Research on data fusion of power wireless sensor networks based on Kalman filter

Article type: Research Article

Authors: Wang, Haoran^{; *}

Affiliations: Information Center of Guizhou Power Grid Co., Ltd., Guiyang 550002, China

Correspondence: [*] Corresponding author. E-mail: youwei@36haojie.com.

Abstract: In order to overcome the problems existing in traditional methods such as large mean error and long time of network data fusion, a data fusion of power wireless sensor networks based on Kalman filter is proposed. Firstly, the composition of power wireless sensor is analyzed, and the data of power wireless sensor network is preprocessed. Then, the data fusion process of Kalman filter is designed, and the schematic diagram of the data fusion process is given. Finally, l-M method is used to modify the network data fusion prediction covariance matrix to realize the power wireless sensor network data fusion. Experimental results show that when the amount of data is 600 GB, the data fusion time of the proposed method is 1.89 s. When the number of Kalman recursion is 120, the mean square error of data fusion of the proposed method is 0.04, and the practical application effect is good.

Keywords: Kalman filter, power, wireless sensor network, data fusion

DOI: 10.3233/WEB-220035

Journal: Web Intelligence, vol. 21, no. 2, pp. 103-114, 2023