Batch Gamma Spectrum Decomposition And Procedures Optimization Based On The LabVIEW

The exploration of the Moon is beneficial to the utilization and development oflunar resources, also has a very important significance in human society forsustainable development. In order to detect the lunar surface distribution of thecharacteristic elements, Chang’E-1Carrying the gamma-ray spectrometer (CGRS) todetect the gamma-ray spectroscopy which come from menology main elements andtrace elements. Through the exploration gamma ray obtained the correspondingelement, from gamma rays flow get corresponding elements abundance.The amount of the lunar gamma-ray spectrum data from CGRS are very large,only the2C data files has28.9G-byte. After a series of preliminary data processing,we found that energy spectrum has many overlapping peaks, which is not conduciveto the extraction of characteristic peaks. When the computer processed the batch data,it will take too much times, also take up too much memory, even unable to run. All ofabove problem go against data processing, made the processing slowly and also not soefficient, it will increase the difficulty for the last procedure that getting abundancemap of characteristic element.Analyzed the overlapping peaks accurately, then got each peak for relatedparameters information, can allow us to judge the elements of information, so as toachieve the element abundances value of moonscape, and finally obtain thecharacteristic elements abundance figure, realize the moon; At the same time, becauselunar data was detected by detector after a long time, so it will be very large. Based ona batch data processing, all need to optimize the process after build in data processingsoftware platform, otherwise, it will make the whole data processing become difficult.So for the processing of batch data which comes from CGRS, decomposed overlappedpeaks and simultaneously optimized the program are very important. This paper mainly focused on the batch gamma spectrum data with the problemof overlapping peaks based on LabVIEW software technology, after analyzed somematerial about decomposition of overlapping peaks as well as program optimization.Following two aspects of the conclusions can be drawn.1, Due to the characteristic values of some elements closing too much，there isoverlapping peaks in the obtained spectrum. To this phenomenon, based on analyzingof domestic and foreign literature regarding overlapping peaks decompositionmethods, this paper had a study of the decomposition of overlapping peaks. We choseGaussian function being fitting function, used combined methods to find the peak ofoverlapping peaks. Also proposed when used the wavelet method to determine thepeak position, can add Gaussian least squares iterative optimization method to searchpeaks, made the peak position more accurately, the peak relative error greatlydecreased, had ensured the accuracy of the peak position.Changed the separating degree of overlapping peaks, then added the noise, or notto add the noise, on different condition, to compare the result of peak searching by thewavelet method, and the second order spline wavelet convolution, also differentialmethod. Ultimately found three kind of method has its own advantages, among threemethod, wavelet method is the most ideal way for peak searching, it can determine thepeak of the number and a peak position better. Especially in noisy situation, it getsrelatively good results, for it basically free from the impact of noise in spectrum signal.The use of the second derivative method improved the resolution of overlappingpeaks, it has advantage that being sensitive to waveform changes, and we took fulladvantage of this merit to determine the full width, combined with the waveletmethod, added wavelet noise reduction for each level of the differential, and ensurethe advantages of the differential method. Bonding with the L-M least squares methodfor parameter optimization, at last decomposed and fitted the corresponding singlepeak. The combination of methods can overcome the defects of the traditional singlemethod, improved the precision of the decomposition of overlapping peaks.In order to verify the validity and accuracy, simulated the known single peak andoverlapping peak, decomposed them in noisy and noisy-free condition, after estimatedthe parameters, inserted parameters into the Gaussian function fitting. In the processof fitting, had reach a conclusion that for those gamma elements which have betterstatistical radioactive generate characteristic energy spectrum, use Gaussian functionas fitting function is feasible, and will have the good parameter estimation. For thosewho have large separate degree, the detected peak position is the son of peak position; and for those who have small separate degree, the detected peak position need tocorrect precisely to minimize the error, because the accuracy of the peak positionaffect the result of whole overlapping peak-fitting process directly.In addition, considered the factors (like the resolution, peak height ratio, noise,etc.) that can affect the decomposed result of pair-overlapping peaks. Finallydecomposed and fitted the energy value of662keV-1000keV (including U and Fe, Thpeak) which is in the Chang’E lunar exploration spectrum. By piecewise fitting way,divided the spectrum into the decomposition of overlapping peaks, and the fitting ofsingle peak, estimated the parameter of spectrum, fitted and decomposed theoverlapping peaks successful.2, For batch characteristics of the gamma spectrometry data, when processed anddiscussed them, the storage and extraction of batch gamma data are very big, soinvolved arrays will be very large, the processing program also were relatively complex.Owing to the processing of graph take much memory, when used the graphs of spectraldata frequently, it will lead to the execution speed slow, also will consume too muchcomputer memory. In order to solve this problem, had established the decomposition ofoverlapping peaks on the software platform, optimized the program from three aspects(improve the speed of program, reduce the consumption of computer memory,procedural standardization), the optimization which can effectively enhance the wholeprogram performance, accelerating the extraction of information and data processingspeed, makes the program have more efficient and smaller application error rate.Ensure the results of gamma spectra data, also make certain that the processing of batchdata is precise and the execution speed is fast, ultimately made the processing of batchgamma spectroscopy data effective and real-time.