Volterra积分方程的蒙特卡罗数值求解方法

本文讨论了第一类、第二类以及具有奇异核的Volterra积分方程的数值解问题.利用重要抽样蒙特卡罗随机模拟方法获得积分方程解的近似计算结果.通过对文献中算例的实现表明文中所提方法扩展了Volterra型积分方程的数值求解方法,     

住户调查中户内样本抽样方法的比较研究

抽样调查是获取社会经济调查数据的主要手段,其抽样设计一般采用分层多阶段不等概的抽样设计。但是,在抽样设计和实际抽样中,人们往往忽视末端样本个体的抽样,本文主要基于中国家庭动态跟踪调查数据对末端样本的概率抽样方法进行比较研究。     

A bootstrapping soft shrinkage approach for variable selection in chemical modeling

In this study, a new variable selection method called bootstrapping soft shrinkage (BOSS) method is developed. It is derived from the idea of weighted bootstrap sampling (WBS) and model population analysis (MPA). The weights of variables are determined based on the absolute values of regression coefficients. WBS is applied according to the weights to generate sub-models and MPA is used to analyze the sub-models to update weights for variables. The optimization procedure follows the rule of soft shrinkage, in which less important variables are not eliminated directly but are assigned smaller weights. The algorithm runs iteratively and terminates until the number of variables reaches one. The optimal variable set with the lowest root mean squared error of cross-validation (RMSECV) is selected. The method was tested on three groups of near infrared (NIR) spectroscopic datasets, i.e. corn datasets, diesel fuels datasets and soy datasets. Three high performing variable selection methods, i.e. Monte Carlo uninformative variable elimination (MCUVE), competitive adaptive reweighted sampling (CARS) and genetic algorithm partial least squares (GA-PLS) are used for comparison. The results show that BOSS is promising with improved prediction performance. The Matlab codes for implementing BOSS are freely available on the website: http://www.mathworks.com/matlabcentral/fileexchange/52770-boss.     

A critical assessment of hidden markov model sub‐optimal sampling strategies applied to the generation of peptide 3D models

Hidden Markov Model derived structural alphabets are a probabilistic framework in which the complete conformational space of a peptidic chain is described in terms of probability distributions that can be sampled to identify conformations of largest probabilities. Here, we assess how three strategies to sample sub‐optimal conformations—Viterbi k‐best, forward backtrack and a taboo sampling approach—can lead to the efficient generation of peptide conformations. We show that the diversity of sampling is essential to compensate biases introduced in the estimates of the probabilities, and we find that only the forward backtrack and a taboo sampling strategies can efficiently generate native or near‐native models. Finally, we also find such approaches are as efficient as former protocols, while being one order of magnitude faster, opening the door to the large scale de novo modeling of peptides and mini‐proteins. © 2016 Wiley Periodicals, Inc.     

无烟煤掺混白酒酒糟制备生物质水煤浆

利用白酒酒糟与贵州无烟煤按不同比例掺混制备生物质水煤浆(BCWS),考查了酒糟掺混量、添加剂种类及含量对水煤浆成浆性能的影响。结果表明,在萘磺酸系添加剂(MF、NNO)和木质素磺酸钙(LS)三种添加剂中,MF的分散效果要优于其他两种,且添加量为0.5%时最佳;表观黏度为1 000 m Pa·s时,无烟煤单独制浆最大成浆浓度为70%,而掺混3%(干基)酒糟的BCWS定黏浓度为65.8%,浆体稳定性在3d以上;BCWS属于宾汉塑性流体,酒糟含有大量的亲水性含氧官能团,以及其管束和脉络结构,会引起BCWS表观黏度的升高,有利于提高BCWS的稳定性。     

Ambient air monitoring with Auto-gas chromatography running in trigger mode

Speciated volatile organic compounds (VOC), either as ozone precursors or air toxics in the air, are commonly monitored by triggered canister method or continuous ozone precursor analyzer (commonly known as Auto-gas chromatography (GC)) method. In the triggered canister method, a canister sample is collected when a total non-methane organic compound (TNMOC) concentration exceeds a pre-determined trigger level. The canister sample is then analyzed in a lab in a later time. In the Auto-GC method, an online GC runs in a “continuous” mode with a sampling and analysis cycle of 1 h. Within the cycle hour, samples are collected only during the first 40 min.A new approach of Auto-GC running in trigger mode is developed in this study. This new approach uses Auto-GC but operates it in a trigger mode similar to the triggered canister sampling method. Compared to the triggered canister sample method, this system provides near real-time speciated VOC data, which are critical for responding to a high VOC concentration episode. Although the canister system generally costs less, its cost advantage may diminish if trigger events are frequent and the monitoring duration is long. Compared to continuous Auto-GC, triggered GC has its niche—it is better for capturing transient plumes with a small footprint. The continuous GC either misses a transient plume if the plume does not arrive at the sampling site during the sampling cycle or flattens the plume concentration peak by dilution with non-plume air sample.Field experience with this system for fenceline VOC monitoring is presented. The sampling and calibration strategy for trigger mode operation is described. The chromatograph retention time drift issues are discussed. The system performance is evaluated, including the method detection limit, precision and accuracy. The trigger mode configuration for VOC fenceline or near source monitoring in this work proved effective for local and transient plume identification.     

悬浮液进样-液体阴极辉光放电原子发射光谱法测定高纯氮化硅粉体中微量杂质元素

针对高纯氮化硅粉体中的9种微量杂质元素(Al、Ca、Co、Fe、K、Mg、Mn、Na、Ni),建立了悬浮液进样-液体阴极辉光放电原子发射光谱定量分析方法.考察了制备稳定悬浮液对样品颗粒度的要求,并通过六通阀将悬浮液引入液体阴极辉光放电原子发射光谱装置检测.本方法采用水溶液标准进行定量分析,无需对悬浮液的pH值进行精确调节,能够保持液体阴极辉光等离子体的稳定性.研究了仪器装置的操作电压、载液流速、光电倍增管积分时间等因素对检出限的影响.优化后得到的最佳实验条件为操作电压1080 V,载液流速1.2 mL/min,光电倍增管积分时间800 ms.利用六通阀进样系统对原有的液体阴极辉光放电原子发射光谱装置进行改进,从而实现悬浮液直接进样检测.用此装置对氮化硅实际样品进行检测,得到各种元素的检出限在0.2~53 mg/kg之间,RSD在1.1%~5.0%之间.通过对氮化硅标准参考物质ERM-ED101进行分析,其测定结果与高温高压消解-电感耦合等离子体发射光谱法一致,并与标准参考值吻合,表明此方法可用于氮化硅粉体的悬浮液直接进样检测,结果准确可靠,灵敏度高,具备应用价值.     

薄膜梯度扩散技术结合相研究进展

薄膜梯度扩散(DGT)技术是一种新型原位被动采样技术,已被广泛应用于水体、土壤、沉积物中目标物的采集与测量。结合相是DGT技术的重要组成部分,决定了与目标物的结合能力、结合速度、结合容量以及目标物的形态选择性等。DGT结合相分为固态结合相和液态结合相。本文重点综述了树脂、氧化物、无机盐、活性炭、改性硅胶、分子印迹、共聚物、复合、液态等结合相在DGT技术中的应用,展望了DGT结合相的发展前景。     

Two effective post-filtering strategies for improving direct sampling methods

In this paper, we introduce two novel strategies to reduce artifacts in the direct sampling type methods (DSM). The newly proposed techniques can essentially reduce the artifacts and provide more accurate and reliable physical profiles of the scatterers compared with the original DSM. The techniques can find wide applications in the inverse scattering problems. Moreover, the novel techniques exhibit several strengths: direct, stable, robust, and ease of implementation.