Zırhlı yüksek gerilim kablolarında yalıtım arızalarını önlemek için çok amaçlı optimizasyon ve hibrit yapay zekâ tabanlı kablo topraklama yöntemi.

The metal sheath and armor on the insulating layer of a high voltage cable not only limit the electric field but also protect the cable's insulator against mechanical effects. However, when the load current flows through the cable, currents and voltages occur on the armor and metal sheath, which increase the cable temperature and cause dangerous voltages. These cause insulation failures. In the literature, different grounding methods are suggested to prevent insulation failures. However, due to the increase in the effect of harmonic currents in recent years, it is seen that these grounding methods are insufficient to prevent insulation faults. In this study, a new grounding method based on multi-purpose optimization and artificial intelligence is proposed to prevent insulation failures caused by high harmonic current and voltage. In order to make an optimum grounding with this method, it is necessary to know the voltage, current and current harmonic distortion on the metal sheath and armor. For the determination of these parameter values, estimation methods consisting of hybrid artificial neural networks and regression methods were used. Hybrid artificial bee colony-artificial neural network (H-NE) and gaussian process regression (GPR) methods were selected from these groups according to their minimum training errors and used as objective functions in optimization algorithms. Since there are many objectives in the optimization of the proposed grounding method, the multi-objective optimization method has been used. In multi-objective optimization, optimum values for voltage, current and current harmonics on metal parts are obtained when H-YAK is used as the estimation method. [ABSTRACT FROM AUTHOR]