Percentage of Dissatisfied Model of Thermal Environment Based on Beta Distribution

Human thermal comfort is one of the important contents of environmental health. Thermal environment predicted percentage of dissatisfied (PPD) is a key index of thermal comfort evaluation. Predicted mean vote (PMV) is obtained by measuring environmental and body factors. Then, a prior distribution (pre-estimation results) of PPD forms by introducing the PMV-PPD model. Considering the characteristic of field survey data, beta distribution is introduced to express a likelihood function and a posterior distribution in Bayesian thermal comfort model. The posterior distribution of PPD can be calculated according to characteristics of beta distribution. Prediction results are expressed by a mathematical expectation of the corresponding probability distribution, which is called Bayesian percentage of dissatisfied (BPD). The results of the case study can verify the BPD model. The main results are concluded as follows. Both a variance of a prior distribution and the size of survey samples will affect prediction results. With the increasing of size of survey samples, the affection of the prior distribution on the prediction results will gradually weaken. Considering objective factors of a thermal environment, the model can be applied by fewer field survey samples, and can be applied to design and analysis of building thermal environment.