Reinforcement Learning Fuzzy Algorithm for Adaptive Cabin Management System With Application for Adaptive Interior Lighting

The purpose of this thesis was to design a framework for an adaptive cabin management system that is further explored through a study of light intensity. With the goal of passenger comfort, the system developed would adjust lighting to an average setting and further adapt to an individuals preference. A fuzzy inference system was implemented that utilizes DGI, the passengers age, chronotype and the activity on board to calculate a light intensity. A reinforcement system was then developed to tune the fuzzy inference system parameters (mean and output value K) utilizing a lighting override from the passenger. A cabin mock up was then setup to observe the correctness and e▯ectiveness of the system developed. Overall, the system designed tuned accurately and e▯ectively in all case scenarios tested with varying learning rates. This thesis was concluded with discussions of future work that could further improve the implementations within a cabin.