Papers by Author: Jin Xu

Paper TitlePage

Authors: Cheng Wen Yan, Jian Yao, Jin Xu
Abstract: In the present study a GUI tool for the prediction of building energy performance based on a three-layered BP neural network and MATLAB was developed. The inputs for this tool are the 18 building envelope parameters. The outputs are building heating, cooling and total energy consumptions and the energy saving rate. Compared with the complicated mathematical equations, this tool provides a very easy and effective method for students to learn the effects of building envelope performance parameters on the building energy performance. Thus, this tool can be used in building physics and building energy efficiency courses for the design of energy efficient building.
Authors: Jian Yao, Jin Xu
Abstract: To compare the indoor thermal environment under different building envelope constructions, a Matlab-based tool was presented for building envelope performance simulation. An application study of two cases illustrates energy efficient buildings can provide more suitable indoor environment than non-energy efficient buildings in cold winter and hot summer. In conclusion, this paper provides a new and fast way for the prediction of indoor thermal environment.
Authors: Jian Yao, Jin Xu
Abstract: In this paper we propose a Monte-Carlo method for the simulation of the angle-dependent light transmittance of thermotropic material. The results show that the scattering light increased as temperature rose, and most of light transmitted went through the sample of thermotropic material at the angles between 10~40 deg. The results also indicate that the light transmittance measurement of thermotropic material by spectrophotometer without an integrating sphere is not accurate. As a conclusion, Monte Carlo simulation is an effective method for the determination of angle-dependent light transmittance of thermotropic material, and results of these simulations can be used to calculate the shading coefficient of window for building energy efficiency.
Authors: Jian Yao, Jin Xu
Abstract: In view of the problem that it is difficult to calculate the Fanger’s PMV equation due to its complicated iterative process, a backpropagation neural network (BPNN) model was built to predict PMV. Air temperature, relative humidity, mean radiant temperature, air velocity, metabolic rate and clothing index were used as the input of neural network and PMV output as the output of the neural network. The results show that this prediction approach is very effective and has higher accuracy absolute error below 5%. As a conclusion, this study has a real significance, because it gives a new method with reliability and accuracy in the prediction of PMV.
Showing 1 to 4 of 4 Paper Titles