Prediction of the Space Charge Related Maximum Electric Field of XLPE Insulation by Sparse Bayesian Learning

Space charge dynamics and the related breakdown failure are of great importance in HVDC cable insulation. Considerable works have been studied the space charge behavior and the mechanism underlying the charge transport and accumulation inside the cable insulating materials. However, the formation, accumulation, and dissipation of space charges are still complicated, which are affected by many factors, such as electric field, temperature, voltage reversal, material morphology, etc. Actually. At present, there is still a lack of research on the relationship model of this complex nonlinear network between space charge behavior and breakdown. The maximum electric field strength is extremely important in the study of space charge characteristics and insulation breakdown, which reflects the degree of distortion of the internal electric field. This paper establishes a breakdown prediction model based on the maximum electric field. First, the PEA method is used to measure the experimental results of the space charge distribution for a long time. Then, the sparse Bayesian learning is used employed to learn the mapping relationship between the maximum field strength and the relevant parameters, and t. After that, the prediction model is established. The breakdown conditions are discussed with reference to respect to the relevant studiesprevious work. This work paper is useful to evaluate the aging and predict the insulation failure by space charge effects in HVDC cable insulation.