The Thermal Modeling for Underground Cable Based on ANN Prediction

Abstract

AbstractMany factors affect the ampacity of the underground cable (UC) to carry current, such as the backfill material (classical, thermal, or a combination thereof) and the depth at which it is buried. Moreover, the thermal of the UC is an effective element in the performance and effectiveness of the UC. However, it is difficult to find thermal modeling and prediction in the UC under the influence of many parameters such as soil resistivity \(({{\varvec{\rho}}}_{\text{soil }})\), insulator resistivity \(({{\varvec{\rho}}}_{\text{insulator }})\), and ambient temperature. In this paper, the calculation of the UC steady-state rating current is the most important part of the cable installation design. This paper also applied an artificial neural network (ANN) to develop and predict for 33 kV UC rating models. The proposed system was built by using the MATLAB package. The ANN-based UC rating is achieves the best performance and prediction for the UC rating current. The performance of the proposed model is superior to other models. The experiment was conducted with 200 epochs. The proposed model achieved high performance with low MSE (0.137) and the regression curve gives an excellent performance (0.99).KeywordsUnderground cables performanceCable ampacityThermal modelingThermal backfillHeat transferArtificial neural network (ANN)