The primary objective of this chapter is to introduce Artificial Immune Systems (AIS) as a relatively new bio-inspired optimization technique and to show its appeal to engineering applications. The advantages and disadvantages of the new computing paradigm, compared to other bio-inspired optimization techniques, such as Genetic Algorithms and other evolution computing strategies, are highlighted. Responding to some aforementioned disadvantages, a population adaptive based immune algorithm (PAIA) and its modified version for multi-objective optimization are put forward and discussed. A multi-stage optimization procedure is also proposed in which the first stage can be regarded as a vaccination process. It is argued that PAIA and its variations are the embodiments of some new characteristics which are recognized nowadays as the key to success for any stochastic algorithms dealing with continuous optimization problems, thus breathing new blood into the existing AIS family. The proposed algorithm is compared with the previously established evolutionary based optimization algorithms on ZDT and DTLZ test suites. The promising results encourage us to further extract a general framework from the PAIA as the guild to design immune algorithms. Finally, a real-world engineering problem relating to the building of a transparent fuzzy model for alloy steel is presented to show the merits of the algorithm.
Bio-Inspired Computing lies within the realm of Natural Computing, a field of research that is concerned with both the use of biology as an inspiration for solving computational problems and the use of the natural world experiences to solve real world problems. The increasing interest in this field lies in the fact that nowadays the world is facing more and more complex, large, distributed and ill-structured systems, while on the other hand, people notice that the apparently simple structures and organizations in nature are capable of dealing with most complex systems and tasks with ease. Artificial Immune Systems (AIS) is one among such computing paradigms, which has been receiving more attention recently.
AIS is relatively a new research area which can be traced back to Farmer et al.’s paper published in 1986 (Farmer, J. D. & Packard, N. H., 1986). In this pioneering paper the author proposed a dynamical model for the immune systems based on the Clonal Selection Principle (Bernet, F. M, 1959) and Network Hypothesis (Jerne, N. K., 1974; Perelson, A. S., 1989). However, there were only a few developments since then until 1996 when the first international conference based on artificial immune systems was held in Japan. Following this event, the increasing number of researchers involved in this field indicated the emergence of the new research field: Artificial Immune Systems. But hitherto, no new formal framework based on AIS has been proposed.
There are three main application domains which AIS research effort has focused on, viz. fault diagnosis, computer security, and data analysis. The reason behind this is that it is relatively easy to create a direct link between the real immune system and the aforementioned three application areas, e.g. in the applications of data analysis, clusters to be recognized are easily related to antigens, and the set of solutions to distinguish between these clusters is linked to antibodies. Recently, a few attempts to extend AIS to the optimisation field have been made (de Castro & Von Zuben, 2002; Kelsey, J. & Timmis, J., 2003). However, as mentioned by Emma Hart and Jonathan Timmis (2005), maybe by historic accident, many of the AIS practitioners arrive in the optimisation field by way of working in other biologically inspired fields such as Evolutionary Computing (EC), and thus in terms of optimisation the distinctive line between EC and AIS is vague. In other words, there is not a formal distinctive framework for AIS applied to optimisation. The situation is even worse when it comes to multi-objective optimisation (MOP) case since it is hard to find a way to define Antigen and the affinity due to the implicit Antigen population to be recognized (Chen J. & Mahfouf, M., 2006). Based on such an understanding, this chapter will present a systematic AIS framework to solve MOP with clear definitions and roles of the immune metaphors to be employed, and will highlight the difference between AIS and traditional EC to finally discover the extra advantages which are exclusively inherent in AIS.
Key Terms in this Chapter
Antigen (Ag): Ag is the problem to be optimized.
Transparency of the Fuzzy Model: A fuzzy model is regarded as having a better transparency if it contains less fuzzy rules, less fuzzy sets and less overlapped fuzzy sets
Pareto Front: The plot of the objective functions whose non-dominated vectors are in the Pareto optimal set is called the Pareto front.
Ab-Ab Suppression (Abs’ suppression/Network suppression): When two Abs are very close to each other, they can be recognized by each other. The result is that one of them is suppressed and deleted. Unlike Abs’ affinity, this term is defined as the Euclidian distance in the objective space
Antibody (Ab): Ab is the candidate solutions of the problem to be optimized.
Ab-Ab Affinity (Abs’ affinity): Is defined as the distance (refer to Eqs. (3)) in the decision variable space between one randomly chosen Ab in the first non-dominated front and the one in the remaining population.
Ag-Ab Affinity: For SOP, it is defined as the objective value (fitness value) for MOP, it is determined by using non-dominance concept, i.e. solutions in the first non-dominated front have the highest affinity, then the second front and so on
Complete Chapter List
Fabio Freschi, Carlos A. Coello Coello, Maurizio Repetto
Jun Chen, Mahdi Mahfouf
Licheng Jiao, Maoguo Gong, Wenping Ma
Malgorzata Lucinska, Slawomir T. Wierzchon
Luis Fernando Niño Vasquez, Fredy Fernando Muñoz Mopan, Camilo Eduardo Prieto Salazar, José Guillermo Guarnizo Marín
Fabio Freschi, Maurizio Repetto
Krzysztof Ciesielski, Mieczyslaw A. Klopotek, Slawomir T. Wierzchon
Xiangrong Zhang, Fang Liu
Yong-Sheng Ding, Xiang-Feng Zhang, Li-Hong Ren
Alexander O. Tarakanov
Xin Wang, Wenjian Luo, Zhifang Li, Xufa Wang
Mark Burgin, Eugene Eberbach
Terrence P. Fries
Konstantinos Konstantinidis, Georgios Ch. Sirakoulis, Ioannis Andreadis
Miroslav Bursa, Lenka Lhotska
Martin Macaš, Lenka Lhotská
James F. Peters, Shabnam Shahfar
Tang Mo, Wang Kejun, Zhang Jianmin, Zheng Liying