Font Size: a A A

Research On Autotransformer Winding Deformation Monitoring Based On Parameter Identification

Posted on:2020-06-21Degree:MasterType:Thesis
Country:ChinaCandidate:J J ZhangFull Text:PDF
GTID:2392330578465109Subject:Engineering
Abstract/Summary:PDF Full Text Request
High voltage transmission mode is the inevitable trend of power grid development in the future.The autotransformer is not only efficient,but also easy to transport and install.Widely used in substations above 220 kV and above,it becomes the main transformer form of high voltage and above power network.Because of the special structure of autotransformer,the probability of short-circuit impact damage is higher than that of ordinary power transformer.If there is a weak link inside the winding,it will lead to winding deformation.Therefore,it is very important to monitor the winding deformation of autotransformer timely and accurately to avoid malignant faults.The main framework of this paper is as follows:The 2D and 3D finite element models of the autotransformer are built respectively.The parameters of windings are calculated by macro command.Compared with the theoretical value,within the allowable range of errors,the two-dimensional modeling method is more suitable for engineering applications.The inductance and capacitance parameters of windings are solved by 2D.Analyzing the sensitivity of parameters,the parameters suitable for winding deformation analysis are obtained for subsequent analysis.The results show that leakage inductance and capacitance are sensitive to winding deformation.Based on the analysis of the equivalent circuit of the winding of the auto-transformer,the parameter identification model based on the leakage inductance and equivalent capacitance of the winding is established.The voltage and current of the winding port are input.The electrical parameters of winding are identified by the integral form of least square algorithm.The results were used to establish a database for the winding deformation classification and degree detection.The winding electrical parameters identified on line are taken as characteristic quantities.The support vector classification and regression model for winding deformation of autotransformer under various deformation types are constructed.The corresponding relationship between the electrical parameters and the deformation degree of the winding under different deformation types is obtained.The on-line monitoring of the deformation of the autotransformer is realized.Results show that the accuracy of classification and regression model is higher.It is suitable for on-line monitoring of winding deformation of autotransformer.
Keywords/Search Tags:autotransformer, winding deformation, parameter identification, support vector machine(SVM), support vector regression(SVR)
PDF Full Text Request
Related items