Multiple Populations Coevolution And The Application In The Generative Conception Design

Science and technology is entering the era crossing , permeating , influencing each othermulti-disciplinarily at present, life sciences and project scientific to cross , infiltration and itmutuallies promote to be among them one typical example, it is a remarkable characteristic of themodern science technical development too. Evolve flourishing development of algorithm reflectcharacteristic and trend this of scientific development.The technique of generative and evolutionary design has been widely investigated in the lasttwo decades. In fact, it has become one of the most important techniques for computational andcreative design. There is a large number of authors and literature who contributed to this researchfield, covering such topics as generative systems, genetic programming, creative evolutionarysystems etc. Among these, P.J. Bentley's thesis (Bentley, P.J. [1]) systematically revealed the ideaof using genetic algorithms to create and optimise conceptual designs involving a range ofdifferently-shaped three-dimensional solid objects. He presented a generic evolutionary designsystem with genetic algorithm (GA) as its core and demonstrated the successful evolution of bothconventional and unconventional designs for fifteen different solid-object design tasks, e.g. tables,heatsinks, penta-prisms, boat hulls, aerodynamic cars.In contrast to this generative and evolutionary design approach there is another importantwork carried out by Lecienne Blessing in her thesis in 1994 (Luccinne Blessing [7]), theprocess-based design system. She investigated prescriptive models of the design process,handbooks and catalogues containing knowledge needed for design, methods to support specificdesign activities, and computer based tools. She proposed a solution to the need to improve designthrough an emphasis on the process and through a combination of these means. Herimplementation is a computer based design support system developed around a model of design inwhich the process is the core consisting four stages: (1) determination of the characteristics ofdesign; (2) specification of the type of support; (3) development of a set of requirements andfunctions for the support system; (4) development of the model and the support systems and (5)evaluation of the core of the system.Of the many advantages that evolution can provide, it is a powerful way of providing varietyof possible design solutions that can be optimised and analysed conveniently, and has beensuccessfully applied to various civil and engineering design problems (Frazer, J.H. [3-5]).Evolutionary design facilitates mechanisms to find optimal solutions in a space of possiblesolutions where evolution process is applied to populations. In fact, evolution is not only anoptimisation tool, but also a searching and generative tool in that it synthesises solutions bycombining the configurations of other alternatives. New and unexpected individuals can begenerated by two most useful operations, that is, crossover and mutation. The kernel part of anevolution system, genetic algorithms, can maintain and evolve both populations of designs andpopulations of the design operations.Despite the fact that evolution is probably the most promising tool of the future, itsapplication is often reduced to simple evolution. This restriction is due to algorithmic andtechnical complexity. The limitation resulted in the development of more powerful algorithms for...