摘要
为研究颗粒煤瓦斯解吸规律,基于Fick定律建立了颗粒煤的多扩散系数瓦斯解吸模型,完成了颗粒煤瓦斯解吸模型的数值试验。引入了非负约束最小二乘法反演算法(NNLS),通过试验数据反演得出颗粒煤的扩散参数的B谱,从而确定出颗粒煤瓦斯扩散系数D的准确范围。研究结果表明:颗粒煤瓦斯解吸符合Fick扩散定律,颗粒煤的多扩散系数瓦斯解吸模型能很好地解决单一扩散系数模型的扩散系数随时间衰减的问题,准确反映了颗粒煤瓦斯解吸规律,单一扩散系数瓦斯解吸模型只是多扩散系数瓦斯解吸模型的一个特例;NNLS是一种有效的反演算法,利用NNLS方法可以准确反演出颗粒煤瓦斯解吸过程中的扩散参数的B谱,通过B谱可方便计算出颗粒煤的瓦斯扩散系数。
In order to investigate the laws of gas desorption for coal particles , a multi-diffusion coefficient model of gas de-sorption for coal particles was proposed based on the Fick law .A series of numerical experiments on gas diffusion models for coal particles were carried out .The non-negative constrained least squares method ( NNLS) was introduced , and the B spec-trum of diffusion parameters for coal particles were obtained through the inversion of experimental data , then the accurate range of gas diffusion coefficient D for coal particles was determined .The results showed that the gas desorption of coal parti-cles accords with the Fick diffusion law .The multi-diffusion coefficient model of gas desorption for coal particles can solve the problem that the diffusion coefficient in single-diffusion coefficient model attenuates with time well , and it can accurately reflect the laws of gas desorption for coal particles .The single-diffusion coefficient model of gas desorption is only a special case in the multi-diffusion coefficient model of gas desorption .The NNLS method is an effective inversion algorithm by which the B spectrum of diffusion parameters in the gas desorption process of coal particles can be accurately retrieved , and the gas diffusion coefficient of coal particles can be easily calculated by using the B spectrum.
出处
《中国安全生产科学技术》
CAS
CSCD
北大核心
2016年第7期10-15,共6页
Journal of Safety Science and Technology
基金
国家自然科学基金项目(51574112
51304072)
河南省科技创新杰出青年基金项目(164100510013)
教育部科学技术研究重点项目(213022A)
河南省高等学校青年骨干教师资助计划资助项目(2013GGJS-050)
河南省基础与前沿技术研究计划项目(132300413203)
关键词
颗粒煤
瓦斯解吸
多扩散系数瓦斯解吸模型
非负约束最小二乘法
扩散系数反演
coal particles
gas desorption
multi-diffusion coefficient model of gas desorption
non-negative constrained least squares method
diffusion coefficient inversion