基于神经网络的分振幅光偏振仪的数据处理

分振幅光偏振仪(DOAP)是一种高速测量光波偏振态的传感器。提出了一种基于人工神经网络(ANN)的分振幅光偏振仪的数据处理方法,将分振幅光偏振仪中电路系统输出的电信号作为神经网络的输入,入射光的斯托克斯参数作为神经网络的输出,建立一个前向多层神经网络模型。通过网络训练,使该网络确立了电路系统输出电信号与入射光斯托克斯参数之间的映射关系。由测量时得到的电信号,利用训练后的神经网络可以计算出待测的入射光的斯托克斯参数。测试结果表明,在测量精度方面,该方法获得的斯托克斯参数的总均方根偏差为1.9%,略优于基于矩阵运算的数据处理方法。

Data Processing Method for the Division-of-Amplitude Photopolarimeter Based on an Artificial Neural Network

The division-of-amplitude photopolarimeter (DOAP) is a sensor that can rapidly determine the polarization state of the incident light. A data processing method for the DOAP based on an artificial neural network (ANN) was presented. A multilayer feedforward neural network model was set up whose inputs are the electrical signals produced by an electronics system of the DOAP, and outputs are the Stokes parameters of the incident light. The mapping relationships between the electrical signals and the Stokes parameters can be determined by training the neural network. After the electrical signals were measured, the Stokes parameters of the incident light can be calculated via the neural network which has been trained. The total root-mean-square deviation of Stokes parameters is 1.9%.The testing results show that the data processing method based on the neural network is slightly better than that based on the matrix operation on the aspects of measuring precision.