[Research on the peak search method based on improved simulated annealing model for x-ray fluorescence spectrum].

In energy dispersive X-ray fluorescence analysis technical analysis, the spectrum information processing has been being the focus of researches, spectrum smooth, peak-search, and peak area processing are the top priority, tA peak search method is based on simulated annealing algorithm principle, and to establish a new search peak model algorithm, this algorithm uses the advantages of the global convergence properties of simulated annealing and the Metropolis criterion is used as a peak valley judgment basis. A new decision criteria on peak valley array is used in this method, too. At the same time, convergence is taken from both ends of the spectrum channel address, it ends up when convergence reaches to the same optimal solution. This paper compared the algorithm with the simple comparison method, third derivative method, and the contrast experiment results, showing that this algorithm has a strong ability on X-ray fluorescence spectrum peak-search and a certain value in the actual production.