -

7 Norm(10,1) 1 2

8 Tria(4,6,8) 2 -

Simulation Stage

Page 25: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

Using the scheme which we had developed, the corresponding Arena model is as depicted:

Simulation Stage (Cont.)

Page 26: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

-We defined some priorities which will represent the sequence of the activities which will take

place.- -This priorities was defined using

- Assign module in Arena Basic- Process Panel.

Simulation Stage (Cont.)

Page 27: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

-This priorities will be the controls which will be defined in OptQuest.

-The objective is to minimize the project completion time or makespan.

Simulation Stage (Cont.)

Page 28: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

Optimization Stage

Page 29: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

Optimization Stage (Cont.)

- -Now, getting into OptQuest, the following data are defined:

OBJECTIVE: Minimize the Project Completion Time.

Page 30: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.
Page 31: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

CONTROLS: Predetermined Priorities.

Optimization Stage (Cont.)

Page 32: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.
Page 33: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

RESPONCES:

1 -PCT.

2 -The Project’s Single Entity.

Optimization Stage (Cont.)

Page 34: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.
Page 35: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.
Page 36: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

THE PROJECT PARAMETERS:

1 -Number of Replications.

2 -Tolerance when two solutions are equal.

3 -Others.

Optimization Stage (Cont.)

Page 37: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.
Page 38: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

Optimization Stage (Cont.)

Now, we can run the program and get the results.

Page 39: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.
Page 40: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

-The projects activities sequence is as follows:

-The various solutions for the various number of replications for both approaches are :

Activity 1 2 3 4 5 6 7 8Sequence 1 2 4 3 6 5 7 8

Results and Discussion

Page 41: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

Number of Runs

Average MakespanUsing Simulation

OnlyUsing Simulation

Optimization

100 67.43 57.8500 66.91 57.03

1000 66.26 56.42

e.g. for the 1000 replication)66.26-56.42/(66.26=14.85% had been reduced

from the makespan.

Results and Discussion (Cont.)

Page 42: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

Second Case Study

For this example, we used the same network of the previous case but we added a failure to the machines.

Also, we have chosen another objective for this case.

Page 43: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

Simulation Stage

The machine failure rate is 5 hours for every expo(10) hours up time.

Page 44: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.
Page 45: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

Optimization Stage

Page 46: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

Optimization Stage (Cont.)

- -Now, getting into OptQuest, the following data are defined:

OBJECTIVE: Minimize the Project Cost.

The equation used for calculating total cost is:PCT*100+PCT*10*[Machine1]+PCT*20*[Worker]

Project holding cost

M/C costPerhour

Worker costPerhour

No. of M/Cs No. of workers

Project completion time

Page 47: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.
Page 48: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

CONTROLS:

1-Predetermined Priorities.

2 -Recourses (Machine and Workers)

Optimization Stage (Cont.)

Page 49: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.
Page 50: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

RESPONCES:

1 -PCT.

2 -The Project’s Single Entity.

Optimization Stage (Cont.)

Page 51: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.
Page 52: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.
Page 53: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

THE PROJECT PARAMETERS:

1 -Number of Replications.

2 -Tolerance when two solutions are equal.

3 -Others.

Optimization Stage (Cont.)

Page 54: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.
Page 55: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

Optimization Stage (Cont.)

Now, we can run the program and get the results.

Page 56: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.
Page 57: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

-The projects activities sequence is as follows:

-And the optimal no. of resources is three workers and four machines.

-The various solutions for the various number of replications for both approaches are :

Activity 1 2 3 4 5 6 7 8Sequence 2 3 1 5 4 6 7 8

Results and Discussion

Page 58: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

e.g. for the 1000 replication) 10697.60-8901.903/(10697.60 =16.78% had

been reduced from the total cost.

Number of Runs

Average Cost

Using Simulation Only

Using Simulation Optimization

100 10865.92 9073.379

500 10699.20 8915.489

1000 10697.60 8901.903

Results and Discussion (Cont.)

Page 59: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

Conclusion

-Simulation optimization is a helpful approach in the project scheduling where

the activity times are stochastic .

-It has been found in this project that there is good improvement when optimization tool

is used with simulation model .

-It would be good for further research to develop automatic transformation tool for model primary data to simulation model.

Page 60: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

Conclusion (Cont.)

• -In addition, linking the simulation model with other optimization tool such as Genetic Algorithm will simplify comparisons

between these optimization tools .

Page 61: Solving Stochastic Project Scheduling Problems Using Simulation/Optimization Approach By: Omar Al-Shehri Supervised by: Prof. A. M. Al-Ahmari Winter 1429/2008.

[email protected]

Thank You

Q & A