贝叶斯反卷积在EAST电荷交换复合光谱分析中的应用

Applications of Bayesian deconvolution to the charge exchange recombination spectroscopy on EAST tokamak

电荷交换复合光谱(Charge e Xchange Recombination Spectroscopy,CXRS)诊断是核聚变装置上测量等离子体离子温度和旋转速度的常规诊断之一。然而在实验中,诊断光通过光谱仪后,由于仪器函数的卷积效应,会使测量到的光谱出现明显展宽,影响数据处理的精度,所以需要对实验测量到的光谱进行反卷积处理。本文采用的反卷积方法是基于贝叶斯条件概率公式推导得出,并结合标准灯获取的仪器函数来进行反卷积,分别从仿真和实验两个方面验证了该方法的可靠性。结果表明将贝叶斯反卷积运用到先进实验超导托卡马克(Experimental Advanced Superconducting Tokamak,EAST)电荷交换复合光谱分析中,能有效提高实验测量精度。结合快速极紫外谱仪(Extreme ultraviolet,EUV),对EAST实验中经过贝叶斯反卷积后测量到的光谱进行了杂质谱线识别工作,进一步提高了精度。

Background： Charge eXchange Recombination Spectroscopy （CXRS） is a routine diagnostic method for the measurement of plasma ion temperature and rotation velocity on nuclear fusion devices. The experimental spectrum can be obviously broadened by the instrument function （IF） convoluted, thus the deconvolution is needed for accurate data analysis. Purpose： This study aims to improve accuracy of data analysis by using the Bayesian deconvolution and impurity spectrum identification. Methods： The deconvoluted method utilizes the Bayesian condition probability formula. Standardized neon lamp is applied to get spectrometric IF for deconvolution processing Finally, the impurity spectra is identify fast-time-response extreme ultraviolet （EUV） to further improve the analysis accuracy. Results： Experimental results on Experimental Advanced Superconducting Tokamak （EAST） confirmed the reliability of Bayesian deconvolution that was previously verified by simulation study. Conclusion： Bayesian deconvolution combined with fast-time-response EUV can be effectively applied to the edge CXRS analysis on tokamak.