Content area
Full Text
J Optim Theory Appl (2010) 145: 613615
DOI 10.1007/s10957-010-9701-6
B O O K R E V I E W
Published online: 28 April 2010 Springer Science+Business Media, LLC 2010
This unique book arose out of the pioneering efforts of Yu-Chi Ho and his colleagues (two of whom, Qian-Chuan Zhao and Qing-Shan Jia, are co-authors) and his students to address complex problems that arise routinely in engineering design and that, because of their complexity, cannot be solved using standard optimization tools. A salient feature of the systems to be designed is that they cannot be simply described by, for example, a set of differential equations. Typically the system is described by a complex simulation program. The time taken for one simulation for a turbine blade design problem, one of the case studies considered in the book, is 7 days; conventional optimization requiring, say, merely 100 values of the performance index would take nearly two years. Optimization is further complicated by the fact that often the observed performance is merely an estimate of the true performance either because the design problem is inherently stochastic or because a simple model of the underlying system is employed to estimate performance. This book proposes, and supports by a set of case studies, a useful procedure for design problems of this type; the use of a simple model to estimate performance contributes signicantly to reduction of the computational effort required.
Two principles underly the approach taken in this book. The rst is that, in the presence of uncertainty, it is much easier to estimate whether one design is better than another than to estimate the difference in performance; order is easier than value. The error in choosing one design as superior to another using noisy estimates of their performances decrease rapidly as the difference between their true perfo-
Communicated by A. Miele.
D.Q. Mayne ( )
Department of Electrical and Electronic Engineering, Imperial College London, London SW7 2BT, UKe-mail: mailto:[email protected]
Web End [email protected]
Yu-Chi Ho, Qian-Chuan Zhao and Qing-Shan Jia: Ordinal Optimization: Soft Optimization for Hard...