重离子裂变反应动力学研究

从实验角度评述了重离子裂变反应动力学研究现状．内容包括动力学裂变宽度、裂变时标、核粘滞性的温度和形变相关性以及从裂变实验上研究核粘滞性的方法等． The present status of the dynamics study for heavy ion induced fission reaction includesing the dynamical fission width,the fission time scale,the temperature and deformation dependence of the nuclear viscosity and the methods of the extracting viscosity from fission experiments induced by heavy ion reactions is reviewed.      

光中子源上吸收片功能研究

15 MeV电子直线加速器驱动的光中子源装置,将用于中国科学院战略性先导科技专项“钍基熔盐堆”中的核数据初步测量工作、中子探测器的研制和反应堆相关材料的辐照研究等。光中子源的中子能谱是连续的,中子能量通过中子飞行时间法测量得到,需要利用吸收片确认中子吸收峰,刻度飞行时间,计算等效飞行距离,扣除实验本底等,而实验本底的扣除对最终总截面计算有很大的影响。因此通过Geant4蒙特卡罗模拟软件构建了包括中子源、吸收片在内的模拟实验环境;研究了不同吸收片的吸收谱和吸收片厚度的关系,同理论计算值进行了比较,给出了推荐的吸收片厚度值;模拟计算了中子飞行时间谱,并和实验测量结果比较,确定中子等效飞行距离为5.70 m。Geant4的理论计算也可以模拟出多吸收片本底函数曲线,可用于实验数据的本底扣除和误差分析。实验测量、模拟分析以及理论公式计算的吸收片厚度和中子飞行时间参数得到了完全一致的结果,验证了实验测量的可靠性。A photo-neutron source driven by a 15 MeV electron LINAC is built for the "Strategic Priority Research Program" of the Chinese Academy of Sciences-the "thorium-based molten salt reactor" project to conduct the nuclear data measurement work,develop neutron detector and carry out reactor material irradiation studies.Since the neutron energy spectrum is continuous,the neutron energy is measured by the time of flight (TOF) method,and neutron filters are needed to confirm absorption peaks,calibrate the TOF,calculate the equivalent flight distance,and remove the experimental background which has great influence on the calculation accuracy of the total cross section.Based on the Monte Carlo simulation tool,Geant4 a simulation environment is set up,including neutron source and neutron filters,to study the energy absorption spectra and thickness of different filters and recommended data for the thickness are provided.The neutron TOF spectra are simulated and compared with experimental measurement,deciding the equivalent TOF distance to be 5.7 m.Geant4 can also simulate the background curve of multiple filters and be used to remove background and analyze errors for the experimental data.All the experiments,simulation and theoretical calculation show consistent results on filter thickness and neutron TOF parameters,indicating the accuracy of the measurement.     

20 MeV以下65Cu(n,p)65Ni 反应截面及其协方差数据的评价

为满足核工程对核数据的需求,对20 MeV以下65Cu(n, p)65Ni 反应截面的实验数据进行了分析和评价,包括实验数据的收集、修正、选取、归一、误差调整和数据处理,在分析实验误差特别是关联误差的基础上构造了每家数据的协方差矩阵,最后利用样条拟合程序SPCC给出了评价截面数据和实验数据的协方差矩阵。评价不仅考虑了实验数据的误差,还考虑了数据间的相关性。To meet the needs for nuclear data engineering, the experimental data of the 65Cu(n, p)65Ni cross section below 20 MeV were analyzed and evaluated.The evaluation procedure includes the experimental data collection, correction, selection, normalization, error adjustment and data processing, especially the covariance matrix for each experimental data is constructed with the information on experimental errors and correlation errors. The evaluation cross section data and the corresponding covariance matrices were given using the spline fitting procedure SPCC. The evaluation work was done with the consideration of not only the experimental errors but also the correlation of data.     

400~1500 MeV质子轰击铅靶和钨靶的出射中子能谱的FLUKA和Geant4模拟研究

在ADS散裂靶系统的优化设计中,蒙特卡罗方法结合可靠的散裂反应理论模型进行中子学计算具有重要的作用。本工作利用Geant4程序中的INCLXX模型、BIC模型以及BERT模型和FLUKA程序分别模拟了597 MeV和1 500 MeV质子轰击薄铅靶不同出射角度的中子双微分截面,500,1 500 MeV质子轰击厚铅靶不同出射角度的中子双微分产额,以及400,600,800,1 000和1 200 MeV质子轰击厚钨靶在反角方向（175 °）的中子双微分产额,并与实验数据进行比较。研究表明,对于薄铅靶,Geant4程序的INCLXX模型和FLUKA程序模拟结果与实验符合得更好。能量在10～40 MeV范围内,BIC模型模拟结果明显高于实验数据,而BERT模型模拟结果略微低于实验数据。对于厚铅靶,在40 MeV左右所有的模拟结果都低于实验数据。对于厚钨靶,Geant4程序的BIC模型和FLUKA程序与实验数据符合得较好,INCLXX模型在能量高于60 MeV时模拟结果低于实验数据,BERT模型与实验数据差异较大。总体来看,Geant4程序的INCLXX模型和FLUKA程序进行ADS散裂靶相关的中子学的计算是合理和可靠的。The reliable Monte Carlo simulation codes coupled with nuclear reaction models play an important role in the neutronic calculation for the design and optimization of the ADS spallation target. In this work, the double differential cross sections at different angles produced from a thin lead target bombarded with 597 and 1 500 MeV protons, the neutron energy spectra at different angles produced from a thick lead target bombarded with 500 and 1 500 MeV protons, and the neutron energy spectra in the backward direction(175°) produced from a thick tungsten target bombarded with 400, 600, 800, 1 000 and 1 200 MeV protons are calculated with the Geant4 code coupled INCLXX, BIC and BERT models and the FLUKA code. The calculations are compared with the available experimental data. The results show that, for the thin lead target, the calculations with the Geant4 coupled INCLXX model and FLUKA code well reproduce the experimental results. In a energy range from 10 to 40 MeV, BIC model obviously overestimates the experimental results, and BERT model slightly underestimates the experimental results. For the thick lead target, all of the calculations underestimate the experimental results around 40MeV. For the thick tungsten target, the Geant4 coupled BIC model and FLUKA code well reproduce the experimental results. INCLXX model underestimates the experimental results above 60 MeV. BERT model bad reproduces the experimental results. Overall, the neutronic calculations with the Geant4 code coupled INCLXX model and FLUKA code for the ADS spallation target is reasonable and reliable.     

金纳米粒子的放射增敏效应研究

系统阐述了与金纳米粒子(GNPs) 放射增敏效应相关实验的方法与结果、影响GNPs 放射增敏的因素、GNPs 放射增敏的细胞和动物实验表现及其相关机制。同时结合相关实验,分析和比较了15 nm 柠檬酸钠包被的GNPs 的放射增敏效应,发现GNPs 在高LET 的碳离子束和低LET 的X射线辐照下对Hela细胞的杀伤力随其浓度的增加而增大;在50% 的细胞存活率下,当GNPs 的质量浓度为7.5 g/mL时,其X射线的剂量减少率和碳离子的相对生物学效应值(RBE) 的增加率达到了最大,分别为65.3% 和43.6%,同时GNPs 共培养细胞24 和48 h 后,未出现细胞周期同步化的现象。This paper describes the methods and results of the previous experiments, the experimental phenomena of the cell and animal tests and the relative mechanisms on the radiosensitizing effect of GNPs. Together with our experiments, the radiosensitizing effects of 15 nm citrate-capped GNPs and related mechanisms are analyzed and compared, finding that Hela cell killing of GNPs increase along with their concentration after exposure to high- and low-LET radiation such as carbon ions and X-rays. In addition, the percentages of dose reduction of the X-rays and RBE increment of the carbon ions reached their maximums 65.3% and 43.6%, respectively,at 50% survival level when Hela cells were pre-treated with 7.5 g/mL GNPs. Moreover, Hela cells showed no cell-cycle synchronization after 24 and 48 h exposure to GNPs.     

高电离态原子光谱测量

简要介绍了兰州重离子加速器上的高电离态原子光谱实验测量和取得的部分实验结果.The excitation spectra and level lifetimes of highly charged atoms for heavier elements of Ne, S and Ar have been measured by using the beam foil method on the heavy ion accelerator in Lanzhou. Emission spectra in the wavelength region of 25-400  was studied. On the whole, the agreement between our experimental results and other experiments is fairly good.     

时间分辨原子发射光谱诊断脉冲放电气体束等离子体

建成了一套脉冲放电气体束和原子发射光谱等离子体诊断实验装置,利用这套实验装置测量了不同放电条件下等离子体的时间分辨发射光谱,并采用玻尔兹曼作图法和Hα谱线斯塔克展宽法研究了等离子体的原子温度、离子温度和自由电子密度等参数的演化。实验结果表明,脉冲放电的总电量对等离子体参数的演化有较大影响;脉冲放电气体束等离子体中的离子温度远高于原子温度,自由电子密度相对较低,等离子体处于非局域热平衡状态。A special designed pulsed discharge nozzle (PDN) ion source and a plasma diagnostics system based on the atomic emission spectroscopy were constructed. The time-resolved emission spectra of Argon atoms and ions in the region of 300-800 nm were observed and analyzed. The plasma temperatures, including atomic and ionic temperatures, were simulated by Boltzmann plot method, and the free electron density was simulated through the Stark broadening of Hα line. The evolution of these plasma parameters were investigated by high-resolution time-resolved emission spectra. The results show that the total energy struck on the pulsed gas beam is the most important factor which determines the plasma properties and its evolution. The plasma in PDN is concluded into non-local thermodynamic equilibrium (non-LTE) during the evaluation because of the low electron density and the big difference between the temperatures of Argon atoms and ions in all discharge conditions.     

兰州冷却储存环上可开展的强子物理研究

介绍了当前强子物理的研究现状和兰州冷却储存环的能量特点,以及国内强子物理专家的分析和建议。在兰州冷却储存环上,可利用中能轻离子束核反应产生强子激发态研究强子内部夸克态的结构和性质、强子性质随核环境的变化和手征对称破缺与部分恢复。尤其是通过兰州冷却储存环上限能区附近的P＋P反应,研究奇异夸克的不对称性和形状因子,寻找超子激发态和pentaquark的实验证据,发现双重子态的实验事例。 According to both the development on badrons physics and the aspect of Lanzhou cooling storage ring （CSR） and based on the analysis and propositions given by experts in China, we propose some hadrons physics program at CSR. The hadron spectroscopy produced in light nucleus collisions at CSR used to probe the quark and gluon structure of hadrons, to study the modification of the hadron properties in nuclear matter and to investigate the spontaneous breaking of the chiral symmetry and its partial restoration. Especially, the proton-proton collisions at beam energies per proton below 2.8 GeV at CSR should be used to measure the strangeness asymmetry and strange form lector, to probe the existence of hyperon and pentaquarks and to find the evidence for the existence of dibaryon.     

微束技术在放射生物学中的应用

微束装置可以为生命科学研究提供微米定位、剂量特定的电离辐射,在生物体内的电离辐射靶物质及其敏感度、靶物质的损伤及修复机制研究中具有独特的作用。概述了生物微束装置和实验技术的发展及其在低剂量辐射效应、旁观者效应、信号传导研究中的主要应用;介绍了中国科学院近代物理研究所（IMP）重离子微束装置,该装置可以提供能量7~ 80 MeV/u、传能线密度为30~ 3000 keV/μm的重离子微束,实现了活细胞辐照和在线观察、小鼠定位辐照的实验技术;利用IMP微束装置在重离子诱导旁效应实验、小鼠下丘脑重离子辐照效应和DNA损伤快速修复动态等方面取得了一些实验成果。The microbeam facility can provide micrometer scale localized and predefined ionizing radiation in the life science study, and the microbeam techniques play a unique role in determining the target substances of ionizing radiation, as well as in the study of radiation sensitivity, mechanisms of radiation damage response and repair. This paper summarizes the technical developments of biological microbeam facilities and their applications in the studies of low-dose radiation effect, bystander effect and cellular signaling. This paper also introduces the recent developments at the heavy-ion microbeam facility in the Institute of Modern Physics (IMP), which can provide heavy ion microbeam irradiation with energy of 7~80 MeV/u and LET of 30~3000 keV/μm. The facility can perform radiobiological irradiation and online investigation in living cells and mice, including bystander effect study, sleeping system influence after irradiation to mice hypothalamus and the recruitment dynamics of XRCC1 protein.     

基于Bagging 算法的神经网络在粒子鉴别中的应用

分析了神经网络方法和bagging 算法在实验高能物理和核物理数据分析中的应用现状。分别对神经网络方法和bagging 算法的基本原理进行了介绍。以蒙特卡罗产生器产生的夸克胶子喷注样本为例,详细讨论了神经网络方法以及bagging 算法与神经网络结合对粒子鉴别中信号和背景区分问题的应用过程,并对结果进行了讨论和分析。实验结果表明,应用bagging 算法后,神经网络能够较大幅度地提高实验高能物理和核物理数据分析中粒子鉴别的精度,以及能够得到较高的信噪比。The paper presents the application of neural network and bagging algorithm in experimental high-energy physics and nuclear physics data analysis. Paper also introduces the basic principles of neural network method and bagging algorithm. We use the data samples of quark-gluon jets, which are generated by Monte Carlo generator, to solve the problem of discriminating signal events and background events by the combined algorithm of bagging algorithm and neural network. Experimental results show that, to apply bagging algorithm, neural networks can greatly improve the accuracy of the identification of particles in the experiments of high energy physics and nuclear physical data analysis,and also obtains a larger SNR (Signal to Noise Ratio).     

ICP-MS/ICP-AES测定小麦穗离体培养籽粒营养元素及重金属含量

环境中重金属污染会导致谷物营养缺乏、代谢紊乱.借助ICP-MS/ICP-AES系统和离体穗培养方法,准确测定灌浆期不同浓度铜、镉(0,5,15,45 mg·L-1)处理对籽粒营养元素及重金属元素含量的影响.结果显示,不司浓度铜、镉处理显著降低小麦粒重;外源铜促进了籽粒P,K,Mg,Ca,Na,Mn和Zn含量的增加,抑制了 Fe和B的吸收.镉的加入对小麦籽粒P,K,Ca,Mg和Mn的吸收起促进作用,然而对Na,Fe,Zn和B的吸收起拮抗作用.外源铜、镉抑制小麦籽粒对Hg的积累,但籽粒Cu和Cd含量最高分别超过国家食品卫生标准的55和62倍,存在极大安全隐患.     

塔河地区主要岩石金属元素含量、物性参数与光谱特征间的关系研究

测试了采集自黑龙江省塔河地区的300块并进行抛光处理后的岩石样本350~2 500nm间光谱反射率、磁化率、密度、孔隙率和金属元素含量(Fe,Mn,Ti,Zr,V,Zn,Pb,Nb,Co,Bi),并计算了其光谱吸收深度。在此基础上,以相关性分析方法为依据,探讨了所采集岩石样本的金属元素含量、物性参数、反射光谱间的特征响应关系,计算了岩石样本金属元素和光谱吸收深度间的相关性系数、物性参数与光谱反射率的相关性系数,获得成果如下:(1)在410nm附近,闪长玢岩各金属元素与吸收深度间的相关系数都存在尖锐的波峰和波谷,相关系数达到极值。(2)岩石样本金属元素和吸收深度的相关性研究中,侵入岩的相关系数则显著高于其他岩石类型。(3)1 400nm附近,岩石样本金属元素与吸收深度、各物性参数与光谱反射率的相关性都存在尖锐的波峰和波谷。其中磁化率、密度、视孔隙率与光谱反射率的相关系数在可见光范围内波动变化较大。(4)在1 900~2 500nm范围内,金属元素与光谱吸收深度、各物性参数与光谱反射率间的相关系数波动较大,其中金属元素和光谱吸收深度呈显著相关,相关系数达到极值。进一步研究了岩石金属元素和物理特性与其光谱特征的关系,对于不同岩性的不同波段的反射率与不同金属元素间分布状态的探测,具有一定意义。     

INAA for determination of trace elements in bottom sediments of the Selenga river basin in Mongolia

Instrumental neutron activation analysis (INAA) was used for the determination of major, minor and trace elements in samples of bottom sediments of the inflows of the Selenga river basin to assess the impact of the contamination from the industrial complex Erdenet and other industrial enterprises in Mongolia. A total of 42 elements (Na, Mg, Cl, K, Ca, Al, Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, As, Se, Br, Sr, Rb, Sr, Zr, Nb, Sb, Cs, Ba, La, Ce, Nd, Sm, Eu, Tb, Dy, Tm, Hf, Ta, W, Au, Hg, Th, and U) was determined by combination of conventional and epithermal neutron activation analysis at the IBR-2 reactor, FLNP JINR, Dubna. For the first time such a large set of elements was used for characterization of the bottom sediments as accumulating media which may reflect industrial contamination of the water basin. The concentrations of heavy metals and other trace elements in the samples from three inflows of the Selenga river basin were compared with data from the previous studies. It was shown that the industrial zone of the Erdenet Mining Corporation (EMC) and gold mining zone Zaamar are the sources of strong environmental contamination. The concentrations of Cu and Sb determined in sediment samples of the River Khangal and Govil near the EMC exceed average crustal rock and soil values by factors of 50 and 15, respectively. In the area of the gold mining zone Zaamar concentrations of Au, As, and Sb exceed crustal rock and soil values by factors of 4, 25, and 6, respectively. The relatively high levels of As, V, Zn, V, and Sr in the sediments of the studied rivers are obviously due to the discharges of untreated wastewater of desalination plant, electrical power station, textile industry and mining activities as well as domestic wastewater.     

ICP法和原子吸收光谱法测定褐马鸡羽毛中的10种元素

珍稀濒危鸟类-褐马鸡是中国特有的野生鸟类,是国家一级重点保护动物。采用ICP法和原子吸收光谱法对来自芦芽山、庞泉沟国家自然保护区的野生褐马鸡和太原市动物园圈养褐马鸡羽毛中Mo,Zn,Ni,Fe,Mn,Cr,Cu,K,Pb和Cd等10种元素的含量进行了测定,并对这3个不同地理分布区的样品中10种元素的含量进行了比较。结果表明,在所有羽毛样品中Fe元素含量最高,Mo和Cr含量较低,Cd未检出。动物园饲养的褐马鸡羽毛中除K和Cu外,其余八种元素的含量均低于芦芽山和庞泉沟国家自然保护区野生的褐马鸡羽毛中的含量。生境的不同对褐马鸡机体微量元素有一定影响,野生环境更有利于它们的生长发育。     

唐卡主色矿物颜料光谱分析

唐卡作为一件艺术品,具有较高的历史价值和艺术价值。对唐卡的矿物颜料进行鉴别分析,对唐卡的鉴定、修复、数字化存档、再现等具有非常重要的意义。该研究对唐卡主色矿物颜料进行体系性的光谱分析,选用唐卡绘制过程中5种主色常用的矿物颜料,深入分析矿物颜料的可见光、近红外、短波红外光谱特征产生机理,总结了不同色系矿物颜料可见光、近红外、短波红外谱段光谱特征。通过分析同一矿物颜料粉末、调和骨胶颜料以及颜料上布色卡光谱特征,发现粉状颜料调和骨胶后,反射率整体下降,在1 447和1 928 nm附近出现两个水的强吸收特征。而当骨胶溶物涂绘上布后,随着膏状颜料中水分的减少,上述两个吸收特征均变弱,个别颜料在1 447 nm处的吸收特征甚至消失。因此,矿物颜料粉末和颜料上布色卡光谱极为接近,可以在后期的唐卡颜料分析中直接利用唐卡矿物颜料粉末光谱进行匹配分析。唐卡红色矿物颜料为朱砂,矿物成分主要为HgS,其光谱在可见光波段先降后升,500 nm附近形成一个较深的吸收特征,且吸收峰较宽（430～530 nm）,红光谱段附近反射率急速升高,近红外波段反射率变化较为平直,在1 940和2 250 nm附近有弱吸收特征。唐卡黄色矿物颜料主要有三种：土黄（雄黄、雌黄）,赭石及金箔,主要成分分别为硫化砷、氧化铁及金。其特征光谱在可见光谱段集中在400～500 nm之间,不同颜料的吸收特征位置和吸收深度均不同。赭石在近红外波段的反射率整体较低,且860 nm附近还出现了吸收特征;而雄黄、雌黄和土黄则在近红外和短波红外谱段表现出反射率较高且波形平直,在1 890和2 230 nm附近有弱吸收特征;金箔在可见光波段的吸收特征窄浅,可作为区分的依据。唐卡蓝色矿物颜料为石青,主要矿物成分为蓝铜矿,其光谱在500～1 000,1 500,2 040,2 285和2 350 nm附近均有较强吸收特征,而在1 885和1 980 nm处有弱吸收特征。唐卡绿色矿物颜料为石绿,主要矿物成分为孔雀石,其光谱在550～1 000 nm有较强的宽吸收特征,在2 270和2 350 nm有明显吸收特征。尽管石青和石绿主要矿物成分皆为碳酸铜,但石绿在900～1 900 nm红-近红外谱段反射率增加较缓,1 500 nm无吸收特征,可以作为区分石青和石绿的依据。唐卡白色矿物颜料为砗磲和白土,主要矿物成分分别为碳酸钙和高岭土。在可见光谱段范围,砗磲在370 nm有弱吸收特征,而白土则在370和730 nm处有两个明显的吸收特征,可作为区分。在短波红外和近红外谱段,白土在1 425,1 930和2 230 nm均具有明显吸收特征,砗磲则在1 930和2 320 nm有明显吸收特征,1 440 nm处吸收特征较弱。且同种矿物颜料粉末,矿物粉末颗粒越大,颜料颜色越深,其光谱特征反射率越低。     

器物理原理演示教学仪器的研制

介绍了一种采用应变片、铂电阻、光敏电阻、磁敏电阻四种电阻型传感器来测量压力、温度、照度、磁场的物理演示教学仪器,可以定量地演示出压力和应变片电阻值、温度和铂电阻阻值、照度和光敏电阻阻值、磁场强度和磁敏电阻阻值之间的关系及给出电阻阻值与对应的物理量之间的曲线。介绍了演示仪器的检测原理和结构,该仪器用于本科生教学,可以使学生将物理学理论与工程实际应用有机结合起来,了解不同属性外场,引起电阻值的变化。拓展出形变、照度、温度、磁场检测等在工程技术中的应用。     

Simulation of stress waves in attenuating drill strings, including piezoelectric sources and sensors

A key element in drill steering and prediction of lithology ahead-of-the-bit is the transmission of while-drilling information from the bottom of the well to the rig operator and the geophysicists. Mud-pulse telemetry, based on pressure pulses along the drilling mud and extensional waves through the drill string, is the most used technique. The last method, properly designed, could transmit data rates up to 100 bits per second, against the 1 or 2 bits per second achieved with pressure pulses. In this work, a time-domain algorithm is developed for the propagation of one-dimensional axial, torsional, and flexural stress waves, including transducer sources and sensors. In addition, the equations include relaxation mechanisms simulating the viscoelastic behavior of the steel, dielectric losses, and any other losses, such as those produced by the presence of the drilling mud, the casing, and the formation. Moreover, the algorithm simulates the passbands and stopbands due to the presence of the coupling joints and pulse distortion and delay due to nonuniform cross-section areas. Acoustic and electric pulses, generated at one location in the string, can be propagated and detected at any other location by piezoelectric and acoustic sensors, such as PCB accelerometers, clamp-on ammeters, force, and strain transducers.     

近红外光谱通用模型在农产品及食品检测中的研究进展

我国人口基数大,农产品和食品的需求量多。农产品和食品的质量与安全与人们的日常生活息息相关。如何实现使用无损、快速、环境友好型、高通量的检测方法对农产品和食品的品质进行检测,是当代社会的发展需要。传统的检测分析方法存在着耗时耗力、检测的样品不能再次出售、产生次品漏检的现象等缺点。近红外光谱分析技术作为一种快速无损的检测手段,逐渐被一些学者以及相关行业人员所重视。然而,近红外光谱分析方法大多数只针对于单一物料建立数学模型。对于数量庞大且种类众多的农产品和食品而言,如不同地域、不同年份、不同温度、不同加工方法、不同成分组成甚至是不同品种,这种相对传统的近红外分析方法无疑会增加建模的工作量。随着计算机技术、光谱仪硬件、化学计量学以及互联网技术的发展,相关学者已经开始着手于近红外光谱通用模型的研究与开发来解决这一问题,即建立一个近红外通用模型,能够对多种物料的同一指标或多种指标进行检测。相比于传统的近红外光谱模型,通用模型具有建模成本低、工作量小等优点,特别使近红外光谱技术在农产品和食品领域中应用以及推广方面具有重要的意义。针对近红外光谱通用模型在农产品和食品检测中的研究进行综述,通过比较传统模型建模方法与通用模型建模方法,分别就建立通用模型过程中样品信息的获取、模型的建立以及样品信息的预测三大建模步骤中使用的方法进行总结,并归纳了近红外光谱通用模型在建模步骤中的要点。近红外光谱通用型模型在农产品以及食品品质检测方面的研究尚处于发展阶段。也提出对于通用模型开发与研究的一些建议,并就近红外光谱通用模型预测方法在检测领域的发展趋势进行了进一步展望。     

风力发电机偏航控制系统设计仿真技术研究

偏航控制系统是风力发电机的核心机构之一,其硬件设计、软件开发与控制算法的合理选用将直接影响偏航控制系统研发成本、周期,以及风力发电机的风能利用率和安全可靠性等。仿真技术作为一种有效的性能预测、试验验证和分析的综合性技术,其具有高效、安全、受环境条件的约束较少等优点,在复杂系统的研发过程中已成为不可或缺的重要手段。因此,偏航控制系统研发前期,为确保系统软硬件开发、控制算法设计的合理性,缩短研发周期、提高研制效率、降低资源成本,将仿真作为主要验证手段。利用Proteus对偏航控制系统的硬件电路和系统软件进行仿真验证,利用MATLAB对控制算法设计的合理性进行考核,为实际偏航控制系统研发提供依据和指导。     

目标跟踪中目标模型更新问题的半监督学习算法研究

本文针对长期稳定的目标跟踪中的目标形变、尺度缩放、旋转等问题, 提出一种步步为营的反馈式学习方法, 该方法通过正、负约束实现对于目标模型和分类器的判别能力和容错能力提高的同时, 使更新带来的误差尽量小, 并证明了其收敛性. 通过实验表明, 对于同一种跟踪算法使用本文提出的目标更新方法进行更新学习的比不更新学习的跟踪效果要稳定得多, 对于目标的尺度变化、形变、旋转、视角变化、模糊等都有较好的适应性, 并通过与现有的较流行的方法进行比较, 本文方法鲁棒性较好, 有很高的研究和应用价值.