Multi-Objective Optimization of Powertrain Components for Electric Vehicles Using a Two-Stage Analysis Model

Kihan Kwon,Minsik Seo,Seungjae Min
DOI: https://doi.org/10.1007/s12239-020-0141-5
2020-11-12
International Journal of Automotive Technology
Abstract:<p class="a-plus-plus">An electric vehicle (EV) powertrain is comprised of a motor and reduction gear. Thus, it must be designed by considering both components to improve its dynamic and economic performances. To obtain the optimal design of powertrain components for an EV, this study employs a two-stage analysis model focusing on the motor and vehicle at each stage for accuracy and efficiency. In the first stage, a motor system model analyzes the motor characteristics, such as the maximum and minimum torque and motor losses. Using the motor design parameters, these characteristics are converted to torque curves and an efficiency map. In the second stage, a vehicle system model analyzes the target performance using converted motor data for efficient analysis of the performance. An optimization problem is formulated to minimize the maximum motor power, acceleration time, and energy consumption with dynamic constraints, including the maximum vehicle speed and ascendable gradient. To reduce the excessive computational effort when conducting the multi-objective optimization, surrogate models with respect to performance are effectively constructed by using the adaptive sampling method. From the optimization results, a Pareto front having various solutions among the objective functions is obtained.</p>
engineering, mechanical,transportation science & technology
What problem does this paper attempt to address?