Particle methods for multiscale simulation of complex flows

The multi-scale structures of complex flows have been great challenges to both theoretical and engineering researches, and multi-scale modeling is the natural way in response. Particle methods (PMs) are ideal constitutors and powerful probes of multi-scale models, owing to their physical insight and computational simplicity. In this paper, the role of different PMs for multi-scale modeling of complex flows is critically reviewed and possible development of PMs in this background is prospected, with the emphasis on pseudo-particle modeling (PPM). The performances of some different PMs are compared in simulations and new development in the fundamentals and applications of PPM is also reported, demonstrating PPM as a unique PM for multi-scale modeling.     

不同溶剂对七台河煤萃取效果的影响

采用常压沸点下回流萃取方法,选用6种不同的有机溶剂对七台河煤进行了萃取研究,并利用CC／MS联用法对各萃取物的组成进行了分析。但不同的有机溶剂对七台河煤的萃取率相差很大。溶剂的电子受—授能力、溶解度参数、沸点、酸碱性都会对萃取效果产生不同程度的影响。实验证明,在萃取过程中,含有良好的氢键受体、很强的电负性羰基氧的有机溶剂,含n个孤对电子的胺类有机溶剂对七台河煤的萃取率都较大;溶剂萃取过程遵循“相似相溶”规律;呈碱性的有机溶剂对七台河煤的萃取率较大。结果表明,利用有机溶剂萃取法从煤中得到高附加值产品在技术上是可行的。     

Dating recent lake sediments using spheroidal carbonaceous particle （SCP）

Dating lake sediment using sedimentary event is the supplement and calibration to traditional dating by radionuclide such as ^210Pb and ^137Cs. Based on the change of spheroidal carbonaceous particle (SCP) concentration, the age sequence of lake sediments can be deduced. It is one of the dating methods using sedimentary event. SCP is formed from combustion of fossil fuel at high temperature up to1750℃ and at a rate of heating of approaching 104℃/s. It can be dispersed to several hundred kilometers away from its source and deposited with precipitation or dryly deposited, and kept in sediments. Compared with Cs or Pb, there is no evidence for SCP that it decays in lake sediments and is unremovable once stored except by physical disturbance because it is mainly composed of element carbon. Handy method to extract, identify and calculate has been developed. Although fossil fuel has been used early in China, combustion at high temperature started later since emergence of electricity generation. The productivity of SCP is positively related with the generated thermal power, which is reflected as the SCP concentration in lake sediments increases with the increase of generated thermal power. Therefore, reliable sediment markers from the start of the SCP record and the remarkable variation can be used for dating purpose. In China, electricity industry started from the 1950s, and rapid increase of generated power took place since 1978. Based on these time markers, SCP time sequences of lake sediment cores LH and LL-4 from Longgan Lake, the middle reach of the Yangtze River, have been established, which is comparable with the results from ^137Cs and ^210pb, and has eliminated the errors of dating using ^137Cs and ^210pb.     

非线性参数拟合问题的改进阻尼最小二乘方法和共轭梯度方法

本文首先给出了一个改进的阻尼最小二乘方法,并把它用于非线性曲线拟合问题。由于A~TA为对称正定的就有理由一开始把法方程组做乔累斯基分解,并使法方程的条件数有所改善,为此目的而引入了一个非负的阻尼因子。 对于共轭梯度方法,由于非线性函数在极小点附近表现为二次函数的特性,所以在非线性拟合问题中引入了共轭关系P_j~TAP_j=0共轭梯度方法的优点是收敛速度较快,当它用于二次函数极小化问题时总是在有限步内收敛。     

挤压、剪切作用在增稠现象中的作用研究

学者们一般认为浓厚颗粒悬浮液的稠化现象与颗粒间的剪切作用有关,然而有研究者发现,在冲击试验中,颗粒间的挤压作用可能是悬浮液的稠化主要原因。为了区分挤压和剪切行为在浓厚颗粒悬浮液稠化现象中的作用,本文设计了拖球、拖片的对比试验,并且考虑了围压及加载速率的影响。研究结果表明,在稠化过程中,颗粒间的挤压作用占据了主导地位,即颗粒间的挤压作用比剪切作用更容易造成浓厚悬浮液的增稠。围压的存在会影响挤压和剪切的作用力,围压的增大对增稠现象有明显的促进作用;相比于剪切作用,挤压作用对加载速率的促进作用更加敏感。     

甘肃燕儿沟泥石流形成条件及特征参数分析

"5.12"大地震后,甘肃陇南山区地质环境急剧恶化,崩塌、落石、滑坡和地裂缝频发,导致燕儿沟流域内山体更加破碎,为泥石流发育提供了充足的物源条件。在强降雨条件下,极易形成泥石流,严重威胁灾区的生产生活。根据现场调研,对燕儿沟泥石流形成条件进行了分析;并对泥石流的运动学和动力学参数进行了分析计算。结合燕儿沟泥石流形成发育特点,提出了以稳定沟床和排导为主;主、支沟治理相结合;工程治理及生物水保措施相结合的综合治理建议,也为同类工程提供了参考依据。     

NSGA-II与改进粒子群算法相结合的储能优化配置

储能系统的选址定容在接入新能源的电力系统中具有重要意义,合理选择储能容量可以很好地平抑风电接入后的系统有功功率波动,同时减少能源浪费。本文提出了一种混合智能优化算法(Hybrid Intelligent Optimization Algorithm,HIOA),以电压偏差、负荷波动最小及最少储能配置容量为目标进行优化配置。该算法将多目标粒子群算法改进后(IMOPSO)与精英非支配排序算法(NSGA-II)结合,与常规多目标粒子群算法相比,大幅提高了算法的寻优性能,保证了Pareto 解的多样性和分布性。通过非支配排序求解获得Pareto 最优解集,采用逼近理想解排序(technique for order preference by similarity to an ideal solution,TOPSIS)选出储能的最优接入方案。通过IEEE-33 节点测试系统的仿真实验验证了本文所提算法的准确性及有效性,在求解配电网储能选址定容问题中有很好的收敛性和全局寻优能力。     

基于特征层融合和随机投影的行为识别算法

提出一种基于特征层融合和随机投影的行为识别算法;该方法提取视频序列的时空梯度特征和Gabor特征;然后进行特征层融合,得到分类能力更强的特征,有效地表征人体行为;同时,使用随机投影对融合后的特征进行降维;最后,为了解决主题模型参数估计迭代复杂的问题,将贝叶斯参数估计法应用于LDA(latent dirichlet allocation)主题模型中,对视频中的行为进行分类。在公开的KTH和Weizmann数据集上进行了实验,结果表明方法不仅比单一局部时空特征描述符识别性能好,而且在相同实验设置下,也优于其他基本分类器。