Home // ADAPTIVE 2018, The Tenth International Conference on Adaptive and Self-Adaptive Systems and Applications // View article

Improving Thermal Management of Electric Vehicles by Prediction of Thermal Disturbance Variables

Authors:
Peter Engel
Sebastian Meise
Andreas Rausch
Wilhelm Tegethoff Tegethoff

Keywords: Model Predictive Control; BEV; Applied Machine Learning; HVAC; Mobile Data Mining

Abstract:
In addition to the powertrain, heating and air-conditioning represents the second-largest energy consumer in electric vehicles. Optimization in this area can therefore contribute significantly to enhance the range of these vehicles. A new approach exploiting this optimization potential is the use of a model predictive controls. These controllers are based on a mathematical process model, which predicts the trajectories of the output variables. The predicted output variable trajectories are then evaluated by a non-linear cost function in order to find the corresponding optimal manipulated variable trajectory. Since external disturbances also affect the system in addition to manipulated variable, it is also necessary to predict these disturbances with sufficient precision. This is the core problem of this control approach and is not adequately addressed in previous approaches. For vehicle cabin heating and air-conditioning, the disturbances correspond to the thermal loads. These loads are mainly caused by the energy input of solar radiation, outside temperature, wind speed and humidity. In the following work, we will show how the coupling of methods of machine learning with Car2X technologies can lead to a high-precision prediction of thermal disturbances for an electric vehicle.

Pages: 75 to 83

Copyright: Copyright (c) IARIA, 2018

Publication date: February 18, 2018

Published in: conference

ISSN: 2308-4146

ISBN: 978-1-61208-610-1

Location: Barcelona, Spain

Dates: from February 18, 2018 to February 22, 2018