This article describes an application of Non-linear Model Predictive Control algorithms on energy efficient control of fully electric vehicle cabin temperature and air quality. Since fully electric vehicles can not utilize waste heat from a powertrain (or there is not enough waste heat) as ICE vehicles do, it is necessary to employ advanced control approaches (especially for cabin heating) due to the possible mileage lost by using energy from the batteries for cabin conditioning. The basic idea behind this is to avoid the heat losses caused by excessive air exchange and to ensure a satisfactory air quality in combination with a user defined temperature. The Non-linear Model Predictive control algorithms were successfully implemented into an Infineon AURIX Tricore microcontroller and tested within a Processor in the Loop simulation. Index Terms-non-linear model predictive control, fully electric vehicle, battery electric vehicle, vehicle cabin model, Extended Kalman filter, air quality control, temperature control NOMENCLATURE COP Coefficient

Vehicle thermal management systems of Fully Electric Vehicles bring increased demands on control algorithms to operate the vehicle efficiently. Especially, if there are multiple heat sources and sinks (cabin, batteries, electric drive, thermal energy storage, etc.), it is necessary to select the system operating mode (configuration of actuators), under which the system will operate efficiently with respecting defined constraints and references tracking. This paper brings a novel approach to the decision-making algorithm, which is based on the Hybrid Model Predictive Control and optimally solves the problem with regards to the defined objective function.

No abstract