基于径向基函数网络多光谱辐射测温技术理论研究

在多光谱辐射测温技术中常需要假设光谱发射率及波长之间的数学模型。文中介绍了一种基于神经网络的多光谱辐射测温数据处理方法，代替了假设发射率模型的方法。利用径向基函数网络，可由网络的输出直接得到目标的真温和光谱发射率，并分析了各种测试情形对测结果的影响。计算机仿真结构表明此方法是一种比较好的获知真温与发射率的方法．

Theoretical Study of Multi-wavelength Radiation Thermometry based on RBF Neural Network

In general,the assumption of mathematical models between spectral emissivity and wavelength is re- quired in radiation thermometry.In this paper the data process method based on the neural network is presented instead of the assumption approach.By taking advantages of the RBF network the true temperature and spectral emissivity are available simultaneously.The effects of measurement conditions on calculated temperature and e- missivity are also analyzed.Computer simulation results proved the theory to be an effective method to calculate both temperature and emissivity.