Parameterizing External Factors Influencing E-motor Consumption

In recent times, a standard drive cycle is an excellent way to measure the electric range of EVs. This process is standardized and repeatable; however, it has the drawback of testing in a controlled environment with a low content of active functions. This can cause a significant variation in predicted range between drive cycles and actual on-road tests. This problem can be addressed via on-road testing and by using customer-based velocity cycles which involve more active functions. While on-road tests provide a more accurate idea of real-world consumption, the repeatability of these tests is low due to excessive randomness. External factors such as ambient temperature, driver behavior, traffic, terrain, altitude, and load conditions affect the vehicle during on-road testing, making repeatability difficult. No two measurements can have the same consumption even if done on the same road with the same vehicle due to these factors. This paper presents a machine learning-based method to parameterize the external factors affecting e-motor consumption. By parameterizing these factors, on-road test results are normalized and used for comparative studies. The paper details the data collection process, the parameterization of external factors using ML models for different driving scenarios and temperature ranges. The ML models are developed in MATLAB and can be reproduced in other tools. The merits and demerits of each ML model are discussed, along with ways to mitigate each external factor, making the testing procedure more robust and reliable. This helps make automobiles more energy efficient.