The Research Of Information Fusion Method In The Building Environment Based On Neural Network

The building environment, specially the environment quality has been received more and more values. The matter of people's concern becomes the temperature in the work or the room, the humidity, the air quality, the noise level and so on. But the information sources in the building environment is complex and inconsistent, which effects the analysis, the appraisal, the control for the information of the building environmental. One of the efficient paths to reduce the building energy consumption and to enhance the building environment intelligent control level is that we can study one kind of information fusion processing method to analyze the building ecology and the energy conservation, and put the intelligent control to it, thus forms the one whole set to carry on the fusion, the appraisal, the control system to the building environmental information.This paper's main research content is in the system's information fusion method research. First, it studies the building environmental information. Due to the characteristic of the information sources in the building environment, it advances that using the information fusion method to analysis the information in building environment.Second, the paper researches the information fusion method of the building environment. It advances that the neural network can be used to analysis the information in building environment because of the strong learning capability ,the parallel distribution processing information, the auto-adapted and so on characteristics.At last, it has established the neural network model, has found the complex relations between the indoor air temperature, the indoor air relative humidity, the human body active mass (i.e. metabolism rate), the attire situation (i.e. clothes thermal resistance), average radiation temperature and wind speed and the human body to the air conditioning environment different demand (i.e. the indoor PMV value). The result indicates that the neural network can be used to analysis the information in building environment.