8 - Evolutionary and Adaptive Synthesis Methods

Summary

The simplest scheme of evolution is one that depends on two processes; a generator and a test. The task of the generator is to produce variety, new forms that have not existed previously, whereas the task of the test is to cull out the newly generated forms so that only those that are well fitted to the environment will survive.

— Herbert A. Simon, The Sciences of the Artificial, 1969, MIT Press

INTRODUCTION

The synthesis of novel engineering designs can take on the character of an exploration, especially when the exploration has the capability to experiment, either in a random or directed manner, not simply with the values of parameters that describe a single design configuration, but with a wide range of different configurations.

Since the 1960s, there has been an increasing interest in simulating the natural evolutionary process to solve optimization problems, leading to the development of adaptive and stochastic exploration techniques (Fogel, 1962, 1994, 1998; Holland, 1962; Rechenberg, 1965; Schwefel, 1965; Goldberg, 1989; Hammel, Bäck, and Schwefel, 1997) that can often outperform conventional optimization methods when applied to challenging real-world problems. Evolutionary and adaptive search methods, such as genetic algorithms and simulated annealing, stochastically refine or alter individual candidate solutions in a population, evaluate the fitness or performance of these new candidates, and keep only those with good fitness values for the next iteration.