Investigation Of Identifying The Heat Transfer Coefficient Of Wall Based On The ANN Method

It is great of significance to accelerate to carry out the work of buildings energy saving for solving the increasing energy crisis. The heat transfer coefficient of building envelop is one of the key indexes for the energy saving buildings. The on-the-spot detection method of heat transfer coefficient of wall is an important technique in the evaluation of energy saving buildings. It is not only the last pass to ensure the quality of energy saving buildings, but also the powerful guarantee for implementing the standard of energy saving buildings.Firstly, the advantages and disadvantages of the existing on-the-spot detection methods were compared in the thesis, especially the limitation of the Heat Flow Meter method in hot summer and cold winter area, on the basis of the research in our group. A further research for heat source method with constant heat flux was also carried out. The variation rule of wall in the process of three-dimension heat transfer was analyzed and obtained the key input parameters with the ANN method to gain the heat transfer coefficient of wall. In addition, how to establish the appropriate sample in the ANN method was investigated. The sample data base would be set up for difference structure walls. Furthermore, four types of exterior heat-insulation wall were tested. The detection results have a good repeatability and agree with the theoretical value of wall, which proved the reliability and repeatability of the on-the-spot detection method. The program of numerical computation for the heat transfer process of wall was improved. Meanwhile, the author finished the integrative program of the ANN method which was used to solve the heat transfer coefficient of wall. Moreover, the analysis of error with the detection system was performed. The sources of error in the detection process were given. The expressions of sensitivity coefficient to show effect of detected arguments with detected result was obtained. The variation rule of the sensitivity coefficient was analyzed in succession. The optimal detection extents of the detection variables were also given. Finally, the on-the-spot detection of the practical project was carried out in the actual project. Comparing with the Heat Flow Meter method, this method has advantages of shorter...