Research for multi-sensor data fusion based on Huffman tree clustering algorithm in greenhouses

Information monitoring system in the greenhouse will often deploy multiple redundant collection nodes in the same region in order to gather environmental parameters. It will improve data collection accuracy and enhance system stability through data fusion. This paper proposes a multi-sensor data fusion algorithm for greenhouse environment monitoring. First, triple exponential smoothing is applied to a variety of greenhouse environmental parameters. Then the smoothed data are sent to a router or gateway, removing invalid data, then idea of the Huffman tree is introduced to determine the order of the integration of each sensor, so as to realise multi-sensor data fusion. In the Institute for Farmland Irrigation Chinese Academy of Agricultural Sciences, the tests indicated that three exponential smoothing fluctuations significantly reduce the data to improve the accuracy of the data fusion, meanwhile, running time of data fusion algorithm was significantly shorter, and better than weighted average method.