Fast adaptive particle spectrum fitting algorithm based on moment-estimated initial parameters

Rui Shi; Xianguo Tuo; Honglong Zheng; Huailiang Li; Yangyang Xu; Qibiao Wang; Chao Deng

doi:10.1016/j.apradiso.2017.07.053

Fast adaptive particle spectrum fitting algorithm based on moment-estimated initial parameters

Appl Radiat Isot. 2017 Nov:129:1-5. doi: 10.1016/j.apradiso.2017.07.053. Epub 2017 Jul 31.

Authors

Rui Shi¹, Xianguo Tuo², Honglong Zheng³, Huailiang Li⁴, Yangyang Xu⁴, Qibiao Wang⁵, Chao Deng³

Affiliations

¹ Fundamental Science on Nuclear Wastes and Environmental Safety Laboratory, Southwest University of Science and Technology, Mianyang, Sichuan, China; Sichuan University of Science & Engineering, Zigong, Sichuan, China; State Key Laboratory of Geohazard Prevention and Geoenvironment Protection, Chengdu University of Technology, Chengdu, Sichuan, China.
² Fundamental Science on Nuclear Wastes and Environmental Safety Laboratory, Southwest University of Science and Technology, Mianyang, Sichuan, China; Sichuan University of Science & Engineering, Zigong, Sichuan, China; State Key Laboratory of Geohazard Prevention and Geoenvironment Protection, Chengdu University of Technology, Chengdu, Sichuan, China. Electronic address: tuoxg@cdut.edu.cn.
³ Sichuan University of Science & Engineering, Zigong, Sichuan, China.
⁴ Fundamental Science on Nuclear Wastes and Environmental Safety Laboratory, Southwest University of Science and Technology, Mianyang, Sichuan, China.
⁵ Sichuan University of Science & Engineering, Zigong, Sichuan, China; State Key Laboratory of Geohazard Prevention and Geoenvironment Protection, Chengdu University of Technology, Chengdu, Sichuan, China.

PMID: 28772146
DOI: 10.1016/j.apradiso.2017.07.053

Abstract

An algorithm based on moment estimation is presented to determine the initial parameters of the particle spectrum peak shape function for the iteration fitting procedure. The algorithm calculates the mean value, variance, and third-order central moment by using the spectrum peak data, solves the parameters of the fitting function, and then provides them as the initial values to the Levenberg-Marquardt algorithm to ensure convergence and optimized fitting. The effectiveness of the proposed algorithm was tested by gamma and alpha spectra. The algorithm can be used in automated peak curve fitting and spectral analysis.

Keywords: Initial parameters; Moment estimation; Particle spectrum; Peak shape function.

Publication types

Research Support, Non-U.S. Gov't