Skip to Main content Skip to Navigation

Eco-driving strategy for electric motorcycles

Abstract : The eco-driving strategies are dedicated approaches based on algorithms capable to use external and internal vehicle data to create recommendations and/or limitations over the driver or to generate automatically a complete reference to be followed in the case of autonomous vehicle. They let to reduce the energy consummation and to limit the pollution emissions, but currently, their usage is not directly related to the autonomy and the performance required by the driver in real time. In this context, this thesis work proposes an Eco-Driving strategy suitable for electric motorcycle with usage limitations. In fact, this strategy uses an optimal controller able to make an online optimization process. This controller is oriented to ensure that the energy available is enough to complete a demanded trip and to adapt the speed profile according to the usage requirements and the energetic constraints. The developed strategy integrates dynamic models carrying out an optimization under multi-physic constraints (electric, mechanic, thermal …etc) present in the main elements of the power chain: electric machine, power electronics converters and battery. Those models let to make a global energy behavior representation. The NMPC optimization technic (non-linear model predictive controller) is supported on a non-linear mechanical model joined to a weight unidimensional non-linear optimization. This means that the NMPC cost function weights can evolve in function of energy behavior and external constraints. This multi-objective cost function integrates (in a progressive way) different behaviors that related to the energy consumption classified by a weighting sensitivity study. This process lets to find a good balance between the optimization goal and the compilation time to satisfy real-time needs. The developed approach has been validated in simulation and in experimental with a reduced scale developed test bench. The validation shows that the algorithm is able to increase around 20% the autonomy with a maximum limit of 30% of the speed and acceleration for the strict usage cases (dynamic driving). In addition, the algorithm is capable to ensure completing the travel in the 98% of cases with a distance error lower than 1.5% in the presence of sensors and actuator noises.
Complete list of metadata
Contributor : ABES STAR :  Contact
Submitted on : Wednesday, December 1, 2021 - 1:01:52 AM
Last modification on : Wednesday, December 1, 2021 - 3:04:49 AM
Long-term archiving on: : Wednesday, March 2, 2022 - 6:21:13 PM


Version validated by the jury (STAR)


  • HAL Id : tel-03459430, version 1



Cristhian Yesid Bello Ceferino. Eco-driving strategy for electric motorcycles. Electronics. Université Paris-Saclay; Pontificia universidad javeriana (Bogotá), 2020. English. ⟨NNT : 2020UPAST072⟩. ⟨tel-03459430⟩



Record views


Files downloads