The Research Of Coal Rock Interface Recognition Method Based On Information Fusion And Neural Network

In order to mechanize and intelligent the coal mining face, the first thing to do is to solve the a uto system of shearer roller, which need to track the interface between coal and rock all the time. Hig hly reliable system for recognizing the coal rock interface not only can improve the work efficiency and economic benefits in the process of coal mining, but also improve the underground working con ditions, ensuring safe operation. Therefore, it is very important to study it.Currently the research methods used mostly for analysis and identification are using single sensor, not only recognition accuracy is not high but also the stability and reliability are very low for single sensor. So, in this paper, coal rock interface recognition method which is using existing sensors of coal winning machine and additional sensor which constitute a multi-sensor system of shearer, and combining the technique of information fusion and neural network technology is studied. According to the structure of coal winning machine and the structure of the roller, the instantaneous stress analysis was carried out on the shearer drum to determine the state of cutting parameter of the need to collect; by comparing the signal processing analysis, choose the parameters of the wavelet packet analysis for the collected signal processing; then identify information fusion processing model and the neural network structure of concrete, and the treatment on the neural network training is a good data input to allow for the new data for identification; finally, we use the theory of uncertainty assessment of its recognition results, to observe the system’s reliability. Through the study of the coal rock interface recognition system in this paper, it could lay a good foundation for the future realization of underground coal mining intelligence.