A neural network approach for the inductive coupling between overhead power lines and nearby metallic pipelines

Abstract

An artificial intelligence (AI) based approach has been applied in order to investigate the electromagnetic interference problems between high voltage overhead power lines (HV OPL) and nearby underground metallic pipelines (MP). The implemented artificial neural network (ANN) solution evaluates the inductive coupling matrix describing the OPL-MP electromagnetic interference problem in case of different problem geometries and multi-layer soil structures. The ANN provided results were compared to data obtained through a finite element method (FEM) based analysis, considered as reference. This artificial intelligence technique, proposed by the authors, has the advantage of a simplified mathematical solver compared to FEM, and implicitly a lower required computing time. Finally the ANN provided inductive coupling data was used to evaluate the induced AC currents and voltages induced in an underground gas pipeline.