Abstract
This paper discusses comfort estimators to construct a control system that regulates thermal comfort of a passenger
in the conditions where disturbances exist. We consider two types of estimators; an observer-based estimator and an
estimator with a neural network. The observer-based estimator is designed on the basis of the model proposed in our
previous research. The neural-network-based estimator learns/estimates the relation between inputs and outputs
regarding the system as a black box. The performance of the control system with each estimator is validated with
simulations.
