Portable Spectrometer Based On Energy Spectrum Data Processing

In recent years, Nal (TI) γ digital portable spectrometer is a hot topic of research. Digital spectrometer is much more efficiency than the previous analog spectrometer detection in some way. Eliminating the process that used to need to transfer the data to the computer for processing and effectively achieves the results of quickly detected in any field measurement The successful implementation of a portable measurement to bring greater convenience because of the constant development and innovation of the technology, as well as examine the various aspects of embedded systems continue to make new achievements, to digital portable spectrometer the study provides a lot of convenience. Currently, we studied this NaI (TI) γ digital portable spectrometer based on a combination of previous experience, not only achieved a certain amount of technical innovations in terms of hardware, performing data processing for conventional spectrometer for radioactive content is low, while a wide range of test results are not accurate, and portable spectrometer itself lower measurement accuracy issues in-depth research.Based on the analysis and comparison of conventional γ spectroscopy data processing algorithms, based on a detailed account of the digital portable spectrometer to identify the spectrum of radionuclide data processing method for the solution of the difficulty spectrum of Nal (TI Digital Portable) y spectrometer in of spectral data smoothing, peak recognition algorithm, to establish a unique nuclide library, as well as qualitative final nuclide identification issues in-depth analysis, research, and verified the relevant algorithm. For CPU computing power and capacity issues portable instruments, the data presented by principal component analysis and then substituted into the BP neural network for classification and identification methods. The results show that this method of ensuring the premise of the original degree of confidence to effectively improve the recognition efficiency. Meanwhile this paper proposes a new approach to direct matching qualitatively identify the location of radionuclide by the experimental data and compare results on neural network method, this new method has a high recognition accuracy, and with respect to the neural network methods, this new method of directly matching in ensuring recognition accuracy relative premise, with recognition fast, easy programming and so on.