CCTEG Fushun Testing Center Co., Ltd.
CCTEG Shenyang Research Institute
State Key Laboratory of Coal Mine Safety Technology

近年来TDLAS技术已成为实现高精度甲烷气体测量最重要的技术之一,但是压力的变化会影响甲烷气体的吸收系数,造成检测出现误差,因此设计高精度、稳定性好、易于实现的甲烷气体压力补偿方法是急需解决的关键问题。针对传统BP神经网络算法易于陷入局部极小化,局部优化效率低的问题,提出一种采用具有出色全局寻优力的GA遗传算法与BP神经网络融合的方法对甲烷气体测得数据进行压力补偿。试验结果表明,采用GA-BP算法得到的压力补偿效果明显优于采用传统的BP神经网络算法,具有更加优越的压力补偿精度,最终证明了使用该种方式的可行性,为甲烷气体检测时的压力补偿手段提供了参考价值。

In recent years, TDLAS technology has become one of the most important technologies for achievinghigh-precision methane gas measurement. However, changes in pressure can affect the absorption coefficient ofmethane gas, causing detection errors. Therefore, designing high-precision, stable, and easy to implement methanegas pressure compensation methods is a key issue that urgently needs to be solved. In response to the problem oftraditional BP neural network algorithms being prone to local minimization and low efficiency in local optimization.A method was proposed for pressure compensation of methane gas measurement data using a fusion of GA geneticalgorithm with excellent global optimization ability and BP neural network. The experimental results showed that thepressure compensation effect obtained by using GA-BP algorithm was significantly better than that obtained by usingtraditional BP neural network algorithm, with superior pressure compensation accuracy, which ultimately proved thefeasibility of using this method and provided reference value for pressure compensation methods in methane gasdetection.

*国家重点研发计划项目(2022YFF0605300);国家发改委实验室建设项目(发改投资[2019]704号)

李志福.基于GA-BP激光甲烷检测压力补偿方法的研究[J].煤矿机电,2024,45(4):1-5.