In the optimization design of the traction motor, performance such as torque and power is mainly selected as an objective function. Recently, in the electric drive system, the driving mileage is a very important item due to limited batter capacity, an...
In the optimization design of the traction motor, performance such as torque and power is mainly selected as an objective function. Recently, in the electric drive system, the driving mileage is a very important item due to limited batter capacity, and for this reason, it is necessary to design the traction motor in consideration of energy efficiency. However, in order to calculate the energy efficiency, it is necessary to calculate the motor losses at all operating points of the driving cycle, which requires a lot of calculation time for optimal design. In this study, in order to simplify the calculation of the energy efficiency of the motor, two methods for selecting representative points that can replace the entire driving cycle (WMTC mode) were proposed and the accuracy was verified. As the first method, the representative point selection using frequency weight at each division which was obtained division optimization was proposed, and as the second method, the K-Means clustering machine learning algorithm was proposed. In addition, 2 proposed methods were validated and the accuracy were checked with the values that were calculated by only the selected representative operating points. The research was conducted on the traction motor for EV2Wheeler and carried out through FEA (Finite Element Analysis) electromagnetic field analysis tool, optimization tool and Simulink modeling.