Wireless sensor network node classification fusion method based on perceptual attributes

The invention discloses a wireless sensor network node classification fusion method based on perceptual attributes. The method comprises the following steps of: firstly, performing dimensionless processing on measured data of nodes; performing dimension reduction on the processed data to obtain w-dimension data; determining comprehensive perceptual attributes of all the nodes; selecting a plurality of center nodes through an elitist strategy based on the comprehensive perceptual attributes; determining slave nodes of the center nodes; and lastly, determining a fused node set and performing data fusion. Through adoption of the method, the nodes can be classified adaptively according to the comprehensive perceptual attributes of the real-time measured data of the nodes, so that the fusion efficiency is increased. The nodes with the same comprehensive perceptual attributes are fused, so that interference data are reduced; the fusion accuracy and the information acquisition accuracy are increased; the fusion erro