Wavelet transform based noise reduction method for temperature data sequence in intelligent building

It is a kind of important basis for the maintenance of comfortable environment and the management of energy efficiency in intelligent building that the temperature and its variety at workspace in that building. Temperature in workspace can be sampled frequently and instantly with wireless sensor network. With some inherent electrical characteristics of sensors and sensor node as well as the complexity thermal feature of environment in intelligent building, there are noise in temperature data sampled by wireless sensor network. To acquire temperature in workspace accuracy, characteristic of temperature and its variety in intelligent building under ideal condition is analyzed and wavelet based noise reduction method for observation temperature data sequence is discussed in this paper. Wavelet function, wavelet transform scale and effective threshold method are confirmed. Experiments results show that wavelet transform based noise reduction method can reduce noise in observation temperature data sequence in intelligent building effectively.