Gray-box modeling of the vehicle cabin and comparison of model accuracy using different zonal distributions

Thermal models are essential for the climate control system of vehicles, as they provide crucial information on the occupants’ thermal comfort and help optimize the heating, ventilation, and air conditioning (HVAC) system performance. A control-oriented cabin model should deliver accurate predictions while ensuring fast computational performance. With increasing expectations for thermal comfort, the predictability of zonal temperatures has become an essential aspect. This paper established various cabin models (mono-zone, 4-zone, and 8-zone model) using the gray-box modeling method. The parameters of these models were identified using measurements from real-world test drives based on available vehicle sensors and ten additional air temperature sensors. The goal is to develop a suitable cabin model and to determine a reasonable zonal distribution. Three model variants yielded promising results, with the average coefficient of determination (R²) calculated as 0.83 (mono-zone), 0.75 (4-zone), and 0.74 (8-zone).