摘要
针对泡克耳斯(Pockels)效应光学电压传感器中温度补偿问题,在双光路硬件补偿的基础上,提出一种基于BP神经网络的软件补偿方案。该方案利用BP神经网络算法训练出在不同环境温度下传感器输出电压值与实际测量电压值之间的映射关系,并通过引入虚拟仪器技术在Labview软件开发平台中调用Matlab进行信号的处理与分析,提高了系统的可视性。仿真结果表明,系统采用神经网络温度补偿后,大大提高了光学电压传感器的稳定性和精确度。
To solve the temperature compensation problem in Pockels effect optical voltage sensor,a software compensation scheme based on BP neural network on the basis of double optical circuit hardware compensation is proposed.In this scheme,BP neural network algorithm is used to train the mapping relationship between the output voltage of sensor and the actual measured voltage under different ambient temperatures.By introducing virtual instrument technology to call MATLAB in Labview software for signal processing and analysis,the visibility of the system is improved.The simulation results show that the stability and accuracy of the optical voltage sensor are greatly improved by using neural network temperature compensation.
作者
黄艳
HUANG Yan(Electronic Engineering Department,Zhangzhou Institute of Technology,Zhangzhou 363000,China)
出处
《三明学院学报》
2019年第4期33-40,共8页
Journal of Sanming University
