Multi-objective optimization using evolutionary algorithms.

*(English)*Zbl 0970.90091
Chichester: Wiley. xix, 497 p. (2001).

The author discusses two multi-objective optimization procedures, namely, the ideal procedure and the preference-based one. In the ideal multi-objective optimization procedure he considers that the effort must be made in finding the set of trade-off optimal solutions by considering all objectives to be important. After a set of such trade-off solutions are found, a user can then use higher-level qualitative considerations to make a choice. In the preference-based multi-objective optimization procedure, based on the higher-level information, a relative preference factor among the objectives is first chosen, which leads to a composite objective function to be optimized in order to find a single trade-off optimal solution. The author argues why the ideal procedure is less subjective, more methodical and more practical than the preference-based procedure.

Most of the book is devoted to finding multiple trade-off solutions for multi-objective optimization problems, although the book includes a number of techniques to find a preferred distribution of trade-off solutions, if information about the relative importance of the objectives is available. Based on the ability of an Evolutionary Algorithm (EA) to find multiple optimal solutions in one single simulation run, the book presents various techniques of finding multiple trade-off solutions using EAs.

Chapter 2 provides the principles of multi-objective optimization. Chapter 3 describes various methods according to the preference-based approach. Chapter 4 presents four EAs by discussing their differences with classical search and optimization methods. In chapters 5, 6 and 7 the author presents multi-objective EAs. Salient issues related to design, development and application in Multi-Objective Evolutionary Algorithms (MOEAs) are discussed in chapter 8. Chapter 9 is devoted to five applications of MOEAs.

Most of the book is devoted to finding multiple trade-off solutions for multi-objective optimization problems, although the book includes a number of techniques to find a preferred distribution of trade-off solutions, if information about the relative importance of the objectives is available. Based on the ability of an Evolutionary Algorithm (EA) to find multiple optimal solutions in one single simulation run, the book presents various techniques of finding multiple trade-off solutions using EAs.

Chapter 2 provides the principles of multi-objective optimization. Chapter 3 describes various methods according to the preference-based approach. Chapter 4 presents four EAs by discussing their differences with classical search and optimization methods. In chapters 5, 6 and 7 the author presents multi-objective EAs. Salient issues related to design, development and application in Multi-Objective Evolutionary Algorithms (MOEAs) are discussed in chapter 8. Chapter 9 is devoted to five applications of MOEAs.

Reviewer: F.Guerra (Puebla)