Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
The estimation of the lightning performance of a power distribution network is of great importance to design its protection system against lightning. An accurate evaluation of the number of lightning events that can create dangerous overvoltages requires a huge computational effort, as it implies the adoption of a Monte Carlo procedure. Such a procedure consists of generating many different random lightning events and calculating the corresponding overvoltages. The paper proposes a methodology to deal with the problem in two computationally efficient ways: (i) finding out the minimum number of Monte Carlo runs that lead to reliable results; and (ii) setting up a procedure that bypasses the lightning field-to-line coupling problem for each Monte Carlo run. The proposed approach is shown to provide results consistent with existing approaches while exhibiting superior Central Processing Unit (CPU) time performances.
Marcos Rubinstein, Carlo Alberto Nucci
Richard Lee Davis, Engin Walter Bumbacher, Jérôme Guillaume Brender