Application of on-line Adaptive Weighted Fusion in Mine Gas Measurement System

Abstract:

In mine environmental monitoring system, the concentration of mine gas is an important indicator. Aiming at the redundant information from multi-gas sensors in the measurement system, adaptive weighted fusion algorithm was presented. Using this algorithm, it was unnecessary to be aware of any pre-defined knowledge about these datas measured by the sensors. That the algorithm could adjust the fused sensor’s weight in time according to the variation in sensors’ variances makes the mean square error minimal. It was also proved theoretically that this fusion algorithm is linear and unbiased, in respect of the least mean square errors. Simulation results showed that this fusion algorithm is effective and the result of fused data is superior to the mean estimate algorithm in respect of accuracy and fault tolerance.

• Periodical: Applied Mechanics and Materials (Volumes 239-240)
• Pages: 1395-1398
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.239-240.1395
• Online since: December 2012
• Authors: Yan Ju Wang, Li Kun Yang, Yu Tian Wang
• Keywords: Adaptive Weighted Fusion, Mine-Gas, Multi-Sensor, Optimization, Weighted Factor

References

[1] Liao YanBiao. Fiber Optics. Tsinghua university publishing company.(2000)

[2] Wang Yutian. Optoelectronics and optical fiber technology. National defense industry publishing company.(2003)

[3] Richardson J M, March K A. Fusion of multisensor data [J] .The International Journal of Robotics Research (S0278-3649 ), 1988, 7 (6): 78-96

[4] Yang Xir rong, Luo Gure. Multisensor information fusion technology and its application.SPIE,2001,4414：449～454

[5] Zhai Yili, Dai Yisong. Acta Metrologica Sinica [J]. Study of Adaptive Weighted Fusion Estimated Algorithm of Multisensor Data.1998, 19(1):69-74