Machine Learning Based Variation Studies for Electric Vehicle Drive Unit Virtual System Models

<div class="section abstract"><div class="htmlview paragraph">In electric vehicles development, manufacturing variations pose big challenge in designing various mechanical components as these variations directly impact various customer perceivable performance outputs. If the manufacturing variations can be included in design phase itself, overall robustness of the design can be enhanced.</div><div class="htmlview paragraph">This paper delineates machine learning based methods to include manufacturing variations in designing drive units for electric vehicles. In an electric vehicle, the drive unit transfers power or torque from a battery through an inverter to wheels. The drive units are subjected to different types of loads under various vehicle maneuvers. To evaluate the drive unit system virtually, system level simulations are performed. Traditionally, nominal values of the several inputs such as bearing parameters, gear parameters and clearances etc. are used. However, the drive unit must be designed in such a way that outputs meet target considering all the variations of inputs. To assess this, large number of simulations are to be performed which is quite difficult and time consuming.</div><div class="htmlview paragraph">To address this challenge, machine learning offer alternative to traditional simulation techniques. Machine learning models can be developed and used to predict outputs for all the combinations of required variations in inputs with minimal use of simulation software and computation resources. This paper proposes methodology in which machine learning models are developed based on data generated by sampling. Different regression algorithms are used and evaluated statistically. Best algorithm is selected based on its performance metrics and used to predict outputs for any variations.</div></div>