| With the continuous improvement of China’s power grid system and the gradual improvement of voltage levels,extra-high voltage and ultra-high voltage transformers have emerged one after another,which brings new challenges to the structural design of transformers.As an indispensable device in the power grid,the quality of the transformer’s insulation structure determines the safe and stable operation of the whole power system,which is not only related to economic problems,but also to the normal life and work of the people.Therefore,it is the most basic and important requirement for transformer manufacturing to analyze the insulation of transformer and design the insulation structure reasonably so that it can work reliably under high voltage level and ensure that the insulation is not damaged.In this paper,response surface methodology(RSM)is used to optimize the insulation structure of the electrostatic ring of a 500 k V converter transformer.First of all,the boundary value problem and finite element analysis method of electric field are introduced.The traditional optimization method is combined with the call between MATLAB and ANSYS,and ANSYS parametric modeling language(APDL)is used to automatically complete the process of modeling,meshing,applying boundary conditions and extracting calculation results.The influence of each variable to be optimized related to the electrostatic ring on the maximum electric field strength is analyzed respectively,and the variation law of each variable is summarized.Then,the assumptions of response surface optimization method are analyzed and introduced in detail.The RSM based on Taguchi algorithm is used to optimize the electrostatic ring structure of the transformer.After Taguchi algorithm is used to eliminate irrelevant variables,the response surface test is carried out.Design-Expert regression analysis software is used to perform analysis of variance,diagnostic analysis,and perturbation analysis on the response surface model to verify the significance and applicability of the model regression.A response surface model with good fitting effect and high prediction quality is obtained.Finally,the RSM is compared with the traditional optimization method,the results of the two methods are basically the same,the maximum electric field intensity after optimization has been significantly reduced,which verifies the effectiveness of response surface method for the optimization of the electrostatic ring structure.From the comparison results,it can be seen that the optimization time used by the RSM is shorter,the optimization efficiency is higher,the optimization process is more systematic and complete,and it is more convincing and rigorous than the traditional optimization method.It is a safe and reliable optimization method. |
