Predictive modeling for overall thermal sensation of vehicle occupants based on local thermal sensation when warming up

Abstract

At present, the existing predictive models of overall thermal sensation cannot accurately evaluate occupants’ feelings under vehicle cabin conditions. In this work, aiming at the overall thermal sensation evaluation of vehicle occupants in winter heating condition, the interrelations between the occupants’ local thermal sensations and the overall thermal sensation are analyzed using the experimental results of physical tests. It reveals that the overall thermal sensation is significantly affected by the local thermal sensation at extremities, which are called the extreme local thermal sensations. The overall thermal sensation evaluation model is established using the Support Vector Machine method, which is based on the maldistribution of the local thermal sensations, the extreme local thermal sensation, and the mean value of the local thermal sensations. The prediction accuracy of the training set and the verification set are 80% and 71.11% evidence the application potential of the Support Vector Machine evaluation model.