基于有限元和神经网络方法对超高压合成绝缘子均压环结构优化的研究

针对超高压输电线路合成绝缘子所用的均压环,提出了一种优化其结构参数的方法.这种方法首先利用渐进边界条件解决了绝缘子电场模型的开域问题,然后以有限元方法计算出均压环各个参数对绝缘子电场分布的影响,并以此确定了均压环结构优化目标;在此基础上,利用神经网络拟合了均压环各结构参数与优化目标之间的关系,并以此得出均压环结构参数的最优解,解决了穷举优化方法的计算时间冗长的缺点.优化后的均压环使得绝缘子和均压环的表面电场强度均小于电晕起始场强.

OPTIMIZATION OF CORONA RING DESIGN FOR EHV COMPOSITE INSULATOR USING FINITE ELEMENT AND NEURAL NETWORK METHOD

A method to optimize the location and the dimensions of the corona ring is presented for EHV transmission line composite insulators. This method is used to handle the open boundary problem of the insulator model by means of the asymptotic boundary conditions. Then, in order to set the optimal goal, the finite element method is employed to calculate the electrical field distribution along the composite insulator with corona ring which is of various location and dimensions. Moreover, the neural network model is built to map the location as well as the dimensions of the corona ring and the optimal goal, which can avoid the tedious calculation caused by the method of exhaustion. The optimized corona ring makes the maximum E-field stress along the insulator surface and the one along the corona surface both below the corona inception level.