Passive electromagnetic interference (EMI) filters in power electronics must deliver
high electrical performance, such as insertion loss at non-ideal terminations, while
also complying with targets related to size, weight, and cost. To design EMI filters
which meet multiple conflicting criteria in today’s automotive industry, multi-objective
optimization (MOO) techniques can be employed. MOO typically utilizes some type of
nature-inspired metaheuristic to find optimal solutions, which form compromises between
each other. Among these methods, evolutionary algorithms are one of the most popular.
This paper discusses the application of evolutionary algorithms in the design and
topology optimization of EMI filters, highlighting their effectiveness in achieving
the necessary performance, and pointing out the possibilities presented by the approach.
