Machine Learning Based Optimal Energy Storage Devices Selection Assistance for Vehicle Propulsion Systems 2020-01-0748
This study investigates the vehicle propulsion system energy storage devices selection. In recent years, powertrain electrification has been popular in all kinds of vehicles such as commercial vehicles and military utility vehicles. Energy storage devices are necessary for all levels of electrification. However, due to the large number of available energy storage devices (e.g. chemistry, size, energy density, and power density), and various class of vehicles (e.g. weight, range, acceleration, operating road environment), the energy storage devices selection process requires tremendous work if using traditional method. This study aims to assist the energy storage devices selection using the data sets collected from existing vehicles that equipped with energy storage devices. Machine Learning models are used to extract the relationship between the vehicles and the corresponding energy storage devices. After the training, the Machine Learning models can predict the ideal energy storage devices given the target vehicles design parameters as the inputs. The predicted ideal energy storage devices can be treated as the initial design and modification can be made based on the validation results. With the initial design by the Machine Learning models, the energy storage devices selection time and effort can be cut significantly. Dozens of vehicles public data are collected in the internet. In the training process, 80% of vehicles in the database are used. Another 20% vehicles data are used models validation process. As a results, the models predict the battery size and peak power with mean errors of 11.7% and 13.5%, respectively.
Bin Xu, Denise Rizzo, Simona Onori
Clemson University, US Army Ground Systems Ind Enterprise, Stanford University