Multisensor Correlation Analysis and its Application in Coal Mines

A multisensor system in coal mine is composed of a dozen even dozens of gas monitoring sensors. There is coupling in sensors of the gas monitoring systems. And it is interesting. The paper focuses on two topics about correlation analysis for multisensor system. One is about multisensor correlation analysis. Correlation information entropy and condition number of matrix were used. The other is about how to effective find out all combinations correlation sensors in a multisensor system. A theorem was proposed in this paper. Based on the theorem, an algorithm which can find out all combinations of correlative sensors from a multisensor system is proposed. Proposed algorithm can not only find out all combinations of correlation sensors effectively, but obtain the maximum combination of correlation sensors. Through experiment on a real gas monitoring dataset in coal mine, the results of experiments show that performances of proposed algorithm are excellent.