【作者】 孙阳；

【导师】 阎秀恪；

【作者基本信息】 沈阳工业大学 ， 电机与电器， 2012， 硕士

【摘要】 油浸倒立式电流互感器区别于传统的正立式结构,将二次绕组与一次绕组集中置于整个产品的上部,避免了正立式电流互感器主绝缘位于产品底部易受潮的问题,减少了主绝缘因受潮而被击穿的可能性,但同时也给产品的绝缘设计和工艺制造增添了一定的难度。油浸式电流互感器在110kV电压等级以上通常采用电容型油纸绝缘结构,即在主绝缘中嵌入电容屏来改善电场分布。针对于倒立式的特殊结构,增加电容屏的方式一般分为两种:一种为端屏式结构,另一种则为主屏式结构。本文采用主屏式结构,以220kV油浸倒立式电流互感器为研究对象,分析了主绝缘的物理结构特点,采用数值和解析两种方法求解屏间电容,并验证了两种方法的准确性。根据倒立式电流互感器主屏式结构的特点,采用分层分段的方式建立电场分析的有限元模型,将二次侧绕组与下引线部位分开,对于二次侧绕组部位进行三维有限元分析,对于下引线部位进行二维有限元分析,并采用解析法验证了有限元计算电场的准确性。根据电场分布以及最大场强位置,分析影响绝缘性能的主要因素,提出了优化方案。本文采用RBF神经网络动态响应模型与遗传算法相结合的方法对主屏结构的油浸倒立式电流互感器的主绝缘结构进行优化。根据动态神经网络的思想,建立动态响应模型,使响应模型跟随最优点的位置不断更新、细化,避免了传统响应模型采样点过多的缺陷;采用幂律标定方法增强了遗传算法的选择功能。通过优化降低了最大场强,同时也使主绝缘整体的电场分布得到了改善。更多还原

【Abstract】 Oil-immersed inverted current transformer is different from the traditional structure, its primary winding and second winding are set on the top of the current transformer. This structure can keep main insulation away from wetness. However, it also added some difficulties to the insulation design and manufacture process.Oil-immersed current transformer in 110 kV voltage level above usually adopts capacitance type insulation structure. Capacitance screens which embedded in main insulation are connected in series to improve the distribution of the electric field. There are two ways adding capacitance screens in main insulation, one is adding some end capacitance screens, and the other is adding some main screens. This paper adopts adding main screens in insulation of 220kV oil-immersed current transformer, and analyzes the main insulation physical structure. The numerical and analytical have been adopted to compute the capacitance between screens. The accuracy of two method have been verified each other. According to the structure of oil-immersed invert current transformer, the FEM calculation model has been constructed separately. Through 3D and 2D electric field FEM analysis, the distribution of electric field can be obtained, and the location of maximum electric field strength can be found. The results have been compared with the results calculated by analytical method, and the accuracy has been verified. Based on the electric field analysis, an optimization scheme for main insulation has been proposed in this paper.The RBF neural network dynamic response model combined genetic algorithm (GA) has been proposed to optimize the main insulation of oil-immersed invert current transformer with the main screen structure. The dynamic response model had been established by the dynamic neural network, and the response model has been updated and refined following the position of the optimal point constantly. It resolved the problem for traditional response model that too many sample points need mass computation. Made the power function of objective function as the fitness function to enhance the option function of GA. After optimization the maximum of the electric field has been reduced, and the electric field distribution in the main insulation has been improved.更多还原