English

应用科学学报 ›› 1999, Vol. 17 ›› Issue (1): 45-50.

• 论文 • 上一篇    下一篇

基于神经网络辨识的表面温度传感器动态标定及瞬态温度测量

周川1, 胡寿松1, 马勤弟2, 胡维礼2   

  1. 1. 南京航空航天大学;
    2. 南京理工大学
  • 收稿日期:1997-09-18 修回日期:1998-02-23 出版日期:1999-03-31 发布日期:1999-03-31
  • 基金资助:
    江苏省自然科学基金资助项目

A New Method of Surface Temperature Sensor Dynamic Calibration and Transient Temperature Measurement Based on Neural Networks Identification

ZHOU CHUAN1, HU SHOUSONG1, MA QINDI2, HU WEILI2   

  1. 1. Nanjing University of Aeronautics and Astronautics, Nanjing 210016;
    2. Nanjing University of Science and Technology, Nanjing 210014
  • Received:1997-09-18 Revised:1998-02-23 Online:1999-03-31 Published:1999-03-31

PDF

113

被引次数

7

摘要： 针对瞬态温度测量设计了一种薄膜热电偶,提出了一套瞬态温度传感器动态标定的实验方案.采用神经网络辨识的方法建立反映传感器动态特性的数学模型,通过多组测量数据对网络进行自动训练,获取传感器差分多项式模型的系数,实现传感器在线动态标定.并通过大量实验为解决瞬态温度测量这一难题提供了可靠的实验结果和新的理论方法.

关键词: 薄膜热电偶, 瞬态温度测量, 人工神经网络, 动态标定

Abstract: In this paper, a kind of thin film thermocouple is designed for transient temperature measurement, and a new experimental scheme of temperature transducer dynamic calibration is presented. The mathematical model of transducer has been acquired using artificial neural networks identification. By many groups of experimental data, networks are trained to realized accurate dynamic calibration of transducer. It has been shown that the method presented in this paper provides reliable experimental results and new theory for transient temperature measurement.

Key words: transient temperature measurement, artificial neural networks, thin film thermocouple, dynamic calibration