redecs 20011 adaptive ship maintenance rescheduling 24 - 25 october, 2001 residence hotel uniten...

28
REDECS 2001 1 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN) -- Presenter HAJAR MAT JANI (UNITEN) NOR’ASHIKIN ALI (UNITEN)

Upload: alban-hodge

Post on 04-Jan-2016

218 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 1

ADAPTIVE SHIP MAINTENANCE RESCHEDULING

24 - 25 October, 2001RESIDENCE HOTELUNITENKAJANG

PATHIAH ABDUL SAMAT (UPM)ALICIA TANG Y. C. (UNITEN) -- Presenter

HAJAR MAT JANI (UNITEN)NOR’ASHIKIN ALI (UNITEN)

Page 2: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 2

AGENDA (1)

PROBLEM DEFINITION– WHAT IS THE PROBLEM?– OBJECTIVES

BACKGROUND INFORMATION– WHAT HAD BEEN DONE?

OUR APPROACH– CBR + GA– HOPFIELD Neural Network– Operational Research Framework

Page 3: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 3

AGENDA (2)

SOFTWARE CONCLUSION FUTURE WORKS

Page 4: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 4

PROBLEM DEFINITION (1)

Ships - assets in naval defence Ships - expensive They should be fully utilised High rate of availability is anticipated AVAILABILITY

– depends on effectiveness of Preventive Maintenance Schedule (PMS)

Unable to avoid rescheduling!!

Page 5: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 5

PROBLEM DEFINITION (2) If (uncertainty) breakdowns occur

–availability of ship is Low availability and high

maintenance costs are problems in ship maintenance management

This problem can jeopardise the defence system of the country

Page 6: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 6

PROBLEM DEFINITION (3)

SHIP MAINTENANCE (RE)SCHEDULING– is a process of deciding start-times of

maintenance activities that satisfy all precedence and resource constraints & optimize the ship availability.

variables

domains

constraintsresult

Page 7: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 7

Objectives: Proposals–to develop Adaptive Algorithms

• to decide (select) which activity to reschedule

–to develop Hopfield Neural N.• to reschedule

PROBLEM DEFINITION (4)Go There

Click Me

Page 8: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 8

MAINTENANCE SCHEDULE FOR A SHIP

Factors– Running hours of the ships– Operational requirement– Status of parts availability– Status of operational defects– Dockyard availability

Page 9: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 9

BACKGROUND INFORMATION (1)

Scheduling / time-tabling problem – Neural Network

– Constraints Logic Programming

– Graph Coloring

– Heuristics, etcE.g. ILOG, CHIP

Page 10: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 10

BACKGROUND INFORMATION (2)

CONSTRAINT SOLVING– Reduce search domain/space

– therefore faster & save storage

– how? It minimizes backtracking

• Solve problems: ‘design’, ‘diagnosis’ & ‘planning’• Build schedule that satisfies ‘temporal’ and ‘resource’ constraints

Page 11: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 11

BACKGROUND INFORMATION(4)

Improve G.A. by improving chromosome representation

(increase ship availability) Achieved by search space (such as minimising overlapping of

maintenance activity)

WHAT HAD BEEN DONE?

Table 1

overlappingRefer to articles 1 & 3, references section.

Page 12: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 12

OUR APPROACH (1)

USE GA– To “optimise”

USE CBR– To find near optimum

schedule that maximises availability

Hybrid Vsjust CBR

Page 13: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 13

OUR APPROACH (2)

TO RE-SCHEDULE:–USE HOPFIELD NN

• CONSTRAINTS

• NEURON

–BASED ON CBR-GA DERIVED DATA

2 LAYERS

Soumen and Badrul (1996) - rescheduling of power system

Item#7

Page 14: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 14

THE HYBRID G.A. ALGORITHM

Step 1: code the start times and pattern of activity

Step 2: create initial population Step 3: determine start times and pattern of

activity by the GA Step 4: build feasible schedule using CBR Step 5: evaluate the schedule.

Page 15: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 15

Proposal

Analyse data and research question

Changes,i.eUncertaintybreakdown Process constraint

Time andResourceConstraint

Analyse the part of CBR-GA where adaptive enhancement can be made

Testing the adaptive algorithm

Develop adaptive algorithm

Adaptive?

Analyse existing performance measure

Reconstructperformancemeasure

Analyse rescheduling algorithm

Develop rescheduling algorithm

Testing rescheduling algorithm

Reschedule?optimiseReportresult

Develop newperformancemeasure

R. O.

FRAMEWORK

N

N

N

Page 16: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 16

SOFTWARE

PLATFORM–Unix, Windows

NT/ME/2000/9xPROPOSED LANGUAGE

–C++Used in previous works

Page 17: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 17

Proposed Software Components

Scheduling program Ship program (Solver) Constraints program G.A Maintenance program Many header files Adaptive scheduler Rescheduling using Hopfield Neural Net

Keeps repeating until “fit” enough

Page 18: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 18

Procedure GA-CBR

BeginCreate constraint variable S1

Read constraints C(S1)Post constraints C(S1) to variable S1

Post constraint C(Si) to variable S1

t 0initialise P(t)evaluate P(t)while (not termination condition) do

Begint t + 1select P(t) from P(t – 1)alter P(t)copy allele values from P(t) to value v of constraint variable S1

BeginPartitioning values v, into pattern of activity, c and pattern of activity, stuses pattern of activity c to create pattern of activity C1 for each ship activitiescopy st into first maintenance activities, stset type of resource, sumPbpost constraint from file, smscstPROCESS-CONSTRAINTS( Si, v)Total objective-function, obj, for each domain,Get total-objective, objTot for whole activities, mtnReturn objTot

EndEndReturn legal values v to genome in population P(t)Evaluate P(t)End While

End GA-CBR

G.ACBR

G.A

Page 19: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 19

g

g

f

f

f

x1

x2

xn

g

c1

c2

c3

b1

b2

b3

Constraints layer Neurone layer

Constraints Also constraints New Schedule

Page 20: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 20

CONCLUSION (1) Re-design of existing algorithms is

necessary. Therefore, new algorithms need to

be developed. Reschedule of activities based on

the temporal and resource constraints is required so as to adapt to the changes that may occur.

Rescheduling Algorithms

Page 21: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 21

CONCLUSION (2)

•CBR + G.A - to produce near optimum solution.

•Enhancement to be made to CBR.

•Hopfield Neural Network - to reschedule selected activities.

Our solutions:

Page 22: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 22

FUTURE WORKS

Fuzzy Logic - to address “over constraints” of the selection of activities and the rescheduling process.

Application in other areas: School time-tabling, Financial control and planning, Classification & Prediction.

Page 23: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 23

THE END

Thank You.

Questions?

Page 24: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 24

Improve Chromosome Representation

ChromosomeFitness function(minimising number of overlapping)

With pattern activities Ship of class A 0.84 Ship of class B 0.75

Without pattern activities Ship of class A 0.98 Ship of class B 0.82

less

higher

Page 25: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 25

Schedule Overlapping

Ship 1

Ship 2

Ship 3

Ship 4Weeks

Overlapping!!

Page 26: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 26

CBR Vs. Hybrid

Comparison between the CBR and the hybrid approaches:

Approaches Objective function (minimisingno. of overlapping activities)

CBR alone 950.76CBR+GA 0.98

CBR alone 1540.20CBR+GA 0.82

Class A

Class B

Page 27: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 27

Pattern activities and start-time

Start time Pattern-1 Pattern-2 Pattern-3 Pattern-4 Pattern-5 Pattern-61234::16

1 2 3 ……….. 67 8 ……………..13 ……….:::91 ……………………… 96

An allele

Combination of no. of activities + duration of operation

Refer to figure 2, full paper

Page 28: REDECS 20011 ADAPTIVE SHIP MAINTENANCE RESCHEDULING 24 - 25 October, 2001 RESIDENCE HOTEL UNITEN KAJANG PATHIAH ABDUL SAMAT (UPM) ALICIA TANG Y. C. (UNITEN)

REDECS 2001 28

Values of GA parameters for Ship Class A No. of population = 45 No. of generation = 60 Probability of mutation = 0.01 Type of crossover = single-point Type of GA = steady state Size of chromosome = 4 Size of allele = 96 Fitness function = maximise availability Scaling = Linear scaling