ZENG Jian-guo, CHEN Guang-meng, WANG Jian. New signal processing method for thermocouple based on RBF neural network[J]. 2007, (1): 36-38. DOI: 10.13873/j.1000-97872007.01.012.
基于RBF神经网络的热电偶信号处理新方法
摘要
针对热电偶信号处理中的非线性校正和冷端补偿等突出问题
利用径向基函数(RBF)神经网络构造双输入单输出的网络模型
并采用遗传算法对网络结构和参数进行优化训练
同时完成了热电偶测温中的非线性校正和冷端补偿。经仿真实验证明:该方法的测量误差减小至0.095%
在较大范围内提高了热电偶温度测量的精度。
Abstract
A method is presented to compensate non-linearity and cold-side-offset for signal processing of thermocouple.A network model with two inputs and single output is constructed by radial basis function(RBF) neural network(NN)
which is trained by genetic algorithm.Under the NN model
non-linearity compensation and cold-side-offset adjustment of thermocouple are realized simultaneously.The simulation experimenls show that the testing error of this method is 0.095 %