A Strategy Of Genetic Operations Based On Schema

Evolutionary computation is a remarkable subject which after the artificalintelligence system and artificial neural network. It successfully simulates the biologicalevolution which proposed by Darwin. According to the principle of ‘survival of the fittest’,evolutionary computation makes the initial population approached to the optimumsolution by the operators of selection, crossover and mutation. Basically, evolutionarycomputation is a search optimization technology which has adaptive regulatory function.In the USA, Germany, France and other developed countries, evolutionary computationhas been used in many fields, such as machinery, chemical industry, building andcomputers. It successfully solved the structure optimization, nonlinear optimization,parallel computing and complex problems. In the past10years, evolutionary computationdeveloped fastly. In China, evolutionary computation has been paid much attention.Especially the Genetic Algorithm (GA), has been used in many fields. But for thedevelopment of ages, problems have been more and more complex. So the algorithmsmust be more accurate and faster. Because of the ‘premature’ phenomenon and lowaccurate, GA has been seriously restricted in complex optimization problem. Therefor,based on schema theorem, this paper proposed a strategy of genetic operations based onschema. It effectively improved the shortages above.In order to solve the low computing and convergence problems, according to binarycoding, this paper first proposed a protecting strategy based on schema. Then we built anew GA based on schema (BS-GA). In this paper, we have given the idea of schemaselection and the specific implement steps. Then we gave the steps of protecting strategyfor crossover and mutation operators. Finally, we analyze the performance of BS-GA byMarkov chains and simulation technology. Sencondly, according to real coding, wesimilary proposed aprotecting strategy based on schema and built a new GA based onschema (BS10-GA). Then we give all the steps and simulation technology like binarycoding. At last, we found that, BS-GA and BS10-GA can largely improve the computingefficiency and stability of algorithms.