Hybrid Technique based Voltage Stabilizers for XLPE Cable Insulation
Abstract
Cross-Linked Polyethylene (XLPE) power cable is commonly used as the medium for transmitting power supply in the power system network. However, frequent exposure of XLPE cable to the high voltage causes its insulation to degrade over time leading to electricity outages. Electrical treeing is a commonly observed phenomenon associated with dielectric breakdown in solid dielectrics. Electrical treeing tends to exhibit a fractal shape with more branches at a relatively high discharge speed relative to the rate of damage. It establishes the fact that Partial Discharge (PD) activity in insulation highly depends upon certain factors like applied voltage, size of the void and dielectric constant of insulation material. The aim of the development model is to understand the effect of charges accumulation in the XLPE insulation void and voltage stabilization. Hence to accomplish the objective, composite of Elephant Herd Optimization (EHO) algorithm and Recurrent Neural Network (RNN) algorithm is utilized, which controls the stability constraints of the XLPE cable insulation. However, the efficiency of the voltage stabilizers available today is low, as a load is needed to significantly increase the breakdown strength. Furthermore, the addition of a voltage stabilizer should be kept as low as possible in order not to influence other properties of the material. The new voltage controls presented in this paper have increased the electrical tree inception field. The simulation results from MATLAB/Simulink are validated and tested. The evaluation result of proposed methodology is compared with Whale Optimization Algorithm (WOA) with RNN and EHO algorithm.