冯利军, 李竞生, 邵改群. 具有线性功能函数的神经元在矿井水质类型识别中的应用[J]. 煤田地质与勘探, 2002, 30(4): 35-37.
引用本文: 冯利军, 李竞生, 邵改群. 具有线性功能函数的神经元在矿井水质类型识别中的应用[J]. 煤田地质与勘探, 2002, 30(4): 35-37.
FENG Li-jun, LI Jing-sheng, SHAO Gai-qun. The application of Adaline in recognition of mine water quality types[J]. COAL GEOLOGY & EXPLORATION, 2002, 30(4): 35-37.
Citation: FENG Li-jun, LI Jing-sheng, SHAO Gai-qun. The application of Adaline in recognition of mine water quality types[J]. COAL GEOLOGY & EXPLORATION, 2002, 30(4): 35-37.

具有线性功能函数的神经元在矿井水质类型识别中的应用

The application of Adaline in recognition of mine water quality types

  • 摘要: 正确识别矿井水质类型对于判断突水水源具有重要的意义。本文以华北某矿井为例,采用具有线性功能函数的神经元(AdaptiveLinearElement)方法对来自煤层顶、底板含水层的两种不同水质类型矿井水进行了有效的识别。实际输入的4个未知水样中,经识别两个为顶板水,两个为底板水。神经元学习训练时,其收敛性、收敛速度与步长参数α的选取密切相关。此外,神经元本身也具有一定的抗噪性,这使得神经元在较小噪声背景下仍能正常工作。

     

    Abstract: It has an important significance to correctly identify mine water quality types in order to judge water inrush sources.Taking a certain coal mine in North China for example in this paper,two different water quality types of mine inflows from coal seam roof and floor aquifers have been effectively identified by applying Adaline method.Two of four unknown mine water samples are indetified as roof aquifer water,two as floor aquifer water.The convergence and its speed have a close relationship with the choice of step parameter α in the process of neuron learning and training.Besides,with smaller noise background,neuron itself is still able to normally work due to its anti-noise property.

     

/

返回文章
返回