presented by
|
presented by
|
|
(PDF, approx. 1 meg)
Researchers around the world are using Genetic Algorithms (GAs) to optimize many designs including antennas, aircraft engines, computer programs and even job schedules, with dramatic improvements in quality and design time. The GA is a probabilistic, iterative optimization strategy that mimics biological adaptation and evolution through mating and survival-of-the-fittest. It is very robust in difficult design spaces, finding good solutions to complex problems while avoiding local minima and exploring only a very small portion of the design space. Solutions found by GAs also tend to be counter-intuitive, with unexpected parameters that often work better than designs created by engineers at great expense and time. The GA finds these solutions with little information about the problem and minimal involvement from the engineer--not even an initial guess. This talk will introduce GAs by exploring concepts and several examples of GA optimization. |