No abstract

No abstract

VIII Prefaceusing different data structures for chromosome representation of individuals, and various "genetic" operators which operated on them. Because of my background in databases [122], [126], where the constraints playa central role, most evolution programs were developed for constrained problems.The idea of evolution programs (in the sense presented in this book), was conceived quite early [101], [133] and was supported later by a series of experiments. Despite the fact that evolution programs, in general, lack a strong theoretical background, the experimental results were more than encouraging: very often they performed much better than classical genetic algorithms, than commercial systems, and than other, best-known algorithms for a particular class of problems. Some other researchers, at different stages of their research, performed experiments which were perfect examples of the "evolution programming" technique -some of them are discussed in this volume. Chapter 8 presents a survey of evolution strategies -a technique developed in Germany by I. , [162] for parameter optimization problems. Many researchers investigated the properties of evolution systems for ordering problems, including the widely known, "traveling salesman problem" (Chapter 10). In Chapter 11 we present systems for a variety of problems including problems on graphs, scheduling, and partitioning. Chapter 12 describes the construction of an evolution program for inductive learning in attribute based spaces, developed by C. Janikow [97]. In the Conclusions, we briefly discuss evolution programs for generating LISP code to solve problems, developed by J. Koza [108], and present an idea for a new programming environment.

Evolutionary computation techniques have received a great deal of attention regarding their potential as optimization techniques for complex numerical functions. However, they have not produced a significant breakthrough in the area of nonlinear programming due to the fact that they have not addressed the issue of constraints in a systematic way. Only recently have several methods been proposed for handling nonlinear constraints by evolutionary algorithms for numerical optimization problems; however, these methods have several drawbacks, and the experimental results on many test cases have been disappointing. In this paper we (1) discuss difficulties connected with solving the general nonlinear programming problem; (2) survey several approaches that have emerged in the evolutionary computation community; and (3) provide a set of 11 interesting test cases that may serve as a handy reference for future methods.

Summary. The issue of setting the values of various parameters of an evolutionary algorithm is crucial for good performance. In this paper we discuss how to do this, beginning with the issue of whether these values are best set in advance or are best changed during evolution. We provide a classification of different approaches based on a number of complementary features, and pay special attention to setting parameters on-the-fly. This has the potential of adjusting the algorithm to the problem while solving the problem. This paper is intended to present a survey rather than a set of prescriptive details for implementing an EA for a particular type of problem. For this reason we have chosen to interleave a number of examples throughout the text. Thus we hope to both clarify the points we wish to raise as we present them, and also to give the reader a feel for some of the many possibilities available for controlling different parameters.

No abstract

No abstract

During the last five years, several methods have been proposed for handling nonlinear constraints using evolutionary algorithms (EAs) for numerical optimization problems. Recent survey papers classify these methods into four categories: preservation of feasibility, penalty functions, searching for feasibility, and other hybrids. In this paper we investigate a new approach for solving constrained numerical optimization problems which incorporates a homomorphous mapping between n-dimensional cube and a feasible search space. This approach constitutes an example of the fifth decoder-based category of constraint handling techniques. We demonstrate the power of this new approach on several test cases and discuss its further potential.