Multi-Objective Design of Filter Installed in Brush Motor by Artificial Neural Network Accounting for Cable Length

In the design process of automotive products, it is often necessary to find solutions that simultaneously satisfy multi-objective performance goals, which can sometimes include requirements that conflict. Such redundant solutions are expected to cover a wider feasible range of design parameters and meet an assortment of different lead time and price goals. In this work, we apply an artificial neural network (ANN)-based machine learning algorithm to determine the cable length and design ranges of an EMI filter for an automotive-brush-motor system. We were able to find at least three interval solutions that satisfy the performance requirements, including a single interval solution obtained by our previous approach using Preference Set-based Design.