A single and dual parameter data acquisition, ion beam measurements and control system for accelerator mass spectrometry is described. The system hardware has been constructed with the advantage of the lower cost and ...A single and dual parameter data acquisition, ion beam measurements and control system for accelerator mass spectrometry is described. The system hardware has been constructed with the advantage of the lower cost and higher reliability. It is provided with varieties of functions such as selecting acquisition mode carrying out the multiple display, analyzing data and especially viewing isometric spectrum at different directions. It can also be used for ordinary nuclear spectrum system.展开更多
Integral experiments on tungsten slab samples were carried out on the D-T neutron source facility at China Institute of Atomic Energy. Leakage neutron spectra from the irradiated tungsten target were measured by the t...Integral experiments on tungsten slab samples were carried out on the D-T neutron source facility at China Institute of Atomic Energy. Leakage neutron spectra from the irradiated tungsten target were measured by the time-of-flight technique. Accuracy of the nuclear data for tungsten was examined by comparing the measured neutron spectra with the leakage neutron spectra simulated using the MCNP-4C code with evaluated nuclear data of the JEFF-3.2, FENDL-3.0 and TENDL-2014 libraries. The results show that the calculations with JEFF-3.2 agree well with the measurements in the whole energy range and all angles, whereas the spectra calculated with FENDL-3.0 and TENDL-2014 have some discrepancies with the experimental data.展开更多
近期提出的单体相移深度神经网络(single phase-shift deep neural network,SPDNN),因其网络规模小、学习精度高,成为首个复杂中子共振截面拟合与评价的实用深度学习工具。在SPDNN学习共振截面的过程中,诸多因素显著影响网络的训练效果...近期提出的单体相移深度神经网络(single phase-shift deep neural network,SPDNN),因其网络规模小、学习精度高,成为首个复杂中子共振截面拟合与评价的实用深度学习工具。在SPDNN学习共振截面的过程中,诸多因素显著影响网络的训练效果、训练效率以及训练模型的泛化性。这些因素包括:决定网络相移层大小的共振截面频谱范围与频段宽度、隐藏层的数目、每层神经元的数目、激活函数、损失函数、训练步数和训练数据的预处理等。为了进一步提升SPDNN在共振截面研究中的实用性,详细考察了这些因素对网络拟合性能的影响。通过考察,确定了SPDNN在共振截面研究中适宜的网络构建和训练方法,助力推动SPDNN的广泛应用。展开更多
文摘A single and dual parameter data acquisition, ion beam measurements and control system for accelerator mass spectrometry is described. The system hardware has been constructed with the advantage of the lower cost and higher reliability. It is provided with varieties of functions such as selecting acquisition mode carrying out the multiple display, analyzing data and especially viewing isometric spectrum at different directions. It can also be used for ordinary nuclear spectrum system.
基金supported by the National Natural Science Foundation of China(No.11605097,91426301,and 11605257)Doctoral Scientific Research Foundation of Inner Mongolia University for the Nationalities(No.BS365)the‘‘ADS project 302’’of the Chinese Academy of Sciences(No.XDA03030200)
文摘Integral experiments on tungsten slab samples were carried out on the D-T neutron source facility at China Institute of Atomic Energy. Leakage neutron spectra from the irradiated tungsten target were measured by the time-of-flight technique. Accuracy of the nuclear data for tungsten was examined by comparing the measured neutron spectra with the leakage neutron spectra simulated using the MCNP-4C code with evaluated nuclear data of the JEFF-3.2, FENDL-3.0 and TENDL-2014 libraries. The results show that the calculations with JEFF-3.2 agree well with the measurements in the whole energy range and all angles, whereas the spectra calculated with FENDL-3.0 and TENDL-2014 have some discrepancies with the experimental data.
文摘近期提出的单体相移深度神经网络(single phase-shift deep neural network,SPDNN),因其网络规模小、学习精度高,成为首个复杂中子共振截面拟合与评价的实用深度学习工具。在SPDNN学习共振截面的过程中,诸多因素显著影响网络的训练效果、训练效率以及训练模型的泛化性。这些因素包括:决定网络相移层大小的共振截面频谱范围与频段宽度、隐藏层的数目、每层神经元的数目、激活函数、损失函数、训练步数和训练数据的预处理等。为了进一步提升SPDNN在共振截面研究中的实用性,详细考察了这些因素对网络拟合性能的影响。通过考察,确定了SPDNN在共振截面研究中适宜的网络构建和训练方法,助力推动SPDNN的广泛应用。
