WIT Press


The Technique Of Mine Gas Monitoring Based On Multi-sensor Information Fusion

Price

Free (open access)

Paper DOI

10.2495/SAFE050441

Volume

82

Pages

8

Published

2005

Size

494 kb

Author(s)

L. Shao & H. Fu

Abstract

The problem of factors influencing the sensors in a mine gas monitoring system cannot be solved by a single sensor itself. The multi-sensor information fusion technique was put forward to increase the system’s information utilance, precision, dependability and error tolerance. Based on the information source that the mine gas monitoring system has collected, the system’s structure mode was defined, the structure mode was described by the state space method, and the multi-sensor information fusion state space model of the mine gas monitoring system was also built. A fuzzy neural network model for the multi-sensor information fusion was built by adopting the artificial intelligence method. Keyword: multi-sensor information fusion, gas monitoring system in coal main, fuzzy neural network. 1 Introduction Currently, there are two kinds of systems used for mine gas monitoring in our country. The centralizing and distributing monitoring system is based on standard analog signal transmitting, and the distributed gas monitoring system is based on serial communication technology [1]. But they both adopt the output signal of a single kind of sensor directly as the system’s collected information. Because the monitoring system works over a long period of time without breaking and the work environment of the coalmine bottom is comparatively bad, then various interferences would influence the measured value seriously. How to increase the measurement accuracy is the key for the whole monitoring system. Traditionally, a digital filtering method (such as seeking the arithmetic average value) was adopted to overcome the interference in the monitoring system. But when the sensor failure causes none information acquired, this

Keywords

multi-sensor information fusion, gas monitoring system in coal main, fuzzy neural network.