The Research On The Intelligent Charge Simulation Method

In the research of insulation and breaking characteristic of the high voltage apparatus or equipment, the calculation and analyze of the-field with open boundary problems is largely referred to. While the charge simulation method (CSM) has possessed peculiar advantage in the calculation of this kind of problem. However, several problems such as the field with multi-dielectric the field with thin electrode and how the program is realized and so on are uncounted with CSM. In this paper the further study is made on these questions in this paper.First of all, the basic theory of CSM is introduced for solving the problems of multi-dielectric field with conventional CSM, A kind of new method is put forward in this paper. The method is to divide multi-dielectric arrangements into regions based on the structure 'and material characteristic of the arrangement. The field and potential inside each region are calculated by only taking into account the discrete charges that surround it.Besides, when the field with thin electrode is solved by the use of conventional CSM, the discrete charges are required putting inside the electrode, but it is impossible for the electrode with thinnest thickness. So the electrode boundary-displaced CSM is brought forward to settle the problem in the paper. The property of the new method is to divide the field base into several regions on the shape of the thin electrode. For each region, there is a set of discrete charges belonging to it accordingly. The potential and the electric field in each region are calculated with the corresponding charges. Thereby the above problem can be avoided effectively.In consideration of the defects of conventional GA, an improved GA has been investigated in this paper. The evolution speed and quality of the population are directly influenced by the change of the number of chromosomes in GA and by whether the paternal excellent information is passed to the offspring as much as it can. In regard to the questions existing in GA, an algorithm with dynamic population scale is provided in this paper.In the genetic process of reproduction, crossover and mutation of the chromosomes in this method, these operators Pr,Pc and Pm are produced randomly within some space, the scale of population and all kinds of genetic probability are also adjusted randomly so that the diversity individuals of population is ensured. The GA of dynamic population scale passes more information of paternal chromosomes to the offspring, which is beneficial to search the global optimization or quasi-global optimization. The calculation of an instance proves the method valid.The above-mentioned methods are combined to optimize the field of the breaker, and the calculation results are compared with the results calculated by finite element method.The practical calculation shows that the approach provided in thjs paper is a valid method to solve the problems that existed in the classic CSM. Through improving, the kinds of optimization method-GA with dynamicscale of population have a better convergence characteristic than the traditional one.