Professional Documents
Culture Documents
SHUFFLED FROG
LEAPING ALGORITHM
1
OPTIMIZATION BY SFLA
1. INTRODUCTION
2
OPTIMIZATION BY SFLA
2. OPTIMIZATION
2.1 Definition
The optimization of systems and processes is veryimportant to the efficiency and economics
of many scienceand engineering domains. Optimization problems aresolved by using
rigorous or approximate mathematical search techniques. Rigorous approaches have
employed linear programming, integer programming, dynamic programming or branch-and-
bound techniques to arrive at the optimum solution for moderate-size problems. However,
optimizing real-life problems of the scale often encountered in engineering practice is much
more challenging because of the huge and complex solution space. Finding exact solutions to
these problems turn out to be NP-hard. This kind of complex problem requires an exponential
amount of computing power and time, as the number of decision variables increases
(Lovbjerg 2002). To overcome theseproblems, researchers have proposed
approximateevolutionary-based algorithms as a means to search for near-optimum solutions.
3
OPTIMIZATION BY SFLA
o Mathematical Algorithms
Simplex (LP), BFGS (NLP), B&B (DP)
o Drawbacks of Mathematical Algorithms
LP: Too Ideal (All Linear Functions)
NLP: Not for Discrete Var. or Complex Fn., Feasible Initial
Vector, Local Optima
DP: Exhaustive Enumeration, Wrong Direction
o Meta-Heuristic Algorithms
GA, SA, TS, ACO, PSO,
4
OPTIMIZATION BY SFLA
3. S HUFFLED FROG LEAPING ALGORITHM
3.1 Definition
The shuffled frog-leaping algorithm is a memetic metaheuristicthat is designed to seek a
global optimal solutionby performing a heuristic search.It is based on the evolutionof memes
carried by individuals and a global exchange ofinformation among the population.
The SFLA (Shuffled Frog Leaping Algorithm) is presented by Eusuff and Lansey
(2003) is a meta-heuristic iterative method inspired from the memetic evolution of a
group of frogs when seeking for food.
The SFLA is a method which is based on observing, imitating, and modelling the
behaviour of a group of frogs when searching for the location that has the maximum
amount of available food .
The SFL algorithm, in essence, combines the benefits of the genetic-based MAs and the
social behavior-based PSO algorithms. In the SFL, the population consists of a set of frogs
(solutions) that is partitioned into subsets referred to as memeplexes. The different
memeplexes are considered as different cultures of frogs, each performing a local search.
Within each memeplex, the individual frogs hold ideas, that can be influenced by the ideas of
other frogs, and evolve through a process of memetic evolution. After a defined number of
memetic evolution steps, ideas are passed among memeplexes in a shuffling process.The
local search and the shuffling processes continue until defined convergence criteria are
satisfied.
5
OPTIMIZATION BY SFLA
Improve the worst frog position;
Repeat for a specific number of iterations;
End;
Combine the evolved memeplexes;
Sort the population P in descending order of their
fitness;
Check if terminationZtrue;
End;;
Within each memeplex, position of frogs with the best and worst fitness is determined as Xb
and Xw,respectively. Also position of frog with the global best fitness is determined as Xg.
Then, in eachmemeplex, a process is applied to improve only the frog with the worst fitness
(not all frogs) in eachcycle as follows:
6
OPTIMIZATION BY SFLA
where Rand() is a random number between 0 and 1. If this process generates a better solution,
the worstfrog will be replaced. Otherwise, the calculations in (1) and (2) are repeated with
replacement of Xbby Xg. If no improvement becomes possible in this case, then a new
solution is randomly generatedwithin the feasible space to replace the worst frog.
7
OPTIMIZATION BY SFLA
factor, as follows:
Assigning a large value to the factor C at the beginning of the evolution process will
accelerate the global search byallowing for a bigger change in the frogs position
andaccordingly will widen the global search area. Then, as theevolution process continues
and a promising location isidentified, the search acceleration factor, C, will focus
theprocess on a deeper local search as it will allow the frogs tochange its positions.
8
OPTIMIZATION BY SFLA
4. PROCESS - SFLA
9
OPTIMIZATION BY SFLA
10
OPTIMIZATION BY SFLA
Within each memeplex, the frogs with the best and the worst fitnesses are identified as Xb
and Xw, respectively. Also, the frog with the global best fitness is identified as Xg. Then, a
process similar to PSO is applied to improve only the frog with the worst fitness (not all
frogs) in each cycle.
Accordingly, the position of the frog with the worst fitness is
adjusted as follows:
where rand( ) is a random number between 0 and 1; andDmax is the maximum allowed
change in a frogs position. If this process produces a better solution, it replaces the worst
frog. Otherwise, the calculations in Eqs. (8) and (9) are repeated but with respect to the global
best frog (i.e.Xg replaces Xb). If no improvement becomes possible in this case, then a new
solution is randomly generated toreplace that frog. The calculations then continue for
aspecific number of iterations . Accordingly, the mainparameters of SFL are: number of frogs
P; number ofmemeplexes; number of generation for each memeplexbefore shuffling; number
of shuffling iterations; andmaximum step size.
11
OPTIMIZATION BY SFLA
5. .COMPARISION AMONG DIFFERENT EVOLUTIONARY
ALGORITHMS
SFLA has been used as appropriate tools to obtain the best solutions with the least total time
and cost by evaluating unlimited possible options. One of the problems of previous research
is that assumptions make them unrealistic in comparison with actual construction projects.
On the other hand, delay events during execution of activities have an important impact on
total time and cost of projects. Therefore, the authors attempt to make the model better
approximate real projects by considering splitting during execution of activities.
Compare non-dominated solutions ofSFLA by applying splitting to previous works in GA
and NSGA-II. Results in both TCO andTCRO models demonstrate improvement of solutions,
convergence ratio, and the processingtime to reach the optimum solution. It confirms that
SFLA improves results, by comparing results before applying. Since in this case, we do not
have any limit for resources, the impact of splitting on concepts of time-cost trade off and
resource allocation has been investigated. The values of improvement demonstrate that
splitting has significant impact on final results.
Benchmarkcomparisons among the algorithms are presented for both continuous and discrete
optimization problems, in terms of processing time,convergence speed, and quality of the
results. Based on this comparative analysis, the performance of EAs is discussed along with
someguidelines for determining the best operators for each algorithm. The study presents
12
OPTIMIZATION BY SFLA
sophisticated ideas in a simplified form that should bebeneficial to both practitioners and
researchers involved in solving optimization problems.
SFLA has been used as appropriate tools to obtain the best solutions with the least total time
and cost by evaluating unlimited possible options. Thevalues of improvement demonstrate
that shuffled frog leaping algorithm has significant impact on efficiency of various aspects.
.
13
OPTIMIZATION BY SFLA
6. CONLUSIONS
In this paper a optimization method known as shuffled frog leaping algorithm has been
described. It comes under the type evolutionary algorithms which means those algorithms
which are prepared by observing the typical natural biological evolutions and behaviour of
natural species such as ants, birds, bees, frogs etc.
A brief description of shuffled frog leaping algorithm method is presented along with a
pseudocode to facilitate its implementation. Visual Basic programs were written to
implement each algorithm. Also presented were the comparative results found when a
discrete optimization test problem was solved using all five algorithms. The PSO method was
generally found to perform better than other algorithms in terms of success rate and solution
quality, while being second best in terms of processing time.
Evolutionary algorithms (EAs) are stochastic search methods that mimic the natural
biological evolution and/or the social behavior ofspecies. Such algorithms have been
developed to arrive at near-optimum solutions to large-scale optimization problems, for
which traditionamathematical techniques may fail.
14
OPTIMIZATION BY SFLA
7. REFERENCES
[1] Al-Tabtabai H, Alex PA. Using genetic algorithms to solve optimization problems in
construction. EngConstrArchit Manage 19996:12132.
[2] Duan, Q.Y., Gupta, V.K. and Sorooshian, S., Shuffled complex evolution approach
for effective and efficient global minimization. J. Optimization Theory Appns, 1993,
76, 502 521.
[3] Elbeltagi, E., Hegazy, T. and Grierson, D., Comparison among five evolutionary-
based optimization algorithms. J. Adv. Engng. Informatics, 2005, 19, 43 53.
[4] Eusuff, M.M. and Lansey, K.E., Optimization of water distribution network design
using the shuffled frog leaping algorithm. J. Water Resour. Planning Mgmt, 2003,
129, 210 225.
[5] Feng, C., Liu, L. and Burns, S., Using genetic algorithms to solve construction time
cost trade-off problems. J. Comput. Civil Engng, 1997, 11, 184 189.
[6] Hegazy T. Optimization of construction time-cost trade-off analysis using genetic
algorithms. Can J Civil Eng 1999;26:68597.
[7] Hegazy, T., Elbeltagi, E. and Elbehairy, H., Bridge deck management system with
integrated life cycle cost optimization. Transportation.
[8] Holland, J., Adaptation in Natural and Artificial Systems, 1975 (University of
Michigan Press: Ann Arbor, MI).
[9] Kennedy, J. and Eberhart, R., Particle swarm optimization, in Proceedings IEEE
International Conference on Neural Networks, IEEE Service Center, Piscataway, NJ,
pp. 1942 1948, 1995.
15