A genetic algorithm‐based improved charge simulation method and its application

• Fan Yang ^[1] ; Wei He ^[1] ; Wendong Deng ^[2] ; Tao Chen ^[3]
  1. [1] Chongqing University

     Chongqing University

     China

     
  2. [2] Chongqing Jiangbei Power Supply Bureau
  3. [3] Chongqing Electric Power Research Institute

  Mostrar afiliaciones +
• Localización: Compel: International journal for computation and mathematics in electrical and electronic engineering, ISSN 0332-1649, Vol. 28, Nº 6 (Special Issue: Selected Papers from ICSC 2008), 2009, págs. 1701-1709
• Idioma: inglés
• Enlaces
  • Texto completo
• Resumen

  • Purpose – The purpose of this paper is to propose an improved charge simulation method (CSM) based on the real‐coded genetic algorithm (GA), in which the tedious testing and adjusting work in the traditional CSM is avoided. In addition, less simulation charges are used in the simulation model for the same object compared with the traditional CSM.

    Design/methodology/approach – In the improved CSM, the information of testing points are combined with the matching points, hence the size and locations of simulation charges can be only computed by the information of matching points, and the testing calculation in the traditional CSM is avoided. According to the Maxwell equations, an overdetermined equation is reformulated, and an improved GA is used to solve it. The process to create the initial population is improved, boundary condition information of part matching points and human experience are added into the initial population, which enhanced the convergence rate of the search calculation.

    Findings – Combining the information of testing points into that of the matching points in the improved CSM, the tedious testing calculation can be avoided, and the GA is effective to solve the overdetermined equation for the improved CSM. By importing the information of part matching points and human experience, the search process can be accelerated and the convergence of the method is ensured. Comparison between the efficiency of traditional CSM and the improved CSM is presented, less simulation charges are used compared with the traditional CSM for the same object.

    Research limitations/implications – The improved CSM cannot adjust the count of simulation charges in the equivalent model automatically.

    Practical implications – Calculations of a sphere‐plane electrode model and a 64 kV insulator string are carried out to verify the validity of the improved CSM. In addition, it can also be used to compute the power frequency electric field of HV transmission apparatuses.

    Originality/value – By improving the process of the traditional CSM and importing the GA, an improved CSM for the calculation of power frequency electric fields produced by high‐voltage apparatuses is presented in the paper, which can avoid the tedious testing and adjusting work in the traditional CSM.