Другие журналы

scientific edition of Bauman MSTU


Bauman Moscow State Technical University.   El № FS 77 - 48211.   ISSN 1994-0408

Research and Setting the Modified Algorithm "Predator-Prey" in the Problem of the Multi-Objective Optimization

# 12, December 2016
DOI: 10.7463/1216.0852423
Article file: SE-BMSTU...o181.pdf (1848.94Kb)
authors: A.P. Karpenko1,*, A.V. Pugachev1

1 Bauman Moscow State Technical University, Moscow, Russia

We consider a class of algorithms for multi-objective optimization - Pareto-approximation algorithms, which suppose a preliminary building of finite-dimensional approximation of a Pareto set, thereby also a Pareto front of the problem. The article gives an overview of population and non-population algorithms of the Pareto-approximation, identifies their strengths and weaknesses, and presents a canonical algorithm "predator-prey", showing its shortcomings.
We offer a number of modifications of the canonical algorithm "predator-prey" with the aim to overcome the drawbacks of this algorithm, present the results of a broad study of the efficiency of these modifications of the algorithm. The peculiarity of the study is the use of the quality indicators of the Pareto-approximation, which previous publications have not used. In addition, we present the results of the meta-optimization of the modified algorithm, i.e. determining the optimal values of some free parameters of the algorithm.
The study of efficiency of the modified algorithm "predator-prey" has shown that the proposed modifications allow us to improve the following indicators of the basic algorithm: cardinality of a set of the archive solutions, uniformity of archive solutions, and computation time. By and large, the research results have shown that the modified and meta-optimized algorithm enables achieving exactly the same approximation as the basic algorithm, but with the number of preys being one order less. Computational costs are proportionally reduced.

  1. Deb K. Multi-objective optimization using evolutionary algorithms. Chichester; N.Y.: Wiley, 2001. 497 p.
  2. Karpenko A.P., Semenikhin A.S., Mitina E.V. 77-30569/363023. Population methods of Pareto set approximation in multi-objective optimization problem: Review. Nauka i obrazovanie MGTU im. N.E. Baumana = Science and Education of the Bauman MSTU,2012, no. 4.DOI: 10.7463/0412.0363023
  3. Laumanns M., Rudolph G., Schwefel H.P. A spatial predator-prey approach to multi-objective optimization: a preliminary study. Parallel problem solving from nature: 5th Intern. conf. on parallel problem solving from nature (Amsterdam, Netherlands, Sept. 27 – 30, 1998): proceedings. B.: Springer, 1998. Pp. 241 -- 249. DOI: 10.1007/BFb0056867
  4. Xiaodong Li. A real-coded predator-prey genetic algorithm for multiobjective optimization. Evolutionary multi-criterion optimization: 2nd Intern. conf. on evolutionary multi-criterion optimization. EMO-2003 (Faro, Portugal, April 8-11, 2003): proceedings. B.: Springer, 2003. Pp. 207-221. DOI: 10.1007/3-540-36970-815
  5. Kalyanmoy Deb, Udaya Bhaskara Rao N. Investigating predator-prey algorithms for multi-objective optimization. KanGAL Report. 2005. No. 2005010. 12 p.
  6. Chowdhury S., Dulikravich G. S., Moral R.J. Modified predator-prey algorithm for constrained and unconstrained multi-objective optimization. Int. J. of Mathematical Modelling and Numerical Optimisation, 2009, vol. 1, no. 1/2, pp. 1 -- 38. DOI: 10.1504/IJMMNO.2009.030085
  7. Surafel Luleseged Tilahun, Hong Choon Ong. Prey-predator algorithm: A new metaheuristic algorithm for optimization problem. Intern. J. of Information Technology & Decision Making, 2015, vol. 14, no. 6, pp. 1331 -- 1352. DOI: 10.1142/S021962201450031X
  8. Silva A., Neves A., Costa E. An empirical comparison of particle swarm and predator prey optimization. Artificial intelligence and cognitive science: 13thIrish Intern. conf. AICS 2002 (Limerick, Ireland, Sept. 12-13, 2002): proceedings. B.: Springer, 2002. Pp. 103-110. DOI: 10.1007/3-540-45750-X_13
  9. Grimme C., Schmitt K. Inside a predator-prey model for multi-objective optimization: A second study. Genetic and evolutionary computation conf. GECCO’06(Seattle, Washington, USA, July 8–12, 2006): proceedings. Vol. 1. Pp. 707 -- 714. DOI: 10.1145/1143997.1144121
  10. Eiben A. E., Michalewicz Z., Schoenauer M., Smith J. E. Parameter control in evolutionary algorithms // Parameter setting in evolutionary algorithms. B.: Springer, 2007. Pp. 19 -- 46. DOI: 10.1007/978-3-540-69432-8 2
  11. Eshelman L.J., Schaffer J.D. Real-coded genetic algorithms and interval- schemata. Foundations of genetic algorithms - 2: Workshop on foundations of genetic algorithms: FOGA-92 (Vail, Colo., 1992). San Mateo: Morgan Kaufmann, 1993. Pp. 187 -- 202. DOI: 10.1016/B978-0-08-094832-4.50018-0
  12. Michalewicz Z. Genetic algorithms + data structures = evolutionary programs. B.; N.Y.: Springer, 1992. 250 p.
  13. Sobol’ I.M., Statnikov R.B. Vybor optimal’nykh parametrov v zadachakh so mnogimi kriteriiami[The choice of optimal parameters in tasks with many criteria]. 2nd ed. Moscow: Drofa Publ., 2006. 175 p.
  14.  Kalyanmoy Deb, Thiele L., Laumanns M., Zitzler E. Scalable test problems for evolutionary multi-objective optimization. Evolutionary multiobjective optimization. L.: Springer, 2005. Pp. 105-145. DOI: 10.1007/1-84628-137-7_6
  15.  Zitzler E., Kalyanmoy Deb, Thiele L. Comparison of multiobjective evolutionary algorithms: Empirical results. Evolutionary Computation, 2000, vol. 8, iss. 2, pp. 173-195. DOI: 10.1162/106365600568202
  16.  Karpenko A.P. Sovremennye algoritmy poiskoskovoj optimizatsii[Modern algorithms of search engine optimization]. Moscow: MSTU im. N.E. Baumana, 2014. 446 p.
elibrary crossref ulrichsweb neicon rusycon

About Project
Rambler's Top100
Phone: +7 (915) 336-07-65 (строго: среда; пятница c 11-00 до 17-00)
© 2003-2019 «Наука и образование»
Перепечатка материалов журнала без согласования с редакцией запрещена
 Phone: +7 (915) 336-07-65 (строго: среда; пятница c 11-00 до 17-00)