Comparisons of Models of Electric Drives for Electric Vehicles

Desreveaux, Anatole; Ruba, Mircea; Bouscayrol, Alain; Sirbu, Gabriel Mihai; Martis, Claudia · 2019 · Crossref

DOI: 10.1109/vppc46532.2019.8952540

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper addresses the need for efficient simulation tools to accelerate the development of electric vehicles (EVs), a critical step in reducing greenhouse gas emissions. The authors compare different modeling approaches for the electric drive (e-drive) of an EV, specifically evaluating the trade-offs between computational speed and accuracy in energy consumption predictions. The study focuses on the Renault Zoe, a commercial EV equipped with a 65 kW Permanent Magnet Synchronous Machine (PMSM). The research aims to determine if simplified static models can replace complex dynamic models for system-level analysis, such as energy management, without significant loss of fidelity. The methodology utilizes the Energetic Macroscopic Representation (EMR) formalism to organize and simulate the vehicle’s traction system, which includes the battery, e-drive, gearbox, wheels, and chassis. Three distinct e-drive models are implemented in Matlab/Simulink: a dynamic model serving as the reference, a static model assuming a constant efficiency of 90%, and a static model utilizing a detailed efficiency map. The battery is modeled with a series resistance and open-circuit voltage dependent on the state of charge. The simulations are conducted using two real-world driving cycles: an urban cycle (14 km, 40 minutes) and an extra-urban cycle (17.54 km, 12 minutes). Regenerative braking is limited to 50% of the total braking force in all simulations. The results demonstrate that static models significantly reduce computation time while maintaining high accuracy. For the urban cycle, the dynamic model required 244.66 seconds, whereas the static models with constant efficiency and efficiency maps required only 1.76 and 2.16 seconds, respectively. This represents a reduction in computation time by a factor of approximately 100. The energy consumption errors for the static models were 4% and 2%, respectively, compared to the dynamic reference. In the extra-urban cycle, the dynamic model took 137.38 seconds, while the static models took 1.16 and 1.47 seconds. The corresponding errors were 3.1% and 1.4%. The static model using the efficiency map consistently yielded lower errors than the constant efficiency model. The study concludes that static models are viable alternatives to dynamic models for EV simulation when the objective is energy analysis or system design rather than control scheme development. The use of an efficiency map in static models improves accuracy, halving the error compared to constant efficiency assumptions. These findings suggest that engineers can select simplified models to drastically speed up simulation processes, facilitating faster iteration and development of electrified vehicle systems without compromising the reliability of energy consumption estimates.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• simulator validity fidelity