Problems regarding the electromagnetic interferences between high voltage power lines and underground metallic pipelines are studied. To evaluate the magnetic vector potential for different constructive geometries of a specific interference problem, a neural networks based artificial intelligence technique is implemented. To find the optimal solution, different neural network architectures are tested. Results gained with neural networks are compared to finite element solutions considered as standard ones.

Keywords: Electromagnetic interference, High voltage power lines, Artificial intelligence, Neural networks, Underground metallic pipelines