Immune response-based algorithm for optimization of dynamic environments

Abstract

A novel immune algorithm suitable for dynamic environments (AIDE) was proposed based on a biological immune response principle. The dynamic process of artificial immune response with operators such as immune cloning, multi-scale variation and gradient-based diversity was modeled. Because the immune cloning operator was derived from a stimulation and suppression effect between antibodies and antigens, a sigmoid model that can clearly describe clonal proliferation was proposed. In addition, with the introduction of multiple populations and multi-scale variation, the algorithm can well maintain the population diversity during the dynamic searching process. Unlike traditional artificial immune algorithms, which require randomly generated cells added to the current population to explore its fitness landscape, AIDE uses a gradient-based diversity operator to speed up the optimization in the dynamic environments. Several reported algorithms were compared with AIDE by using Moving Peaks Benchmarks. Preliminary experiments show that AIDE can maintain high population diversity during the search process, simultaneously can speed up the optimization. Thus, AIDE is useful for the optimization of dynamic environments.

Keywords

dynamic optimization / artificial immune algorithms / immune response / multi-scale variation

References

[1] BRANKE J. Evolutionary optimization in dynamic environments [M]. Kluwer Academic Publishers, 2001: 10–12.

[2] BACK T U H. Evolution strategies applied to perturbed objective functions [C]// Proceedings of the Congress on Evolution Computing. Orlando, Florida, USA, 1994: 40–45.

[3] ANGELINE P J. Tracking extrema in dynamic environments [C]// Proceedings of the Evolutionary Programming VI. London, UK, 1997: 335–345.

[4] BRANKE J, KAULER T, SCHMIDTH C. A multi-population approach to dynamic optimization problems [C]// Proceedings of the 5th International Conference on Adaptive Computing in Design and Manufacturing. Plymouth, UK, 2000: 299–308.

[5] YANG S. Memory-based immigrants for genetic algorithms in dynamic environments [C]// Proceedings of the 2005 Genetic and Evolutionary Computation Conference. New York, NY, USA, 2005: 1115–1122.

[6] YangS., YaoX.. Experimental study on population-based incremental learning algorithms for dynamic optimization problems [J]. Soft Computing, 2005, 9(11): 815-834

[7] YaochuJ., BrankeJ.. Evolutionary optimization in uncertain environments-a survey [J]. IEEE Transactions on Evolutionary Computation, 2005, 9(3): 303-317

[8] DE CASTRO L, VON ZUBEN F. Artificial Immune Systems: PART II-A survey of applications [R]. DCA-RT, 2000.

[9] GASPAR A, COLLARD P. From gas to artificial immune systems: Improving adaptation in time dependent optimization [C]// Proceedings of the Congress on Evolutionary Computation. Washington D C, USA, 1999: 1859–1866.

[10] WALKER J, GARRETT S. Dynamic function optimization: comparing the performance of clonal selection and evolution strategies [C]// The Second International Conference in Artificial Immune Systems. Berlin, Germany, 2003: 273–284.

[11] GongM.-g., JiaoL.-c., ZhangX.-rong.. A population-based artificial immune system for numerical optimization [J]. Neurocomputing, 2008, 72(12): 149-161

[12] DE CASTRO L, TIMMIS J. An artificial immune network for multimodal optimization [C]// Congress Evolutionary Computation. New York, NY, USA, 2002: 699–704.

[13] DE FRANCA F, ZUBEN F J V, de CASTRO L N. An artificial immune network for multimodal function optimization on dynamic environments [C]// Proceeding of GECCO. New York, NY, USA, 2005: 289–296.

[14] LuoY., LiR., ZhangW.. Dynamic function optimization algorithm based on immune mechanism [J]. Journal of Xi’an Traffic University, 2005, 39(4): 384-388

[15] SimõesAnabela, CostaErnesto. An Immune System-Based Genetic Algorithm to Deal with Dynamic Environments: Diversity and Memory. Artificial Neural Nets and Genetic Algorithms, 2003, Vienna, Springer Vienna: 168-174

[16] ZhangZ.-h., QianS.-qu.. Immune algorithm with dynamic environments and its application to greenhouse control [J]. Optim Eng, 2008, 11(1): 125-144

[17] TrojanowskiK., WierzchonS. T.. Immune-based algorithms for dynamic optimization [J]. Information Sciences, 2009, 179: 1495-1515

[18] FarmerJ. D., PackardN. H., PerelsonA. S.. The immune system, adaptation, and machine learning [J]. Physica, 1986, 22(3): 187-204

[19] KesmirC., BoerR.. A mathematical model on germinal center kinetics and termination [J]. The Journal of Immunology, 1999, 163(5): 2463-2469

[20] SunY.-z., DaiX.-h., WeiWei.. Model of artificial immune response [J]. Journal of Zhejiang University: Engineering Science, 2004, 38(6): 682-685

[21] BlackwellT.. Particle swarm optimization in dynamic environments [J]. Studies in Computational Intelligence, 2007, 51: 29-49

[22] PARROTT D, LI X. A particle swarm model for tracking multiple peaks in a dynamic environment using speciation [C]// Congress on Evolutionary Computation. New York, NY, USA, 2004: 98–103.

[23] BRANKE J, KAULER T, SCHMIDT C. A multi-population approach to dynamic optimization problems [M]// Adaptive Computing in Design and Manufacturing, 2000.

[24] MORRISON R W, JONG D E. Triggered hypermutation revisited [C]// Congress on Evolutionary Computation. New York, NY, USA, 2000: 1025–1032.

[25] GeraisM.. A gradient-based artificial immune system applied to optimal power flow problems [J]. LNCS, 2007, 4628: 1-12

[26] Pradyumn Kumar Shukla.. Gradient based stochastic mutation operators in evolutionary multi-objective optimization [J]. LNCS, 2007, 4431: 58-66

[27] WangX., GaoX. Z., OvaskaS. J.. A Hybrid optimization algorithm based on ant colony and immune principles [J]. International Journal of Computer Science & Applications, 2007, 4(3): 30-44

[28] DE FRANÇA F O. VON ZUBEN F J, DE CASTRO L N. An artificial immune network for multimodal function optimization on dynamic approach [C]// Proceedings of the Genetic and Evolutionary Computation Conference. GECCO 2005. New York, NY, USA, 2005: 289–296.

[29] BRANKE J. Memory enhanced evolutionary algorithm for changing optimization problems [C]// Proceedings of the Congress on Evolutionary Computation. Washington D C, USA, 1999: 1875–1882.