LIU Gang, LIU Xue-ren, JI Ying-hua, et al. Nonlinear correction of methane sensor based on improved BP neural network[J]. 2007, (1): 15-17+20. DOI: 10.13873/j.1000-97872007.01.005.
基于改进型BP神经网络的瓦斯传感器的非线性校正
摘要
提出了一种基于改进型BP神经网络的瓦斯传感器的非线性校正方法
利用神经网络良好的非线性映射能力
逼近反非线性函数完成非线行校正。仿真实验结果表明:与传统的分段线性与BP算法相比
改进型的BP神经网络收敛速度快、逼近精度高
准确度由原来分段线性校正的±5.020%提高到现在的±0.130%
且易于动态调校。
Abstract
The nonlinear correction method of methane sensor based on improved BP neural network is introduced to approach inverse nonlinear function by use of nonlinear mapping ability of neural network.The experimental results show that network-learning speed can be sped up markedly and nonlinear precision of the sensor is(±0.130 %) nonlinegr precision of classic paragraph algorithm and BP algorithm is(5.020 %).