节点文献
小波神经网络在多波长辐射测温中的应用
Application of Wavelet Neural Networks to Multi-wavelength Radiation Thermometry
【摘要】 辐射测温的主要问题是解决未知的或变化的表面发射率所引起的误差测量。提出了一种小波神经网络,并对算法进行了优化,用以解决低温目标真温与辐射量之间的映射关系,有效地从辐射信息中分离出发射率和真温,给出了具体的算法。通过仿真实验证明此方法是获知真温的一种较好的方法。
【Abstract】 The main problem in radiation thermometry is how to minimize the measuring error resulting from unknown or variant surface emissivity. A kind of wavelet neural network is designed to get the mapping relation between the true temperature and radiation can be effectively separated from emission information. The detailed algorithm is presented. With proper optimization through the simulation experiment, this method is proved to be a better method to get true temperature.
【关键词】 计量学;
小波神经网络;
拟牛顿算法;
辐射测温;
真温;
【Key words】 Metrology; Wavelet neural networks; BFGS; Radiation thermometry; True temperature;
【Key words】 Metrology; Wavelet neural networks; BFGS; Radiation thermometry; True temperature;
【基金】 国家自然科学基金(6977020);哈尔滨工业大学基金(HIT.2002.22)
- 【文献出处】 计量学报 ,Acta Metrologica Sinica , 编辑部邮箱 ,2003年04期
- 【分类号】TP183
- 【被引频次】16
- 【下载频次】185