Appendix C: Biologically Inspired Optimization
Witold Pedrycz, Fernando A. C. Gomide · 2007
T#o fully benefit from the potential of fuzzy sets and information granules as well as all constructs emerging there, there is a genuine need for effective mechanisms of global optimization.It is equally important that such an optimization framework comes with substantial capabilities of structural optimization of fuzzy systems.It is highly advantageous to have systems whose structure could be seamlessly modified to fully exploit the capabilities of the constructs of fuzzy sets.It would be highly desirable to consider constructs whose scalability can be easily realized.Biologically inspired optimization offers a wealth of optimization mechanisms that tend to fulfill these essential needs.The underlying principles of these algorithms relate to the biologically motivated schemes of system emergence, survival, and refinement.Quite commonly, we refer to the suite of these techniques as Evolutionary Computing to directly emphasize the inspiring role of various mechanisms encountered in the Nature that are also considered as pillars of the methodology and algorithms.The most visible feature of most, if not all, such algorithms is that in their optimization pursuits they rely on a collection of individuals that interact between themselves in the synchronization of joint activities of finding solutions.They communicate between themselves by exchanging their local findings.They are also influenced by each other. Evolutionary OptimizationEvolutionary optimization offers a comprehensive optimization environment in which we encounter a stochastic search that mimics natural phenomena of genetic inheritance and Darwinian strife for survival.The objective of evolutionary optimization is to find a maximum of a certain objective function f defined in some search space E. Ideally, we are interested in the determination of a global maximum of f.