The Role of Temperature Prediction Models in Energy Consumption for BEV Cooling Systems

Battery electric vehicles (BEVs) are gaining popularity in the automotive industry as they play a crucial role in achieving zero-emission targets set by upcoming regulatory legislation. BEVs not only offer environmental benefits but also demonstrate significantly higher efficiency compared to conventional vehicles when generating 1 kW of power [1]. Although BEVs are more efficient than conventional vehicles, ensuring the longevity and optimal performance of their components is critical. Vehicle thermal management systems (VTMS) are needed to be designed to maintain optimum operating temperatures, which directly influence battery life, efficiency, and energy consumption. Effective VTMS design helps to prevent overheating and ensures that all components operate within their ideal temperature ranges. This does not only extend battery life but also contributes to better overall vehicle performance, reducing the need for energy to cool down component’s temperature. A crucial part of an effective VTMS is predicting a temperature horizon of the system. A temperature prediction and an optimization model estimates the future temperature, or horizon temperature, of a system based on various specific inputs. These inputs can include the vehicle speed, route information from GPS, initial system temperature and environmental conditions. The predictive capability of the optimization model enables a proactive thermal management system control, ensuring that components like pumps and fans can adjust their operation points to maintain optimum system temperature.This paper compares the impact of a predictive temperature model on energy consumption and the impact of the accuracy of a semi-empirical coolant circuit model behind the predictive temperature model. The predictive temperature model estimates the time horizon of temperature of the coolant circuit, which requires certain inputs. A major input is the powertrain inlet temperature, calculated using a semi-empirical model that combines theoretical knowledge and empirical maps to estimate the temperatures of various components in the cooling circuit of the VTMS, such as the e-motor, battery, radiator, and DC-DC converter.