Principles of Optimal Design: Modeling and Computation

This text discusses modelling for design optimization. It presents a condensed version of classical optimization theory and numerical algorithms, which it integrates with the newer ideas of monotonicity analysis and model boundedness. Careful definition of new concepts and rigorous proof of simple but important principles are followed by immediate applications. It begins with the definition of modelling and the optimization problem and outlines the limitations of this approach. The authors then move on to discuss the important but rarely emphasized concepts of boundedness checking, the idea that the parameters of every model should be verified and simplified; and monotonicity analysis, a method of determining which variables actively constrain a model. Then the discussion turns to the classical theory of differential optimization and hence to powerful numerical optimization techniques. Extensive examples and exercises aid the student and provide practice. A knowledge of differential calculus is helpful.

As computers have become more powerful, design engineers have been able to tackle more complex systems, and the term "optimization" is now routinely used to denote a design process with increased speed and quality . This second edition of Principles of Optimal Design takes account of these developments and brings the original text thoroughly up to date. New material includes a discussion of trust region and convex approximation algorithms, how to construct optimal design models, case studies to illustrate the creation of optimization models, and an expanded chapter on computation of derivatives, interpretation of algorithmic results, and selection of algorithms and software.