Font Size: a A A

HPGe ? Fingerprint Correction Based On System Transformation And Its Fuzzy Recognition

Posted on:2021-01-08Degree:MasterType:Thesis
Country:ChinaCandidate:J H WangFull Text:PDF
GTID:2392330626464980Subject:Particle Physics and Nuclear Physics
Abstract/Summary:PDF Full Text Request
? fingerprint,as the radiation fingerprint of nuclear materials,has the function of marking and identifying radioactive objects.Therefore,the technology of the analysis and identification of ? fingerprint has been widely used in many aspects of nuclear security,such as nuclear material identification in the fields of nuclear terrorism,nuclear non-proliferation and nuclear weapons verification.However,in the process of ? fingerprint measurement,due to the influence of the instability and nonlinearity of the ? spectrometer system,? fingerprint will have spectral line drift,deformation and distortion,which will reduce its identification confidence of the fingerprint.Based on system transformation theory,this dissertation proposes a ? fingerprint correction method to solve this problem.The deposition energy of ? rays in the detector and their corresponding channel addresses in the multichannel analyzer are regarded as two random variables,and the normalized energy deposition fingerprint and the normalized measured ? fingerprint are regarded as probability density functions of the two random variables,respectively.The changes of the parameters of ? spectrometer can be regarded as the changes of the system parameters.Therefore,the theoretical deposition spectrum of ?-ray in the detector,which was obtained through the system transformation theory of the random signal,eliminated the influence of the the drift,deformation and distortion of ? fingerprint caused by the instability and nonlinearity of ? spectrometer system on the recognition of ?fingerprint,which lead to the effective correction of ? fingerprint.The detailed work is as follows:?1?Demonstrating the necessity of ? fingerprint correction through studying the influence of the instability of ?-ray spectrometer on the results of the identification of ? fingerprint.Taking 60Co and 137Cs as research objects,two groups of ? fingerprints are simulated by Monte Carlo simulation technology: one group is ? fingerprints which simulates the changes of system parameters of ? spectrometer while the nuclear materials remain unchanged,the other group is ? fingerprints which simulate the subtle changes of nuclear materials while the system parameters of ? spectrometer remain unchanged.The two groups of simulated ?fingerprints are corrected through the ? fingerprint correction method,based on the system transformation theory.Next,the fuzzy identification method is used to identify the two groupsof simulated ? fingerprints before and after correction,respectively.Then,the necessity of ?fingerprint modification is demonstrated,through the comparison and analysis of the identification results before and after correction,respectively.?2?Demonstrating the correction method of this dissertation through the correction and recognition of the measured ? fingerprint.Selecting 60Co,137Cs and 152Eu as research objects,three groups of data of the ?fingerprint are measured: the first group is to measure the statistical fluctuation ? fingerprints of 152Eu point sources repeatedly.On the basis of the measured data of the first group,the second group and the third group take the 60Co point sources and 137Cs point sources instead of 152Eu point sources,respectively.Two groups of nuclear material ? fingerprints with subtle differences are obtained by increasing the measurement time of 60Co and 137Cs.Three groups of measured ? fingerprints are corrected through the correction method proposed in this dissertation.Next,the fuzzy identification method is used to identify the corrected three groups of measured ? fingerprints before and after correction,repectively.Then,the feasibility of the correction method is further demonstrated through the comparison and analysis of the identification results of the three groups of measured ? fingerprints before and after correction,repectively.The results show that the modified ? fingerprint is transformed from channel address domain to energy domain,which solves the problem of the drift of peak position.In addition,the peak height,peak half width and peak area have been effectively modified.The modified? fingerprint eliminate the influence on the recognition of ? fingerprint caused by the instability and nonlinearity of the system of the ? spectrometer,and improve the recognition confidence.The results show that the modified fingerprint transforms the trace address domain to the energy domain,theoretically eliminates the influence of the system instability and nonlinearity on the fingerprint,and solves the problem of spectral line drift.In addition,the peak height,the peak half width and the peak area are effectively corrected,and the identification confidence of nuclear materials is improved.
Keywords/Search Tags:? fingerprint, random process, system transformation, fuzzy recognition, membership degree
PDF Full Text Request
Related items