关键词： PMV, 热舒适度, Elman神经网络, 预测模型

Abstract: The traditional numerical calculation method of PMV has the defects of high computational complexity and large time delay.In this paper,according to the time-varying characteristic of PMV index,PMV prediction model is established based on Elman neural network and the on-line monitoring of thermal comfort is realized.The temperature,air velocity,relative humidity and mean radiant temperature are selected as the inputs of the prediction model and the PMV value is assigned as output.The prediction model has good generalization capacity.Simulation results show that the predictive results of the proposed method are in agreement with the results of numerical calculation;meanwhile the computation time of the proposed method is superior to that of the traditional method after the Elman neural network is trained sufficiently.

Key words: predicted mean vote, thermal comfort level, Elman neural network, prediction model