分类： 物理学 >> 核物理学 提交时间： 2025-06-25
摘要： Automatic identification of radioactive isotopes through energy spectrum analysis is vital for remote, unmanned monitoring of radioactive contamination and rapid early warning. In recent years, deep learning methods have advanced significantly, outperforming traditional approaches in recognition accuracy. However, their purely data-driven nature and the "black-box" characteristics of neural networks result in poor interpretability, a high risk of overfitting, and uncontrollable errors, limiting their use in high-reliability fields like the nuclear industry. We present FhyMetric-Net, a novel multi-label classification model that integrates physical constraints with data-driven techniques. This model automatically infers the probability of mixed nuclides and provides weight interpretations consistent with expert knowledge. Our approach is groundbreaking in embedding prior characteristic peak physical information into neural networks, effectively constraining the feature weight optimization space for improved reliability and interpretability. We also introduce a novel metric constraint method in the feature space, tailored for mixed nuclide samples, which enhances the model’s ability to extract discriminative features. By establishing a clear causal link between predicted probabilities and channel addresses, FhyMetric-Net overcomes the interpretability challenges of traditional dense fully connected layers. We conducted more challenging quantitative tests than previous studies. When faced with challenges such as an increased number of mixed radionuclides, variations in Gaussian broadening coefficients, and differences in detector types, the proposed model consistently maintained an F1 score above 95%, achieving state-of-the-art (SOTA) performance, while the model's parameter count was only 1.58% of the ResNet-18 model. In scenarios with low gross count and low signal-to-noise ratio (SNR), its overall performance also demonstrated significant advantages. Qualitative analysis further confirmed the model's strong physical interpretability. This achievement will advance the application of automated mixed radionuclide identification technology in high-reliability fields of the nuclear industry.
分类： 物理学 >> 核物理学 提交时间： 2025-01-05
摘要： The qualitative identification and quantitative analysis of radioactive nuclides in unknown environments are essential for remote monitoring and prompt early warning of radioactive contamination. In recent years, deep learning techniques have made significant strides in automated qualitative identification. However, the quantitative analysis of radioactive nuclides still depends on traditional methods to determine peak positions and boundaries. These methods often require extensive manual expertise and parameter tuning, which makes it challenging to meet the demands of unmanned remote monitoring. This paper presents a novel framework for automatic full-energy peak segmentation, named YOLOSpecNN. We introduced a multi-Root Mean Square Error (RMSE) joint optimization function and developed a unified regression model capable of simultaneously predicting the central position, boundaries, and confidence of full-energy peaks. To address the challenge of low recall rates due to narrow, weak, and overlapping peaks, we proposed a new multi-scale context feature extraction module (MSNN module). This module effectively enhanced local detail features, significantly improving recall rates. The effectiveness of the proposed method was validated using six artificial radioactive nuclides (241Am,57Co,131I,134Cs,137Cs,and 60Co), along with 40K, to construct a mixed energy spectrum dataset for quantitative evaluation. Experimental results show that the proposed method significantly outperforms traditional approaches, achieving a precision of 0.998, recall of 0.95, and the best F1 score of 0.974@0.427, and the average precision of 0.946. Compared to traditional morphological methods, the proposed method improves precision, recall, and the best F1 score by 0.512, 0.199, and 0.391, respectively. Ablation experiments further reveal that the MSNN module notably enhances recall, with an improvement of 0.067. Moreover, the proposed method performs excellently even in challenging environments with low gross counts and low Signal-to-Noise Ratio (SNR), achieving state-of-the-art (SOTA) results. Additionally, the model achieves an average real-time inference performance of 16.1941 ms on a 15 W low-power device. Overall, the proposed method demonstrates exceptional performance in the automatic search and segmentation of full-energy peaks, offering robust support for the implementation of unmanned remote radiation monitoring systems.
分类： 物理学 >> 核物理学 提交时间： 2024-02-14
摘要： The distribution of the nuclear ground-state spin in a two-body random ensemble (TBRE) was studied using a general classification neural network (NN) model with two-body interaction matrix elements as input features and the corresponding ground-state spins as labels or output predictions. The quantum many-body system problem exceeds the capability of our optimized NNs in terms of accurately predicting the ground-state spin of each sample within the TBRE. However, our NN model effectively captured the statistical properties of the ground-state spin because it learned the empirical regularity of the ground-state spin distribution in TBRE, as discovered by physicists.
分类： 物理学 >> 核物理学 提交时间： 2024-12-12
摘要： Using the charge density from the two-parameter Fermi model, a robust and nontrival correlation between binding energies and charge radii of mirror nuclei is newly proposed. This correlation enables simple yet reliable predictions of the nuclear mass and charge radius of proton-rich nuclei. The validity of these predictions is demonstrated by comparing the predicted binding energies and charge radii with experimental data and predictions from other models. All 197 predicted binding energies and 199 charge radii involved in the comparisons are tabulated in the Supplemental Materials of this paper. The noticeable discrepancies are attributed to the large asymmetry in charge densities of mirror nuclei, suggesting that the proposed correlation could be a sensitive probe for local structural anomaly, such as shell closure and proton halo. The difference in mass dependence of charge radii near the proton dripline compared to those along the \beta-stability line supports the validity of our prediction method.
分类： 核科学技术 >> 辐射物理与技术 提交时间： 2024-09-23
摘要： In scenarios such as vehicle radiation monitoring and unmanned aerial vehicle radiation detection, rapid measurements using a NaI(Tl) detector often result in low photon counts, weak characteristic peaks, and significant statistical fluctuations. These issues can lead to potential failures in peak-searching-based identification methods. To address the low precision associated with short-duration measurements of radionuclides, this paper proposes an identification algorithm that leverages heterogeneous spectral transfer to develop a low-count energy spectral identification model. Comparative experiments demonstrated that transferring samples from 26 classes of simulated heterogeneous gamma spectra aids in creating a reliable model for measured gamma spectra. With only 10% of target domain samples used for training, the accuracy on real low-count spectral samples was 95.56%. This performance shows a significant improvement over widely employed full-spectrum analysis methods trained on target domain samples. The proposed method also exhibits strong generalization capabilities, effectively mitigating overfitting issues in low-count energy spectral classification under short-duration measurements.
分类： 物理学 >> 交叉学科物理及相关领域的科学与技术 分类： 材料科学 >> 陶瓷学 提交时间： 2024-03-20
摘要： 高放废物固化体的长期辐照稳定性是评价其深地质处置安全性的关键指标之一。本文利用 500 keV He2+离子束辐照 2M、 3O 和 4M 型钙钛锆石陶瓷固化体， 研究高放核素长期α 衰变辐照对陶瓷结构的损伤行为。三种钙钛锆石样品通过传统高温固相法（1400 ℃， 48h两次烧结）合成。掠入射 X 射线衍射（GIXRD） 、 Raman 光谱和 XPS 研究结果发现在 α 粒子辐照下， zirconolite-2M 固化体在辐照非晶化的过程中经历了 zirconolite-2M 到 zirconolite-4M 到 pyrochlore 的结构演变； zirconolite-3O 在 5×1015 ions/cm2 剂量下部分晶体结构转变为钙钛矿，在 1×1016 ions/cm2 以上剂量快速转变为无序烧绿石； zirconolite-4M 样品在 1×1017 ions/cm2 剂量时依旧保留 zirconolite-4M 主相，次生无序烧绿石/缺陷萤石晶相。综上， 从维持晶体结构稳定性来看 zirconolite-2M 和 zirconolite-3O 的抗 α 粒子辐照能力相似，略弱于zirconolite-4M 固化体。
分类： 核科学技术 >> 核科学与技术 提交时间： 2024-10-25
摘要： With the rapid development of nuclear energy, the removal of radioactive iodine generated during spentfuel reprocessing has become increasingly important. Based on the unique straw-like structure of populus tomentosa fiber (PTF) and the highly active iodine vapor capture ability of zero-valent silver nanoparticles (PTF@Ag0NP), an Ag0NP composite functional material with highly efficient iodine vapor capture capability was synthesized from biowaste PTF through ultrasonic and high-temperature hydrothermal methods in this study. The iodine capture experiment demonstrated that PTF@Ag0NP exhibits rapid iodine capture efficiency, reaching dynamic equilibrium within 4 h and a maximum capture capacity of 1008.1 mg/g. Density functional theory calculations show that PTF@Ag0NP exhibits extremely high chemical reactivity towards iodine, with a reaction binding energy of −2.88 eV. Additionally, the molecular dynamics of PTF@Ag0NP indicate that there is no atomic displacement at 77 ℃, indicating the excellent temperature stability of the material at the operating temperature. The capture mechanism suggests that iodine vapor primarily reacts with Ag0NP to form AgI, and that the hydroxyl groups in PTF can also effectively capture iodine vapor by adsorption induction. In conclusion, PTF@Ag0NP is expected to be an effective candidate adsorbent material for removing radioactive iodine vapor from exhaust gases during spent-fuel reprocessing.
分类： 物理学 >> 普通物理:统计和量子力学,量子信息等 提交时间： 2023-09-05
摘要： This study introduces a novel algorithm to detect and identify radioactive materials in urban settings using time-series detector response data. To address the challenges posed by varying backgrounds and to enhance the quality and reliability of the energy spectrum data, we devised a temporal energy window. This partitioned the time-series detector response data, resulting in energy spectra that emphasize the vital information pertaining to radioactive materials. We then extracted characteristic features of these energy spectra, relying on the formation mechanism and measurement principles of the gamma-ray instrument spectrum. These features encompassed aggregated counts, peak-to-flat ratios, and peak-to-peak ratios. This methodology not only simplified the interpretation of the energy spectras physical significance but also eliminated the necessity for peak searching and individual peak analyses. Given the requirements of imbalanced multi-classification, we created a detection and identification model using a weighted k-nearest neighbors (KNN) framework. This model recognized that energy spectra of identical radioactive materials exhibit minimal inter-class similarity. Consequently, it considerably boosted the classification accuracy of minority classes, enhancing the classifiers overall efficacy. We also executed a series of comparative experiments. Established methods for radionuclide identification classification, such as standard k-nearest neighbors (KNN), support vector machine (SVM), Bayesian network, andrandom tree, were used for comparison purposes. Our proposed algorithm realized an F1 measure of 0.9868 on the time-series detector response data, reflecting a minimum enhancement of 0.3% in comparison to other techniques. The results conclusively show that our algorithm outperforms others when applied to time-series detector response data in urban contexts.
分类： 核科学技术 >> 粒子加速器 提交时间： 2023-06-18 合作期刊: 《Nuclear Science and Techniques》
摘要： In this paper, a genetic-algorithm-based artificial neural network (GAANN) model radioactivity prediction is proposed, which is verified by measuring results from Long Range Alpha Detector (LRAD). GAANN can integrate capabilities of approximation of Artificial Neural Networks (ANN) and of global optimization of Genetic Algorithms (GA) so that the hybrid model can enhance capability of generalization and prediction accuracy, theoretically. With this model, both the number of hidden nodes and connection weights matrix in ANN are optimized using genetic operation. The real data sets are applied to the introduced method and the results are discussed and compared with the traditional Back Propagation (BP) neural network, showing the feasibility and validity of the proposed approach.
分类： 核科学技术 >> 粒子加速器 提交时间： 2023-06-18 合作期刊: 《Nuclear Science and Techniques》
摘要： In energy dispersive X-ray fiuorescence (EDXRF), quantitative elemental content analysis becomes difficult due to the existence of the noise, the spectrum peak superposition, element matrix effect, etc. In this paper, a hybrid approach of genetic algorithm (GA) and back propagation (BP) neural network is proposed without considering the complex relationship between the elemental content and peak intensity. The aim of GA-optimized BP is to get better network initial weights and thresholds. The starting point of this approach is that the reciprocal of the mean square error of the initialization BP neural network is set as the fitness value of the individuals in GA; and the initial weights and thresholds are replaced by individuals, then the optimal individual is searched by selecting, crossover and mutation operations, finally a new BP neural network model is established with the optimal initial weights and thresholds. The quantitative analysis results of titanium and iron contents in five types of mineral samples show that the relative errors of 76.7% samples are below 2%, compared to chemical analysis data, which demonstrates the effectiveness of the proposed method.
分类： 天文学 >> 天文仪器与技术 提交时间： 2024-12-27 合作期刊: 《Astronomical Techniques and Instruments》
摘要： Radio astronomy necessitates radio frequency bands that are both stable and free from interference at observatory locations. To comprehensively evaluate the radio environment at radio observatories, we employ Monte Carlo methods to assess the quality of observational data and predict potential interference. With an extensive dataset, we used an algorithm to find the interference threshold within the L-band, automatically identifying disruptive signals. Monte Carlo simulations were conducted to estimate whether these interference signals surpass a predetermined threshold of the total observation period, facilitating a detailed analysis of the interference profile. A Monte Carlo analysis was used on 83 hours of continuous monitoring data using a wireless environment testing system, to forecast the proportion of time during which interference signals would surpass established harmful thresholds. Our findings indicate that, within the L-band spectrum at Fenghuang Hill, Kunming City, Yunnan Province, the incidence of interference within the frequency ranges of 1 330–1 440 MHz, 1 610–1 613 MHz, and 1 660–1 670 MHz is acceptably low, with respective confidence levels of 96.9%, 97.4%, and 97.4% that the proportion of time these interference signals occupy does not exceed 5% of the total observational time, as stipulated by the International Telecommunication Union. Conversely, the confidence level for the 1 718–1 722 MHz band not exceeding 5% of the total observational time is significantly lower at 88.5%. This study offers a valuable tool for assessing the radio environment in radio astronomy research and provides a foundational basis for the scientific management and safeguarding of radio frequency bands.
分类： 核科学技术 >> 辐射物理与技术 提交时间： 2025-03-10
摘要： Colloids are prevalent in nuclear waste repositories, with bentonite colloids posing an uncontrollable riskfactor for nuclide migration processes. In this study, static adsorption experiments were coupled with dynamicshower experiments to comprehensively investigate the influence of bentonite colloids on Sr2+ migration ingranite, considering adsorption capacity. Bentonite colloids have a considerably greater adsorption capacity thanboth bentonite and granite, with a maximum adsorption of 30.303 mg/g. The adsorption behavior of bentonitecolloids on Sr2+ is well described by the Langmuir isotherm and pseudo-second-order kinetic models, indicatingthat a single-layer chemical adsorption process is controlled by the site activation energy. The adsorbed Sr2+is unevenly distributed on the colloids, and the adsorption mechanism may involve ion exchange with Ca.Bentonite colloids exhibit superior adsorption in neutral environments. The cations in groundwater inhibit Sr2+adsorption, and the inhibition efficacy decreases in the order Fe3+>Ca2+>Mg2+>K+ . The presence of bentonitecolloids in a granite column slightly influences the retention of Sr2+ in the column while markedly reducingthe Sr2+ penetration time from 70 h to 18 h. However, the coexistence of Co2+, Ni2+, and Cs+ in a multinuclidesystem weakens the ability of the colloids to promote Sr2+ migration. In comigration of colloid and multinuclidesystems, the adsorption of nuclides by bentonite colloids causes the nuclide migration speed to decrease in theorder Sr2+>Cs+ >Ni2+>Co2+. This study provides insights into Sr2+ migration in cave repositories for low- andmedium-level radioactive waste.
分类： 天文学 >> 天文仪器与技术 提交时间： 2024-04-18 合作期刊: 《天文技术与仪器（英文）》
摘要：The high frequency-very high frequency (HF-VHF) frequency band is of significant importance in astronomical observations, with applications studying various phenomena such as space weather, solar radio emissions, planetary eruptions in the solar system, pulsars, transient sources, and reionization of the early universe. This article introduces the HF-VHF frequency band multifunctional radio astronomical terminal system based on a dual-channel high-speed acquisition board with a frequency observation range of 1−250 MHz and a sampling rate of 500 Msps (Mega samples per second). The maximum quantization bit of the system is 14 bits, with a maximum time resolution of 0.1 s and a maximum spectral resolution of 16 kHz. The system combines spectral analysis of solar radio signals and recording of time-domain data of signals interfering with long baselines, and adopts a server-client separation mode to allow remote operation with separate permissions. It is used in the China-Malaysia joint astronomy project, which can carry out single-site observation of solar radio signals as well as interferometric observation of signals from multiple sites.
分类： 天文学 >> 天文仪器与技术 提交时间： 2025-04-23 合作期刊: 《Astronomical Techniques and Instruments》
摘要： Radio antenna arrays have many advantages for astronomical observations, such as high resolution, high sensitivity, multi-target simultaneous observation, and flexible beam formation. Problems surrounding key indices, such as sensitivity enhancement, scanning range extension, and sidelobe level suppression, need to be solved urgently. Here, we propose a sparse optimization scheme based on a genetic algorithm for a 64-array element planar radio antenna array. As optimization targets for the iterative process of the genetic algorithm, we use the maximum sidelobe levels and beamwidth of multiple cross-section patterns that pass through the main beam in three-dimensions, with the maximum sidelobe levels of the patterns at several different scanning angles. Element positions are adjusted for iterations, to select the optimal array configuration. Following sparse layout optimization, the simulated 64-element planar radio antenna array shows that the maximum sidelobe level decreases by 1.79 dB, and the beamwidth narrows by 3°. Within the scan range of ±30°, after sparse array optimization, all sidelobe levels decrease, and all beamwidths narrow. This performance improvement can potentially enhance the sensitivity and spatial resolution of radio telescope systems.
分类： 核科学技术 >> 粒子加速器 提交时间： 2023-06-18 合作期刊: 《Nuclear Science and Techniques》
摘要： Direct demodulation method (DDM) was applied to reconstruct -ray spectra. Boosted Richardson-Lucy iteration was introduced into DDM. Monte Carlo method (here GEANT 4) was proposed to calibrate response function and establish response matrix. First, gauss function was regarded as total energy peak. Spectra line was simulated with nine gauss functions. And afterwards DDM was applied to reconstruct the simulated spectra line and determine peak positions and areas. Compared with original spectra, for case that peak position interval was about 1/3 full width half maximum (FWHM), the error of rebuilding peak position was 2 channels. The rest of peaks could be searched accurately. The relative errors of all peaks area were less than 4%. Then, three key factors, including noise, background, response matrix, were discussed. Finally, DDM was applied to calibrate the field NaI gamma spectrometer. The errors of U, Th, K were less than 5%. Comprehensive studies have shown that it is feasible to reconstruct gamma-ray spectra with DDM. DDM can significantly pseudo-improve energy resolution of gamma spectrometer, effectively decompose doublets whose peak potential interval is 1/3 FHWM, and accurately search peak and calculate areas. DDM can restrain noise strongly but is greatly influenced by background. And DDM can improve the accuracy of qualitative and quantitative analysis in combination with the conventional spectrum analysis method.
分类： 农、林、牧、渔 >> 农、林、牧、渔业科学其他学科 提交时间： 2024-06-17 合作期刊: 《智慧农业（中英文）》
摘要： ［目的/意义］传统的小麦倒伏检测方法需要人工进行田间观测和记录，这种方法存在主观、效率低、劳 动强度大等问题，难以满足大规模的小麦倒伏检测的需求。基于深度学习的小麦倒伏检测技术虽已在一定程度上 得到应用，但普遍局限于对小麦单一发育阶段的倒伏识别，而倒伏可能发生在小麦生长的各个时期，不同时期倒 伏特征变化复杂，这给模型特征捕捉能力带来考验。本研究旨在探索一种基于深度学习技术的多生育期小麦倒伏 区域检测方法。［方法］用无人机采集小麦灌浆期、早熟期、晚熟期这三个关键生长阶段的RGB图像，通过数据 增强等技术构建出多生育期小麦倒伏数据集。提出一种小麦倒伏提取模型Lodging2Former，该模型在Mask2Former 的基础上加以改进，引入分层交互式特征金字塔网络（Hierarchical Interactive Feature Pyramid Network, HI-FPN ）， 用于提高模型在复杂田间背景干扰下对于多个生长阶段小麦倒伏特征的捕捉能力。［结果和讨论］所提出的Lodg⁃ ing2Former模型相较于现存的多种主流算法，如Mask R-CNN （Mask Region-Based Convolutional Neural Network）、 SOLOv2（Segmenting Objects by Locations, Version 2） 以及Mask2Former，在平均精度均值（mean Average Precision, mAP） 上展现出显著优势。在阈值分别为0.5、0.75以及0.5~0.95的条件下，模型的mAP值分别达到了79.5%、 40.2%和43.4%，相比Mask2Former模型，mAP性能提升了1.3%~4.3%。［结论］提出的HI-FPN网络可以有效利用 图像中的上下文语义和细节信息，通过提取丰富的多尺度特征，增强了模型对小麦在不同生长阶段倒伏区域的检 测能力，证实了HI-FPN在多生育期小麦倒伏检测中的应用潜力和价值。
点击量
待评议
点击量
下载量
评论
通过
