BP神经网络在小信号有效值非线性误差校正中的应用

BP Neural Network in the Application of Small Signal TRMS Nonlinear Error Correction

系统传输的非线性是产生测量误差的主要原因之一。为保证测量精度,应对非线性误差进行校正。描述了BP神经网络的基本模型及算法,阐述了神经网络校正非线性误差的基本原理,提出了用BP神经网络来校正交变小信号真有效值非线性误差的方法。最后给出了校正AD637非线性误差的实验数据。结果表明:运用BP神经网络校正非线性误差,提高了交变小信号真有效值的测量精度。

The nonlinear of system transmission is one of the major causes of measurement error. To ensure accuracy, the nonlinear error should be calibrated. In this paper,the basic BP neural network model and algorithm is described, the basic principle of nonlinear error correction using neural network is expatiated, and the method of alternating small signal TRMS nonlinear error correction by BP neural network is shown. Finally, the experimental data of nonlinear error correction of the AD637 show that the measurement accuracy of the alternating small signal TRMS is improved by BP neural network.