利用BP神经网络建立煤炭水分微波测量模型

On the Establishment of the Microwave Measurement Model of Coal Moisture with BP Neural Network

采用常规的多项式建立煤炭水分模型时,存在某些自变量因子选取不合理的问题,而逐步回归分析具有筛选作用显著的变量因子的能力。本文利用BP神经网络具有良好的非线性映射能力,分别建立的衰减法、相移法和双参量法的煤炭水分模型,并与逐步回归分析建立的模型相比,BP神经网络建立的三种模型具有更好的精度。

When the conventional polynomial is adopted to establish the model of coal moisture,there is the problem that some independent variable factor selection is not reasonable,and stepwise regression analysis has the ability to screen effective variable factors.Based on the fact that BP neural network has good nonlinear mapping ability,in the paper,attenuation and phase shift method and the method of double parameters of coal moisture model are respectively established,and compared with the model established by stepwise regression analysis,As it shows,the BP neural network can establish three models with better precision.