摘要：
Zincophilic property and high electrical conductivity are both very important parameters to design novel Zn anode for aqueous Zn-ion batteries (AZIBs). However, single material is difficult to exhibit zincophilic property and high electrical conductivity at the same time. Herein, originating from theoretical calculation, a zincophilic particle regulation strategy is proposed to address these limitations and carbon coated Na3V2(PO4)3 is taken as an example to be a protective layer on zinc metal (NVPC@Zn). Na3V2(PO4)3 (NVP) is a common cathode material for Zn-ion batteries, which is zincophilic. Carbon materials not only offer an electron pathway to help Zn deposition onto NVPC surface, but also enhance the zinc nucleophilicity of Na3V2(PO4)3. Hence, this hybrid coating layer can tune zinc deposition and resist side reactions such as hydrogen generation and Zn metal corrosion. Experimentally, a symmetrical battery with NVPC@Zn electrode displays highly reversible plating/stripping behavior with a long cycle lifespan over 1800 h at 2 mA cm-2, much better than carbon and Na3V2(PO4)3 solely modified Zn electrodes. When the Na3V2(PO4)3 is replaced with zincophobic Al2O3 or zincophilic V2O3, the stability of the modified zinc anodes is also prolonged. This strategy expands the option of zincophilic materials and provides a general and effective way to stabilize the Zn electrode. (c) 2023 Science Press and Dalian Institute of Chemical Physics, Chinese Academy of Sciences. Published by ELSEVIER B.V. and Science Press. All rights reserved.
作者机构：
[Senjie Zhu; Xiao Huang; Lei Xia; Yifei Zhang] University of Science and Technology of China, Hefei, Anhui Province 230026, China;Institute of Modern Physics, Chinese Academy of Sciences, Lanzhou, Gansu Province 730000, China;University of Chinese Academy of Sciences, Beijing 100049, China;Guangdong Provincial Key Laboratory of Nuclear Science, Institute of Quantum Matter, South China Normal University, Guangzhou 510006, China;Southern Center for Nuclear-Science Theory (SCNT), Institute of Modern Physics, Chinese Academy of Sciences, Huizhou, Guangdong Province 516000, China
摘要：
Charm quark production and its hadronization in ep and eA collisions at the future Electron-ion Collider in China (EicC) will help us understand the quark/gluon fragmentation processes and the hadronization mechanisms in the nuclear medium, especially within a poorly constrained kinematic region ( x<0.1 ). In this paper, we report a study on the production of charmed hadrons, D0 and Λc+ , reconstructed with a dedicated geant4 simulation of vertex and tracking detectors designed for EicC. The Λc+/D0 ratios as functions of multiplicity and pT , as well as the D0 double ratio are presented with projected statistical precision.
作者机构：
[Wenchang Xiang] Guizhou University of Finance and Economics, guiyang, Guangzhou, Guangdong, 510725, CHINA;[Dewen Cao] , 101 Hongshan 3rd Street, Guangzhou, 510725, CHINA;[Dai-Cui 周代翠 Zhou#&#&#Wenchang Xiang] 华中师范大学粒子物理研究所, 武汉华中师范大学粒子物理研究所, Wuhan, 430070, CHINA
摘要：
We use the refined hot spot model to study the valence quark shape of the proton with the deeply virtual Compton scattering at high energies in the framework of Color Glass Condensate. To investigate the individual valence quark shape, a novel treatment of the valence quark width is employed. We calculate the cross-sections for the coherent and incoherent deeply virtual Compton scattering, for the first time, by using different widths ($\mathrm{B_u}$ and $\mathrm{B_d}$) for the profile density distributions of the up and down quarks instead of using the same width as in the literature. We find that the cross-sections calculated with $\mathrm{B_u \geq B_d}$ at each collision energy are consistent with each other themselves, which is in agreement with the theoretical expectations. While the ones computed with $\mathrm{B_u < B_d}$ show some discrepancies. This outcome implies that the up quark might emit more gluons than the down quark leading to $\mathrm{B_u \geq B_d}$ at high energy. The energy impact of the outcome is estimated. Our results show that as the collision energy increases, the aforementioned discrepancies are not only significantly broadened, but also shift to relatively smaller momentum transfer range in the future Electron-Ion Collider (EIC) and Large Hadron Electron Collider (LHeC) energies, which indicates that the EIC and LHeC can provide unprecedented chance to access the shape of the valence quark of the proton.
作者机构：
PLAC, Key Laboratory of Quark and Lepton Physics (MOE), Central China Normal University, Wuhan 430079, China;Hubei Provincial Engineering Research Center of Silicon Pixel Chip & Detection Technology, Wuhan 430079, China;[Jiemiao Wang; Hulin Wang; Dongliang Zhang] PLAC, Key Laboratory of Quark and Lepton Physics (MOE), Central China Normal University, Wuhan 430079, China<&wdkj&>Hubei Provincial Engineering Research Center of Silicon Pixel Chip & Detection Technology, Wuhan 430079, China
摘要：
In this study, the possibility of observing a solar neutrino background in a future neutrinoless double beta decay experiment using a high-pressure gaseous 82SeF6 TPC is investigated. Various contributions are simulated, and possible features that could be used for event classification are discussed; two types of backgrounds are identified. The rate of multi-site background events is approximately 0.63 events/(ton·yr) in a 30 keV ROI window. This background could be effectively reduced to less than 0.0001 events/(ton·yr) (95% C.L.) while maintaining a high signal efficiency of 93% by applying a selection based on the number of clusters and energy of the leading cluster. The rate of the single-electron background events is approximately 0.01 events/(ton·yr) in the ROI. Assuming a reduction factor of 10 for the single-electron background events obtained via the algorithms developed for radioactive background rejection, the total background induced by the solar neutrino would be 0.001 events/(ton·yr), which is sufficiently small for conducting ton-level experiments.
作者机构：
[He, Xingchen; Liu, Yanliang; Liu, Wenjun; Li, Dong; Shi, Tongyu; Wang, Jiahong; Huang, Hao] Materials Interfaces Center, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, Guangdong, China;[Gao, Chaosong; Wu, Meng] Key Laboratory of Quark and Lepton Physics, Central China Normal University, Wuhan, Hubei, China;[Zhang, Xin; Zhu, Jiongtao; Liang, Dong] Research Center for Medical Artificial Intelligence, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, Guangdong, China;[Sheng, Zonghai; Liang, Dong] Paul C Lauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, Guangdong, China;[Sheng, Zonghai; Liang, Dong] Key Laboratory of Biomedical Imaging Science and System, Chinese Academy of Sciences, Shenzhen, Guangdong, China
通讯机构：
[Zheng, Hairong] P;[Yu, Xue-Feng] M;[Ge, Yongshuai] R;[Sun, Xiangming] K;Materials Interfaces Center, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, Guangdong, China. xf.
摘要：
High performance X-ray detector with ultra-high spatial and temporal resolution are crucial for biomedical imaging. This study reports a dynamic direct-conversion CMOS X-ray detector assembled with screen-printed CsPbBr(3), whose mobility-lifetime product is 5.2 × 10(-4) cm(2) V(-1) and X-ray sensitivity is 1.6 × 10(4) µC Gy(air)(-1) cm(-2). Samples larger than 5 cm[Formula: see text]10 cm can be rapidly imaged by scanning this detector at a speed of 300 frames per second along the vertical and horizontal directions. In comparison to traditional indirect-conversion CMOS X-ray detector, this perovskite CMOS detector offers high spatial resolution (5.0 lp mm(-1)) X-ray radiographic imaging capability at low radiation dose (260 nGy). Moreover, 3D tomographic images of a biological specimen are also successfully reconstructed. These results highlight the perovskite CMOS detector's potential in high-resolution, large-area, low-dose dynamic biomedical X-ray and CT imaging, as well as in non-destructive X-ray testing and security scanning.
作者机构：
[Xu, Hongbo; Wang, Dong; Huang, Xingxing; Li, Ruijie; Chen, Yun; Zhang, Guoping] Cent China Normal Univ, Coll Phys Sci & Technol, Wuhan 430079, Peoples R China.
通讯机构：
[Zhang, GP ] C;Cent China Normal Univ, Coll Phys Sci & Technol, Wuhan 430079, Peoples R China.
关键词：
Integrated sensing and communication (ISAC);Outage probability (OP);Power optimization;Non -orthogonal multiple access (NOMA)
摘要：
This paper investigates a framework for integrated sensing and communication (ISAC) based on non-orthogonal multiple access (NOMA). In this framework, a dual-function base station (BS) utilizes NOMA technology to send superimposed signals to various users, and this superimposed signal also acts on target sensing simultaneously. Considering the channel estimation error, ensuring the communication performance, and sensing performance requirements, a transmit power optimization problem of ISAC system using NOMA is studied. Specifically, in the statistical channel state information (CSI) error model, the total system communicate power is minimized while ensuring all single users' rate outage probability (OP) constraints and the requirements for the beampattern gains of all single radar targets. Unfortunately, the proposed problem is challenging to solve and non-convex. But we have devised a feasible way to deal with this problem. First, we use Bernstein inequality to transform the rate OP constraint, and this challenging non-convex problem is then successfully solved using a method based on semi-definite relaxation (SDR). The numerical outcomes demonstrate that the system's overall transmission power will increase due to the channel estimation error. The numerical findings also show that the ISAC system performs better with NOMA assistance than with OMA when comparing the NOMA and OMA schemes.
期刊：
Neural Computing and Applications,2024年 ISSN：0941-0643
通讯作者：
Qu, SC
作者机构：
[Qu, Shaocheng; Zheng, Wei; Tang, Qian] Cent China Normal Univ, Coll Phys Sci & Technol, Wuhan 430079, Hubei, Peoples R China.;[Zheng, Wei] Kashi Univ, Coll Phys & Elect Engn, Kashi 844000, Xinjiang, Peoples R China.
通讯机构：
[Qu, SC ] C;Cent China Normal Univ, Coll Phys Sci & Technol, Wuhan 430079, Hubei, Peoples R China.
关键词：
Cellular neural network;Memristor;Sliding mode control;Chaotic synchronization;Secure communication
摘要：
The purpose of this research is to investigate the synchronization and control for uncertain memristor-based cellular neural network and its application in secure communication. To address the issue, a novel sliding mode function is designed, on which the system states can effectively converge to the equilibrium point after reaching the sliding mode surface. Moreover, the corresponding controller is constructed by employing the proposed sliding mode function. The proposed control strategy achieves the synchronization of the uncertain memristor-based cellular neural network, and effectively addresses the integral saturation existing in traditional one. In addition, the control performance, including convergence speed, control accuracy, robustness and security, are significantly enhanced. Furthermore, the stability of the system is discussed based on Lyapunov theory. Finally, comparative tests and application examples are presented to verify the effectiveness of the proposed scheme.
摘要：
In mm -wave massive multiple -input -multiple -output (MIMO) systems, accurate channel state information (CSI) at the base station (BS) is the key knowledge to obtain the performance gain. Consequently, the user is not only required to complete the channel estimation, but also to feedback CSI to the BS. However, large-scale antenna arrays result in a substantial feedback overhead, which poses a challenging issue. Furthermore, the accuracy requirements of channel estimation and CSI feedback depend on the computing capacity of the user. In this paper, we propose a joint channel estimation and multiple -compression -rate feedback (JCEMF) scheme, and adopt centralized learning (CL) and federated learning (FL) strategies for the scheme. According to the limited computational resources available to users, the JCEMF scheme enables various lengths of feedback bits to change the feedback overhead. Additionally, the users in FL train the local models using their own datasets and upload the local model updates to the BS, thereby reducing communication overhead and protecting data privacy. Specifically, an estimation network is designed for the user to estimate the channel from the received signal. In the CSI feedback process, we introduce an MCRF network, which can achieve CSI compression and reconstruction with different numbers of feedback bits. Simulation results verify that the proposed approach shows good performance of joint channel estimation and multiple -compression -rate CSI feedback in different channel conditions.
摘要：
Semi-supervised classification has gained widespread popularity because of their superior ability to handle unlabeled samples in practical problems. This paper has presented a novel estimation error-ranked LSSVM method with double Mahalanobis-kernel which is used for semi-supervised classification. The main point is to construct two Mahalanobis distances in Hilbert space to form double Mahalanobis-kernel by considering the relationship between the characteristics of two sorts of samples, so as to reduce the influence of non-informational dimensions. Furthermore, the implementation of the proposed method is required to solve the label security problem of unlabeled samples. The unlabeled sample with the minimum evaluated error is selected for labeling, which effectively ensures the accuracy of the unlabeled sample labeling. This method not only considers the similarity of sample features, but also focuses on the security of unlabeled samples. And based on the experimental results of four artificial data sets and several UCI data sets, it verifies the effectiveness of the semi-supervised method with double Mahalanobis-kernel. Especially considering the experimental results of five disease diagnosis data sets, it demonstrates the potential of the proposed semi-supervised classification method in medical diagnosis.
作者机构：
School of Physics and Optoelectronic Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, China;[Zhiming Li; Yuanfang Wu] Key Laboratory of Quark and Lepton Physics (MOE) and Institute of Particle Physics, Central China Normal University, Wuhan 430079, China;[Ye-Yin Zhao] School of Physics and Electronic Engineering, Sichuan University of Science and Engineering (SUSE), Zigong 643000, China;[Yunshan Cheng; Gang Wang] Department of Physics and Astronomy, University of California, Los Angeles, California 90095, USA;[Lizhu Chen] School of Physics and Optoelectronic Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, China<&wdkj&>Key Laboratory of Quark and Lepton Physics (MOE) and Institute of Particle Physics, Central China Normal University, Wuhan 430079, China
摘要：
Hyperorder cumulants C5/C1 and C6/C2 of net-baryon distributions are anticipated to offer crucial insights into the phase transition from quark-gluon plasma to hadronic matter in heavy-ion collisions. However, the accuracy of C5 and C6 is highly contingent on the fine shape of the distribution's tail, the detectable range of which could be essentially truncated by low statistics. In this paper, we use the fast Skellam-based simulations, as well as the ultrarelativistic quantum molecular dynamics model, to assess the impact of limited statistics on the measurements of C5/C1 and C6/C2 of net-proton distributions at lower energies available at the BNL Relativistic Heavy Ion Collider. Both ratios decrease from the unity baseline as we reduce statistics and could even turn negative without a pertinent physics mechanism. By incorporating statistics akin to experimental data, we can replicate the net-proton C5/C1 and C6/C2 values comparable to the corresponding measurements for Au+Au collisions at sNN=7.7 , 11.5, and 14.5 GeV. Our findings underscore a caveat to the interpretation of the observed beam energy dependence of hyperorder cumulants.
作者机构：
[Xing, Hongxi; Anderle, Daniele Paolo; Anderle, DP; Zhao, Yuxiang] South China Normal Univ, Inst Quantum Matter, Key Lab Atom & Subatom Struct & Quantum Control, MOE, Guangzhou 510006, Peoples R China.;[Xing, Hongxi; Anderle, Daniele Paolo; Anderle, DP] South China Normal Univ, Inst Quantum Matter, Guangdong Prov Key Lab Nucl Sci, Guangzhou 510006, Peoples R China.;[Xing, Hongxi; Anderle, Daniele Paolo; Anderle, DP] South China Normal Univ, Southern Nucl Sci Comp Ctr, Guangdong Hong Kong Joint Lab Quantum Matter, Guangzhou 510006, Guangdong, Peoples R China.;[Guo, Aiqiang; Liang, Yutie; Ma, Yuming; Zhao, Yuxiang] Chinese Acad Sci, Inst Modern Phys, Lanzhou 730000, Gansu, Peoples R China.;[Guo, Aiqiang; Liang, Yutie; Zhao, Yuxiang] Univ Chinese Acad Sci, Beijing 100049, Peoples R China.
通讯机构：
[Anderle, DP ] S;South China Normal Univ, Inst Quantum Matter, Key Lab Atom & Subatom Struct & Quantum Control, MOE, Guangzhou 510006, Peoples R China.;South China Normal Univ, Inst Quantum Matter, Guangdong Prov Key Lab Nucl Sci, Guangzhou 510006, Peoples R China.;South China Normal Univ, Southern Nucl Sci Comp Ctr, Guangdong Hong Kong Joint Lab Quantum Matter, Guangzhou 510006, Guangdong, Peoples R China.
摘要：
The Electron -Ion Collider in China (EicC) has been proposed to study the inner structure of matter and fundamental laws of strong interactions. In this paper, we will present a conceptual design of the tracking system based on the state -of -art silicon detector and micropattern gaseous detector at the EicC and demonstrate that it will enable us to reconstruct charm hadron with good significance, hence study gluonic parton distribution functions in nucleons and nuclei, as well as gluon helicity distributions. The impact study using reweighting techniques shows that the impact of the EicC will be mainly in the large x region. It complements similar physics programs at the Electron -Ion Collider at Brookhaven National Laboratory.
作者机构：
[Xuening Li; Ying Xie; Zhiqiu Ye; Weifang Huang; Lijian Yang; Xuan Zhan; Ya Jia] Department of Physics, Central China Normal University, Wuhan 430079, China
通讯机构：
[Ya Jia] D;Department of Physics, Central China Normal University, Wuhan 430079, China
摘要：
In recent years, the coexistence of different states in the neural system has attracted widespread interest. Researchers have found a coexisting state of spiking and resting in homogeneous networks, which is known as the chimera-like state. The real cortical network is a much more complex and heterogeneous network. Therefore, the excitatory-inhibitory cortical neuronal network is constructed based on Hodgkin-Huxley neuronal model in this paper, and the chimera-like state is further investigated in the heterogeneous network. It is found that the chimera-like state is related to the balance between excitatory and inhibitory synaptic currents. The excitatory coupling current can counteract the initial condition effect and promote synchronized firing of neurons in the network. The inhibitory coupling current desynchronizes the network and thus induces synaptic noise, resulting in an inverse bell-shaped dependence of the change in the number of spiking neurons. We analyzed the underlying mechanisms of synaptic noise in the phase plane diagram and found it has asymmetry for the neuronal state transition. In addition, neurons with low degrees have a higher probability of undergoing state transitions. Finally, we verified that the chimera-like state is robust to network topology and initial conditions. The results provide a new insight into neuronal interactions in heterogeneous networks and might help to reveal the mechanisms of coexistence of different states in the cortical network.
期刊：
Journal of Alloys and Compounds,2024年984:173913 ISSN：0925-8388
通讯作者：
Yiwen Tang<&wdkj&>Shiyu Wang
作者机构：
[Yue Ming; Yue Cheng; Jinghua Jiang; Jintao Wang; Haizhou Qv; Yan Xue; Shiyu Wang] College of Electronic and Communication Engineering, Shenzhen Polytechnic University, Shenzhen 518055, China;[Wenjian Shen; Hongbing Ran; Yue Zhao; Yiwen Tang] Department Nano–Science & Technologym, College of Physics and Technology, Central China Normal University, Wuhan 430079, China;[Jiale Liu] Wuhan National Laboratory for Optoelectronics, Huazhong University of Science and Technology, Wuhan 430074, China
通讯机构：
[Yiwen Tang] D;[Shiyu Wang] C;Department Nano–Science & Technologym, College of Physics and Technology, Central China Normal University, Wuhan 430079, China<&wdkj&>College of Electronic and Communication Engineering, Shenzhen Polytechnic University, Shenzhen 518055, China
摘要：
Trap–mediated non–radiative loss is a major efficiency and stability obstale for solution–processed perovskite solar cells (PSCs). To address this issue, defect passivation using organic molecules has been identified as an effective strategy due to the ionic nature of the perovskite lattice. Herein, Ostwald ripening by using formamidine iodine (FAI) is reported to inhibit the carrier trap induced recombination via promoting secondary grain growth at perovskite surface. Based on secondary grain growth with FAI treatment, the graded passivation was presented caused by FA+ and I- from FAI. Moreover, the findings from time resolved photoluminescence and space charge–limited current technology indicate enhancements in the crystal quality. Consequently, this approach yielded a remarkable power conversion efficiency (PCE) of 20.62% for the FAI modified MAPbI3 photovoltaics, with significantly improved open–circuit voltage ratio of 3.10%, short–circuit density ratio of 6.11%, fill factor ratio of 9.46% and PCE ratio of 19.06%. Moreover, the device show the excellent stability. This study introduces simple and reproducible procedures for fabricating PSCs, which can be readily extended to diverse optoelectronic device applications.
作者机构：
[Xueyan Hu; Yong Wu; Qianming Ding; Ying Xie; Zhiqiu Ye; Ya Jia] Department of Physics, Central China Normal University, Wuhan 430079, China
通讯机构：
[Ya Jia] D;Department of Physics, Central China Normal University, Wuhan 430079, China
摘要：
Spike-timing-dependent plasticity (STDP) is one of the important rules for the change of synaptic weights between neurons in biological nervous systems. In this paper, we study the effect of STDP on the synchronization phenomenon induced by time delay in the neuronal network which is the scale-free network with small-world property, and nodes of the network are constructed by Izhikevich neuron and connected by chemical synapses. For appropriate time delay values, there exists an optimal range of STDP maximum weight value in which the synchronization of the network is better, and in addition the synchronization is decreased with the increasing of STDP maximum weight value. The network with high synchronization has a centralized distribution of synaptic weights within it, while conversely, an unsynchronized network has a more discrete distribution of synaptic weights. When the STDP maximum weight value is too small, the collective firing pattern of network is not affected by synaptic current, and the synchronization of the network is also not affected. Interestingly, comparing with the small-world network and the scale-free network, it is found that a network has a smaller range of optimal STDP maximum weight values when the network is of larger average clustering coefficient, shorter average shortest path length, and higher small-world property. Our results can illuminate the potential significance of STDP for information processing and transmission in the nervous system.
作者机构：
[Jia-Hao Wang] College of Science, China Three Gorges University, Yichang 443002, China;Center for Astronomy and Space Sciences and Institute of Modern Physics, China Three Gorges University, Yichang 443002, China;Key Laboratory of Quark and Lepton Physics (MOE) and Institute of Particle Physics, Central China Normal University, Wuhan 430079, China;[Sheng-Qin Feng] College of Science, China Three Gorges University, Yichang 443002, China<&wdkj&>Center for Astronomy and Space Sciences and Institute of Modern Physics, China Three Gorges University, Yichang 443002, China<&wdkj&>Key Laboratory of Quark and Lepton Physics (MOE) and Institute of Particle Physics, Central China Normal University, Wuhan 430079, China
通讯机构：
[Sheng-Qin Feng] C;College of Science, China Three Gorges University, Yichang 443002, China<&wdkj&>Center for Astronomy and Space Sciences and Institute of Modern Physics, China Three Gorges University, Yichang 443002, China<&wdkj&>Key Laboratory of Quark and Lepton Physics (MOE) and Institute of Particle Physics, Central China Normal University, Wuhan 430079, China
摘要：
The impact of rotation on the deconfinement phase transition under the Einstein-Maxwell system of the soft and the hard wall models in holographic quantum chromodynamics is studied in this paper. The metric by cylindrical coordinates with rotation is introduced into the system to calculate the Hawking temperature. The first holographic study on the influence of the radius of a homogeneous rotating system on the phase diagram is proposed. It is found that the phase transition temperature hardly changes with the rotation angular velocity for a small rotation radius. Only with a larger rotation radius can the change in rotational angular velocity significantly alter the phase transition temperature. The phase transition temperature decreases rapidly with the increase of rotation angular velocity as the rotation radius increases.
期刊：
Journal of Statistical Mechanics: Theory and Experiment,2024年2024(2) ISSN：1742-5468
通讯作者：
Marsili, M
作者机构：
[Xie, Rongrong] Cent China Normal Univ CCNU, Key Lab Quark & Lepton Phys MOE, Wuhan, Peoples R China.;[Xie, Rongrong] Cent China Normal Univ CCNU, Inst Particle Phys, Wuhan, Peoples R China.;[Marsili, Matteo; Marsili, M] Abdus Salam Int Ctr Theoret Phys, Quantitat Life Sci Sect, I-34151 Trieste, Italy.
通讯机构：
[Marsili, M ] A;Abdus Salam Int Ctr Theoret Phys, Quantitat Life Sci Sect, I-34151 Trieste, Italy.
关键词：
learning theory;machine learning
摘要：
We discuss the concept of probabilistic neural networks with a fixed internal representation being models for machine understanding. Here, 'understanding' is interpretted as the ability to map data to an already existing representation which encodes an a priori organisation of the feature space. We derive the internal representation by requiring that it satisfies the principles of maximal relevance and of maximal ignorance about how different features are combined. We show that, when hidden units are binary variables, these two principles identify a unique model-the hierarchical feature model-which is fully solvable and provides a natural interpretation in terms of features. We argue that learning machines with this architecture possess a number of interesting properties, such as the continuity of the representation with respect to changes in parameters and data, the possibility of controlling the level of compression and the ability to support functions that go beyond generalisation. We explore the behaviour of the model with extensive numerical experiments and argue that models in which the internal representation is fixed reproduce a learning modality which is qualitatively different from that of traditional models, such as restricted Boltzmann machines.
期刊：
Journal of Instrumentation,2024年19(03):C03031 ISSN：1748-0221
作者机构：
Institute of Modern Physics, Chinese Academy of Sciences, Lanzhou 730000, China;School of Nuclear Science and Technology, University of Chinese Academy of Sciences, Beijing 100049, China;[T. Liang; C. Gao; D. Zhang; K. Chen; H. Wang; J. Liu; L. Liu; Y. Qiao; Z. Hu; X. Sun; G. Huang] PLAC, Key Laboratory of Quark and Lepton Physics (MOE), Central China Normal University, Wuhan 430079, China;[X. Tian] School of Resources, Environmental Science and Engineering, Hubei University of Science and Technology, Xianning 437100, China;[Z. Song] College of Electronic Engineering, Naval University of Engineering, Wuhan 430033, China
摘要：
This paper presents the design and electrical test results of a low-noise front-end chip (named Topmetal-S) in a High-pressure Time Projection Chamber (TPC) for searching the neutrinoless double beta decay. The Topmetal-S has been fabricated in a 130 nm CMOS technology. The proposed front-end chip consists of a charge collection electrode, a Charge Sensitive Amplifier (CSA) and peripheral circuits. The test results indicate that the CSA features an input linear dynamic range of approximately 6.64 fC, a charge-conversion gain of about 220 mV/fC and an Equivalent Noise Charge (ENC) of approximately 115 e- after a digital trapezoidal pulse shaper.
作者机构：
[Wu, Yanwen; Ge, Di] Cent China Normal Univ, Sch Phys Sci & Technol, 152 Luoyu Rd, Wuhan 430079, Peoples R China.;[Dong, Zheng] Beijing Bytedance Technol Co Ltd, 48 Zhichun Rd, Beijing 200000, Peoples R China.;[Cheng, Yuhang] SHAANXI GSXZ Technol Co Ltd, 57 Fengchan Rd, Xian 710061, Shaanxi, Peoples R China.;[Wu, Yanwen] Cent China Normal Univ, Natl Digital Learning Engn Technol Res Ctr, 152 Luoyu Rd, Wuhan 430079, Peoples R China.
通讯机构：
[Wu, YW ] C;Cent China Normal Univ, Sch Phys Sci & Technol, 152 Luoyu Rd, Wuhan 430079, Peoples R China.
关键词：
Multivariate time series;Spatio-temporal modeling;Graph contrastive learning;Unsupervised anomaly detection
摘要：
Anomaly detection using multivariate time series plays a crucial role in system security. Conventional deep learning detection techniques mainly depend on temporal dependency and employ reconstruction or prediction-based methods. However, as feature variables grow more intricate, there is a risk of neglecting essential spatio-temporal structural information, potentially leading to insufficient model training in unsupervised settings. Hence, we propose an end-to-end anomaly detection model with multiple pre-training tasks designed for the spatio-temporal dimension to enhance our constraints. Specifically, in the temporal dimension, we employ an autoregressive task to train timestamp associations using data’s concealed autocorrelation and periodicity. In the spatio dimension, we acquire knowledge of a diverse feature-related heterogeneous graph. Subsequently, we design three different graph contrastive learning tasks to tap into the effective information arising from the inherent heterogeneity and hierarchy in spatio structures. Through joint spatio-temporal modeling, we can effectively capture inter and intra-feature associations from series and graph structural features, enhancing model robustness to cope with the complex chain reactions between features. Finally, we assess our model on three real-world datasets: SWaT, WADI(2017, 2019), our F1 scores demonstrate enhancements of 6.17%, 18.3% and 5.35% over the top-tier baseline performance. Our model is applicable for both temporal and graph, is self-supervised learning for sparse data which is suitable for data sparsity and complex scenarios that need to capture spatio-temporal characteristics at the same time, for example, traffic flow detection and anomaly detection of intelligent systems. Further visualization experiments and case studies will provide a better interpretation of our model.
