Research on data fusion of surrounding rock mass in mine based on support degree and adaptive weighted

Abstract

In order to improve the level of coal mine safety inspection and ensure safety production, a dynamic perception model based on Internet of things is set up in view of uncertain areas and time of surrounding rock mass catastrophe. Observations from various types of different sensors are combined to provide a robust and complete description of surrounding rock mass in mine. Moreover, a multi-sensor data fusion algorithm based on the theory of support and adaptive weighted is proposed. Firstly, the consistency of data is checked and the outlier data is eliminated. Then, the mutual support degree is calculated and the low calculation data are replaced. Finally, the adaptive weighted algorithm is used to estimate the final parameters. Thus, the total mean square error of target parameter is minimum. The algorithm improves the accuracy of the fusion data free from any priori conditions. Experimental results verify the accuracy of the proposed algorithm and achieve the desired effects.