基于贝叶斯理论的热环境不满意率预测模型*
Prediction model of thermal environment dissatisfaction rate based on Bayesian theory
摘要：
热环境舒适性是环境健康的重要内容之一，而热环境满意率预测是热舒适评价的重要指标。通过预先测得的环境和人体参数，利用PMV-PPD模型形成热环境预测不满意率的先验分布。结合不满意率现场调研结果，应用贝叶斯公式和马尔可夫链蒙特卡罗方法计算不满意率的后验分布。算例结果证实了该方法的有效性。研究表明，不满意率先验分布方差和调研样本数量均会影响预测结果；随着样本数量的增加，先验分布对预测结果的影响会逐渐减小。该模型考虑了主、客观因素作用，适用于建筑热环境设计与分析。
Abstract：
Thermal environmental comfort is one of the important contents of environmental health, and the thermal environmental prediction satisfaction rate is an important index for thermal comfort evaluation. Through the pre-measured environmental and human body parameters, forms a priori distribution of the dissatisfaction rate of thermal environment prediction by PMV-PPD model. Combined with the results of field investigation of dissatisfaction rate, calculates the posterior distribution of dissatisfaction rate using Bayesian formula and Markov chain Monte Carlo method. The results of numerical examples demonstrate the effectiveness of the method. The study shows that the variance of dissatisfied priori distribution and the number of survey samples affect the prediction results. With the number of samples increasing, the influence of priori distribution on the prediction results gradually decreases. Considering the action of subjective and objective factors, the model is suitable for the design and analysis of building thermal environment.
Keywords:predictionmodel,Bayesiantheory,PMV-PPDmodel,fieldsurvey,thermalcomfort