Personalized Air-Conditioning in Electric Vehicles Using Sensor Fusion and Model Predictive Control

Abstract

This paper proposes a system for personalized air conditioning in electric vehicles based on the real-time evaluation of individual thermal comfort and model-predictive control. The goal is to continuously adjust decentralized air conditioning actuators in such a way that an optimal level of thermal comfort is maintained for the occupant while minimizing the overall energy consumption of the entire system. The latter employs contact-less skin temperature measurements using thermal imaging in combination with face and body recognition [2], seat mounted temperature and humidity sensors as well as additional ambient air velocity, operative temperature, humidity and pressure sensors in order to establish sufficient sensor data coverage. The numerical human model MORPHEUS [1] is applied to predict the thermal state of body regions, which are hidden from the thermal camera [3]. In order to assess the individual’s state of thermal comfort, the simulated and measured data is combined and evaluated using a thermal comfort model that is applicable under inhomogeneous climatic conditions [4]. This information is subsequently used to derive control strategies for local air conditioning of the occupants. Local infrared heating panels are used to heat specific zones of the human body under cold conditions. Seat ventilation as well as fans mounted close to the body are used to cool specific zones under warm conditions. In addition, individual occupant feedback acquired through an electronic feedback system is used to further fine-tune the system with respect to the occupant’s needs. The transient nature of the system at hand allows an early identification of individual-specific thermal comfort trends, which enables a much more efficient way of energy balancing.