EDXRFA Software Development And Algorithm Design

Energy dispersive X-ray fluorescence analysis technology is an important tool of modern analysis ways, which is applied in steel,architecture material,electronic/magnetic material,chemistry industry,petroleum and charcoal,agriculture,environmental sample field .X-ray spectrum analysis is a mathematical method of improve energy dispersive X-ray fluorescence analysis precision and get more information from spectrum. The method compensates measure error in deficiency of detector's energy resolution.The paper discusses least-squares theory, including linear least-square fitting,orthogonal polynomials least-square fitting and least-square fitting using analytical functions, especially L-M algorithm. The EDXRFA software uses Savitsky and Golay polynomials filters to smooth spectrum,peak stripping and orthogonal polynomials least-square fitting to remove background,top-hat filter to search peak after energy calibration. At last, we fit the spectrum with Gaussian functions in analytical functions least-square fitting. The software is accomplished by VC programming, comparing and analyzing fitting effect and error by drawing and listing fitting data.We can see the average square error is about 40. 07 from the testing, improving analysis precision and extending its application.