Font Size: a A A

Research On Key Technologies Of Nuclear Pulse Monte Carlo Sampling Acceleration

Posted on:2021-05-09Degree:DoctorType:Dissertation
Country:ChinaCandidate:W ChenFull Text:PDF
GTID:1522307184453894Subject:Nuclear technology and applications
Abstract/Summary:PDF Full Text Request
Some methods,such as expanding the detector size,increasing radioactivity and extending measurement time,can be used to improving counting rate in radiation measurement.However,these methods are usually limited by the actual situations.The way of improving precision of nuclear analysis method is the important content in radiation measurement.The process of the radiation detection and the reciprocity between particles and detectors are random.Monte Carlo method is widely used for modeling of radiation detection system.However,increasing the counting rate in radiation measurement using Monte Carlo method has not been reported before.Thus,a kind of Nuclear Pulse sequence Sampling technique based on Monte Carlo method(NPSMC)is studied in this thesis to improving counting rate.The NPSMC is verified by the designed virtual detector which is developed on MATLAB platform using Monte Carlo method.The NPSMC is finally realized in multi-channel amplitude analyzer which is designed on Zynq So C platform.The feasibility of the NPSMC is also proved by the platform.The main achievements are as follows:(1)A nuclear pulse sequence sampling technique based on Monte Carlo method is proposed by combining radiation measurement technique and Monte Carlo method.According to the random characteristics of the radioactive decay,radiation detection,reciprocity between particles and detectors are random process,a nuclear pulse sequence sampling technique is proposed by integrating Monte Carlo method into actual radiation measurement,and proposes.Firstly,a dynamic sample pool is established for nuclear pulses that coming from radiation detector,and then the pulses in the pool are sampling according to Monte Carlo sampling method.Both the sampled pulses and the measured pulses are transmitted to the designed random pulse queue manager for amplitude discrimination increasing the counting rate of the energy spectrum.(2)A nuclear pulse generator which depends on the virtual detector and nuclear pulse sequence sampling is designed.The model of the virtual detector is established with Geant4.It is applied to generate amplitude information of the nuclear pulse.The generated amplitude is transmitted to the dynamic sample pool.Then,the amplitude information in the pool is sampled according to Monte Carlo method.The parameters of generating time,decay time constant which are characteristic parameters of a nuclear pulse are coupled with the sampled amplitude information to realize a complete nuclear pulse generator.The effectiveness of NPSMC is verified by the nuclear pulse generator which is established with Na I(Tl)detector and Co-60 source.It is proved that the NPSMC can be used for improving counting rate.(3)A parallel random number generator based on FPGA clock jitter entropy source and hash post-processing is proposed.It is used to generate uniform random numbers for NPSMC.The clock jitter entropy source is realized with the PL part of the Zynq platform based on the ring oscillator clock jitter and sub-sampling technology.The output data is processed with hash algorithm to accomplish a parallel random number generator.The proposed generator can produce rand numbers at100 Mbps and the generated random numbers satisfy the NIST randomness tes.(4)The multi-channel amplitude analyzer associated with NPSMC is developed.The output signal of the detector is conditioned by the front-end circuit and converted to discrete pulse.The nuclear pulses in the dynamic sample pool are sampled according to Monte Carlo method.Both the sampled pulses and the measured ones are transmitted to the multi-channel amplitude analyzer for amplitude discrimination.The NPSMC model is tested on the PL section of Zynq.Some models pulse shaping,baseline estimation,pile-up identification,amplitude discrimination and spectrum generation are also designed.The NPSMC is verified by simulation experiments which are carried out depending on MATLAB platform and the designed virtual detector.The simulated results indicate that the spectrum associating with NPSMC is consistent with the original spectrum.The element logging sample and copper standard sample(No.39X17867)are applied to test the multi-channel amplitude analyzer.The result indicates that the counting rate of the energy spectrum is developed improved by the NPSMC.It can be applied to improve precision of nuclear analysis.
Keywords/Search Tags:FPGA accelerating, random number, nuclear pulse sequence sampling, counting rate, Monte Carlo computation
PDF Full Text Request
Related items