XRFS Determination Of Acid-soluble Aluminum Content With BP Algorithm

Ferric oxide powder contains the acid soluble and acid insoluble aluminum, solublealuminum is elemental aluminum species present in the iron oxide, Acid-insoluble aluminumpresent in the ferric oxide in the form of aluminum oxide, Acid-insoluble aluminum on steelmaking refining will be adversely affected, The total content of elemental aluminum andAl2O3called entire aluminum. Laboratory using the spectrophotometric determination ofacid-soluble aluminum content in the iron oxide powder, the analysis methods are morecomplex and long experimental period and it can not meet the requirements of routinelaboratory analysis of detection of increasingly tight for the experimental period. How can theestablishment of a rapid and accurate detection method instead of the previous complex,time-consuming testing methods to complete the routine testing has been a laboratory forchemical analysis of workers considerThis paper studies the X-ray fluorescence spectroscopy (XRF) combined with the BPneural network to determine the acid-soluble aluminum content in the iron oxide powder.Ferric oxide powder will be made of tablet solids to be determined the total aluminum contentin the iron oxide by X-ray fluorescence spectrometer. By using ANN-BP network model,entering the total aluminum content and directly predicting the content of soluble aluminum.1. The paper studies method of XRF analysis of the determination of aluminumcontent in iron oxide. The parameter was chosen to confirm the optimal parameters of theanalysis method. And establish the analysis method. And the method is reliable bycomparison of two methods.2. Selecting Model of artificial neural network, establishing the BP neural network.The paper studies the selection of BP network used in the X-ray spectrometry detectionmethod of the soluble aluminum content network parameters for optimum network trainingeffect. The method is reliable by experiment.In this paper, artificial neural network used in the actual work network to solve theacid-soluble aluminum content of iron oxide powder in the XRF detection method is simplerand faster, the results meet production requirements and suitable for laboratory analysis.Themethod is the way of an application of artificial neural networks in chemical analysis, methods to establish, can provide a solution for future chemical analysis methods and ideas.