模块化神经网络数据融合在谱线核测量中的应用

Modular Neural Network Data Fusion in the Nuclear Spectrum Measurement

针对谱峰随不同干扰量漂移的问题,尝试使用模块化神经网络的信息融合对谱仪谱峰随温度、电压等干扰漂移进行数据融合。对谱仪峰位漂移,建立了融合的网络模型、算法及其学习方式。最后在建立的数学模型上对谱漂移试验数据进行计算和仿真。实验结果表明建立的数学模型有较好的精度,基于模块化神经网络的融合技术能运用在谱仪的谱峰漂移的数据融合中。

With the different peaks against interference in the issue of drift, Modular neural network will be used for the peak drift spectrometer data fusion, By the fusion of the network model and its learning. Finally, the test data of peak drift is calculated and simulated based on the mathematics models. The result proved that the mathematic models have high precision and the fusion technology of information can be used in fusion of peak drift successfully.