Research On Heterogeneous Data Fusion Algorithm Based On Mixed Support Of And Rough Set

With the rapid development of Internet of Things(IOT) technology and smart city, heterogeneous data fusion technology attracts more and more people’s attention, and also shows great vitality in flight target tracking, route planning, environmental testing, decision support,intelligent transportation and other fields. Heterogeneous data fusion technology can take full advantage of the complementary between multi-sources information, so as to reflect the state information of the perception of objects more comprehensively and truly, which has become one of the core technology of IOT.In this paper, we make a deep and systematic research on heterogeneous data fusion techniques of the wireless sensor networks. Aiming at the problems existing in the fusion models of heterogeneous data, fusion accuracy and abnormal data processing, we put forward the corresponding solutions. By comparing algorithms and analyzing experiments, we prove the effectiveness and feasibility of the algorithm.First, regarding to the actual measured WSN(Wireless Sensor Network) sensing data, we analyze the influence of the temporal characteristics and abnormal data processing on fusion results from the perspective of data fusion. On this basis, this paper proposes data correction algorithms based on gray prediction. Meanwhile, starting from the three-dimensional structure of sensing data, we divide the wireless sensor network data fusion into two parts, of which the one is homogeneous data fusion, and the other one is heterogeneous data fusion. Then we propose a hierarchical model of heterogeneous data fusion.Secondly, aiming at the hierarchical model of heterogeneous data fusion, we propose a heterogeneous data fusion algorithm based on mixed support and rough set. The algorithm perceive the data into homogeneous data and heterogeneous data, aiming at the two types of data, we respectively propose fusion algorithms based on mixed support for homogeneous data and based on rough set for heterogeneous data, reducing the heterogeneous data complexity.Finally, using the actual measurement data obtained from the experimental environment, we do the simulation experiment on heterogeneous data fusion algorithm, and evaluate the fusion algorithm’s performance from fusion accuracy and the recognition rate two aspects. Experimental results show that the heterogeneous data fusion algorithm based on mixed support of and rough set in this paper, improve the accuracy of the fusion, while improving the accuracy of decision support.