引用本文: | 朱庆保.最佳拟合与神经网络相结合实现传感器特性线性化[J].控制理论与应用,2003,20(1):66~69.[点击复制] |
ZHU Qing-bao.Sensor's characteristics linearization using artificial neural networks and best-fitting[J].Control Theory and Technology,2003,20(1):66~69.[点击复制] |
|
最佳拟合与神经网络相结合实现传感器特性线性化 |
Sensor's characteristics linearization using artificial neural networks and best-fitting |
摘要点击 1651 全文点击 1855 投稿时间:2001-02-05 修订日期:2002-04-22 |
查看全文 查看/发表评论 下载PDF阅读器 |
DOI编号 |
2003,20(1):66-69 |
中文关键词 神经网络 最佳拟合 传感器 线性化 仿真 |
英文关键词 neural networks best-fitting sensor linearization simulation |
基金项目 |
|
中文摘要 |
提出了一种传感器特性线性化的方法.该方法把传感器特性分为线性和非线性段,用一种改进的BP神经网络映射传感器特性非线性段的反函数作为校正环节,用最佳拟合方法得到线性段的直线方程,从而实现传感器特性的线性化.经过仿真试用表明,这种方法可使传感器的非线性误差减小近十倍.最后,给出了一些仿真实验和仿真结果. |
英文摘要 |
A method of the sensor's characteristics linearization is proposed. The sensor characteristics are compartmentalized into linear and nonlinear segments, the inverse functions of nonlinear segment are mapped with the improved BP neural networks as correction links, and a straight line equation is produced with best-fitting in the linear interval. The results of simulations show that the nonlinear errors of the sensor are decreased approximately tenfold. Finally, an experiment is given. |