2. Methodology
2.1 Experimental Campaign
Eleven experimental campaigns were carried out from November to December 2013 in two classrooms at
Department of Engineering (University of Perugia), in order to collect the input data for the training of the Networks;
all the environmental parameters were monitored by using DeltaOhm HD32.7 acquisition system; specifically the air
temperature (HP3217R), air velocity (AP3203), globethermometer temperature (TP3275), air pressure (HD32.7),
and relative humidity (AP3203) were acquired with an acquisition rate of 1 minute and for all the duration of
lessons (about 2 hours). The outdoor and indoor air temperature and relative humidity were also monitored by
using two TinyTag Ultra 2 TGU-4500, with the same acquisition rate (1 min). The adaptive approach for the
evaluation of the thermal comfort sensation was also used: after about 1 hour of experimental acquisition,
questionnaires were distributed to people attending the lesson. The main required parameters were: worn clothing,
gender, age, thermal sensation, position within the class.
The adopted thermal sensation scale in the questionnaires, shown in table 1, is the same proposed by Fanger and
then by ASHARE [12]. The position within the classroom was also asked in order to use it for the training of
Networks and to evaluate the variation of the thermal sensation within the environment.