This paper proposed a prediction method to predict a 10-kV oil-immersed transformer hot spot temperature (HST). A set of feature temperature points on the transformer iron shell is proposed based on fluid-thermal field calculation. These feature points, as well as transformer load rate, are taken as the input parameters of a machine learning model established by support vector regression (SVR), thus to describe their relationships with the HST. This model is trained by nine samples selected by L 9 (3 4 ) orthogonal array and applied to predict the HST of 20 test samples. The training samples are all obtained by simulation, and the test samples have consisted of simulation and transformer temperature rise test results. With effective parameter optimization of the SVR model, the predicted results agree well with the experimental and simulation data, the mean absolute percentage error (MAPE) is 1.55%, and the maximum temperature difference is less than 3 • C. The results validated the validity and the generalization performance of the prediction model.
INDEX TERMSHot spot temperature, oil-immersed transformer, support vector regression, multi-physical field analysis.
Abstract:The resistivity of oil impregnated paper will decrease during its aging process. This paper takes paper resistivity as an assessment index to evaluate the insulation condition of oil impregnated paper in power transformer. The feasibility of this method are discussed in two aspects: reliability and sensitivity. Iterative inversion of paper resistivity was combined with finite element simulation. Both the bisection method and Newton's method were used as iterative methods. After the analysis and comparison, Newton's method was selected as the first option of paper resistivity iteration for its faster convergence. In order to consider the spatial distribution characteristic of paper aging and enhance the calculation accuracy, the resistivity calculation is expanded to a multivariate iteration based on Newton's method, in order to consider the spatial distribution characteristic of paper aging and improve the calculation accuracy. This paper presents an exploratory research on condition assessment of oil impregnated paper insulation, and provides some reference to the security and economy operation of power transformers.
The low frequency section of frequency domain spectroscopy (FDS) of oil-paper insulation system can effectively reflect the ageing state and moisture of the oil-immersed paper. Meanwhile, steady-state insulation resistance is the one of indexes for the condition of transformer. In actual measurement, these two both are time-consuming and poor accessibility, thereby restraining their field applications. Aim at this problem, this paper analyses the polarization process of the oil-paper insulation in transformer, and adopts a low frequency equivalent circuit model for the oil-paper insulation to characterize the polarization process at low frequency. The simulated annealing particle swarm optimization (SAPSO) algorithm is proposed to identify the parameters in equivalent circuit, which are affected by change of insulation state. Compared with the experimental results, the errors of the steady-state insulation resistance and the tan δ below 10 −3 Hz, which are calculated by the identified parameters, are less than 2.5%. This method can avoid the influence from external environment and reduce the measurement time. The time and frequency domain parameters calculated can provide reference for the ageing assessment of oil-immersed paper.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.