Investigation Of Modified Genetic Algorithm And Its Application In Optimization To Electric Apparatus

Nowadays, with the rapid development of science and technology, investigation field of electric devices is also changing fast .Design of modern electric apparatus demands strictly in the volume, structure and insulation characteristics of electric devices. Modern optimization theory is introduced to solve the multi-variables problem for which has been beyond the capacity of traditional optimal method. Thus, Modern optimization theory has been a focus in the academic field at home and abroad, and, kinds of optimization algorithms are developed.Genetic algorithm is adaptive during the application and its method and skill are in a difference when dealing with different problems. In this paper, based on the traditional research of genetic algorithm, a modified genetic algorithm is put forward and the definition of population rank is also introduced. By studies on the population rank during the optimization process, decision function used to judge the genetic algorithms premature is given, based on which the optimization control is carried out.MATLAB language is adopted to program the codes. Based on specified function, the relationship between rank and genetic algebra, population scale is gained and whether the population premature or not is easily judged. The introduced modified genetic algorithm deals with the premature problem well, which also increases the population diversity and avoids early maturing. This method proves to be feasible and correct.By applying the modified method to the electrical apparatus electrode optimization, the optimization results are got, from which the electric- field strength is uniformed can be seen.