KoKliFaM – Comfort-based climate control of vehicle occupants using machine-learning algorithmsCopyright: © e3D
In current vehicle architectures, the use of decentralized climate control systems is currently of minor interest, since conventional combustion engines generate sufficient process heat to ensure the heating of the vehicle cabin in winter. In this context, the interior temperature of the vehicle cabin is predominantly used as the controlled variable, without taking into account the actual "thermal comfort" of the occupants. "Thermal comfort" results from a combination of multiple impacts and cannot be considered as an isolated entity. For example, parameters from the fields of sociopsychology, thermophysiology and ergonomics play a corresponding role here.
Efficient algorithms from the field of machine learning (ML) are suitable for working out such interindividual differences of people in the context of "thermal comfort" in the vehicle and for optimizing existing air conditioning concepts. The aim of the project is to cover as many aspects of the "thermal comfort" of a human individual by use of algorithms to be developed without the use of concrete "thermal comfort models" such as the "Predicted Mean Vote" (PMV) described in ISO 7730 or “Mean Thermal Vote” (MTV) described in ISO 14505-2.
12/2018 - 11/2019