基于人工神经网络的大量程光纤实时距离干涉测量仪

提出了一种基于人工神经网络的全光纤化大量程实时距离干涉测量仪.采用双正弦相位调制方法,即通过同时调制半导体激光器的波长和干涉仪的光程差实现外差测量.为了扩大干涉仪的测量范围和消除输出信号中的交叉敏感,采用人工神经网络进行信号处理,把两路经过初步解调的干涉信号作为输入样本,物体距离的实际值作为输出样本,对神经网络进行训练,以使其具有良好的推广能力.实验结果表明神经网络的使用不仅扩大了距离的测量范围而且提高了测量精度.

Fiber Optic Interferometer for Real-time Wide-range Distance Measurements with ANN

A novel optic fiber interferometer for distance measurements over a wide measurement range using an artificial neural network is designed. The optical phase of the interferometer is simultaneously modulated with a frequency modulated laser diode and optical path-length difference. To enlarge the measurement range and eliminate the influence of the cross-sensitivity, the interferometer readings are used as input data and the actual sensed measuring value as desired output data during the training process of the network model. The principle and signal processing system containing the ANN are described. The experimental results show the use of ANN could widen the measurement range and enhance the measurement accuracy.