A Study on the Energy Resolution Correction Method for Measured Spectra of Nuclear Radiations

本文研究了一种修正和提高实测γ能谱能量分辨率的计算方法。该方法是基于一个非线性最小二乘法使用多维对数正态分布与贝叶斯定理。它可以引入不确定性，解决常规技术中经常出现的振动和负值问题。在测量系统中，光谱的分辨率变差。在γ射线能谱测量中，这种响应函数可以近似为高斯函数。作为先验信息的初始协方差数据使用逆Weierstrass变换来估计。该方法已成功地应用于光谱模拟。通过模拟计算发现，校正限与测量谱的不确定度密切相关，分辨率校正后的适用限谱是由强度比大于2倍、距离小于分辨率（1 σ）的谱峰组成的。

A calculational method to correct and improve the energy resolution of measured gamma ray spectra has been investigated. This method is based on a nonlinear least-squares method using a multi-dimensional lognormal distribution with Bayesian theorem. It can introduce the uncertainties and solve the problems of the vibration and the negative values, which sometimes have been occurred with the ordinary techniques. Resolutions of spectra are made worse in measured systems. This response function in gamma ray spectrometry can be approximated the Gaussian. The initial covariance data as a prior information is estimated using the inverse Weierstrass transformation. This method has been successfully applied to simulate spectra. Through the simulate calculations, these are found that the corrective limitation is strongly related to the uncertainty of measured spectra, and that the applicably limited spectra to separate peaks as the results of resolution correction are composed of peaks of which the intense ration are above the factor of 2 and the distances are closer than the resolution (1 sigma).