Journal of University of Chinese Academy of Sciences

Some methods of parameter estimation for stochastic differential equations

CAI Xinrui, WANG Lijin

2017, 34 (5): 529-537. DOI: 10.7523/j.issn.2095-6134.2017.05.001

We propose three methods of parameter estimation based on discrete observation data for stochastic differential equations (SDEs). The first method is designed for linear stochastic differential equations (SDEs). For these equations we deduce distribution of certain operation of the exact solution and assume that the relevant operation of the observed data obey this distribution, from which we estimate the unknown parameters in the drift and diffusion coefficients. In the second method, we suppose that certain operation of the observation data and that of the numerical solution arising from the Euler-Maruyama scheme for the SDEs of Itô sense obey the same distribution, from which the unknown parameters can be estimated. We use the third method for SDEs of Stratonovich sense. For these equations we derive the distribution of relevant operation of the numerical solution produced by the midpoint scheme and let the same operation of the data obey this distribution to get estimation of the unknown parameters. Numerical results show validity of the proposed methods, and illustrate that the estimation error produced by the Euler-Maruyama scheme is about of order O(h^0.5) while that by the midpoint scheme is about of order O(h), with h being the time step size of the numerical methods. Furthermore, the numerical results show that our methods are more accurate than the existing EM-MLE estimator.

Herz space with variable exponent on spaces of homogeneous type

WANG Hongbin

2017, 34 (5): 538-542. DOI: 10.7523/j.issn.2095-6134.2017.05.002

Abstract ( 280 ) PDF (KB) ( 0 )

In this work, a certain Herz space with variable exponent on spaces of homogeneous type is defined, and the block decomposition for this space is established. Using this decomposition, some boundedness for a class of sublinear operators on Herz space with variable exponent on spaces of homogeneous type is obtained.

Photoelectric ptychographical imaging based on pre-known amplitude information constraint

WANG Zhibo, ZHANG Jun, LI Tuo, SHI Yishi

2017, 34 (5): 543-550. DOI: 10.7523/j.issn.2095-6134.2017.05.003

Abstract ( 227 ) PDF (KB) ( 0 )

A visible light ptychography imaging method based on pre-known surrounding amplitude information is proposed,in order to effectively improve accuracy and robustness of the reconstruction.The method employs a known standard object as priori information which surrounds the sample.Ptychography scans the sample and the pre-known surrounding information,and then introduces them into the iterative algorithm.Finally a high quality reconstruction result is achieved after several iterations.With the simulations and by setting up the optical system,the experiments on different samples are carried out to further study its validity and versatility.In addition,the method shows a good anti-noise performance and anti-probe-offset capability.Due to the pre-known information constraint,a high quality reconstruction is completed without additional cost.The method also maintains the superiority of the traditional ptychography.This method is analyzed and proved to be simple-to-operate,steady,and versatile.The ptychography based on the pre-known information constraint has a quite promising practical value.

Plant species diversity in bauxite residue disposal areas

HAN Fusong, WANG Cheng, XUE Shengguo, HUANG Nan, YE Yuzhen

2017, 34 (5): 551-557. DOI: 10.7523/j.issn.2095-6134.2017.05.004

Abstract ( 331 ) PDF (KB) ( 0 )

Bauxite residue is an alkaline waste by-product generated in the extraction of alumina from bauxite.Due to large volume and inherent environmental risk,the disposal of bauxite residue has become a major concern for sustainable development of the aluminum industry.Bauxite residue poses numerous physical and chemical limitations to plant establishment in disposal areas.Based on the on-the-spot investigation in bauxite residue disposal areas (BRDAs),the important values and indices of plant species evenness and diversity were calculated.A total of 25 species were found naturally colonized in BRDAs.These species belong to 10 families and 22 genera,among which 14 species were annual.Gramineae and Compositae accounted for 52%of the total species.The plant community in BRDAs showed low species diversity and simple structures.The dominant types of plant community were Chloris virgata Sw.-Cynodon dactylon (L.) Pers community and Corispermum declinatum Steph ex Stev-Setaria viridis (L.)Beauv community.C.dactylon is a dominant species in some BRDAs which grows rapidly and has saline-alkaline tolerance,wide distribution,and a broad ecological amplitude.This species has potential in the revegetation of BRDAs.

Mutagenesis, screening, and medium optimization of high[(4'-OH)MeLeu]⁴-CsA producing strain-Nonomuraea dietziae

JIN Lina, WEN Jianping

2017, 34 (5): 558-566. DOI: 10.7523/j.issn.2095-6134.2017.05.005

Abstract ( 232 ) PDF (KB) ( 0 )

An ultraviolet-lithium chloride mutagenesis strategy was carried out on the[(4'-OH)MeLeu]⁴-CsA producing strain, Nonomuraea dietziae.The production of mutant strain reached 208mg/L and increased by 26.8%. Then the fermentation medium was optimized by single-factor and response surface methods. The optimal medium was determined as follows (materials/concentrations in g/L):glycerol/26.5, dextrin/11.6, peptone/9.7, yeast extract powder/3.5, soybean meal/15.0, soybean oil/2.0, MgSO₄·7H₂O/0.40, MnSO₄·H₂O/0.010, and (NH₄)₂MoO₄/0.010. The final production reached 324mg/L and increased by 97.6% compared with the initial value.

Characterization of typical biomass burning tracers among atmospheric particles in urban Guangzhou

WANG Lijuan, WU Dalei, ZHANG Zhisheng

2017, 34 (5): 567-572. DOI: 10.7523/j.issn.2095-6134.2017.05.006

Abstract ( 389 ) PDF (KB) ( 0 )

Size-resolved particle samples were collected at an urban site of Guangzhou during the summer and autumn.Samples were subjected to chemical analysis for obtaining various typical biomass burning tracers (levoglucosan (LG),mannosan (MN),and water-soluble potassium ion (K⁺)) and carbonaceous species including organic carbon (OC) in order to identify the major biomass burning particle type and quantify the contribution of biomass burning to OC abundance.Results showed that biomass burning tracers had dominant particle sizes of 0.44~1.00μm.High LG/MN and K⁺/LG ratios in the fine particle mode were found,which is consistent with the particle source profile emitted from agricultural residue burning.The LG/OC level during the autumn in urban Guangzhou was comparable to those in other Chinese cities during the biomass burning seasons.Based on a receptor model,the average contributions of biomass burning to OC during the summer and autumn were estimated to be 10.8% and 28.1h%,respectively,which implies that biomass burning sources,especially agricultural residue burning,cannot be ignored in the urban atmosphere in Guangzhou.Furthermore,relevant prevention suggestions were provided.

Changes and coupling characteristics of industrial production space and industrial structure during the economic transformation period in Pearl River Delta

DING Jun, WANG Kaiyong

2017, 34 (5): 573-581. DOI: 10.7523/j.issn.2095-6134.2017.05.007

Abstract ( 212 ) PDF (KB) ( 0 )

By using the integrated variation coefficients of industrial production space and the industrial structure adjustment index,we analyzed the changes in industrial production space and industrial structure during the transformation period in Pearl River Delta (PRD) and evaluated their coupling coordination degree by using the coupling model.The main results obtained in this study are given as follows.1) The industrial production space has an expanding trend from core area to periphery in PRD,and the change degree in periphery is larger than in core area.2) The industrial structure of PRD has been constantly optimized during the transformation period,but cities which are in different stages of industrialization development have different adjustment directions.The cities at higher industrialization level tend to evolve toward higher industrial structures and pay much attention to the development quality and environmental protection.However,the cities at lower industrialization level mainly develop the resource-based industry and high pollution industry.3) The coupling coordination degree between industrial production space change and industrial structure adjustment in periphery of PRD is higher than in core area,but the coupling coordination level and development quality in the periphery of PRD are very low.During the economic transformation period,we should not only show concern for the increase of the "quantity",but also pay more attention to the promotion of the "quality" of the industrial production space and industrial structure.

Quantitative calculations of virtual land in Gansu Province

LUO Kaisheng, TAO Fulu

2017, 34 (5): 582-590. DOI: 10.7523/j.issn.2095-6134.2017.05.008

Abstract ( 226 ) PDF (KB) ( 3 )

The virtual land calculation has not been paid enough attention.In this work,the concept of virtual land and its calculation methods were proposed and applied in Gansu Province based on the statistical analysis method and the economic and social statistical data.Results are given as follows.1) Broomcorn millet,linseed,bean,soybeans,millet,cotton,rape,hemp,tobacco,angelica,and winter wheat belong to the land intensive crops;the virtual lands of yam,spring wheat,sorghum,corn,grapes,apples,and rice are moderate;and vegetable,sugar beet,and honey melon belong to the land sparse crops.2) Virtual lands of cotton,oil crop,rice,sugar beet,grapes,and corn in Gansu Province are 5.91,5.67,1.45,0.23,1.77,and 2.10m²/kg,respectively;virtual lands of live cattle,horses,and sheep in Gansu Province are 69.28,46.68,and 48.56m²/kg,respectively;and virtual lands of beef,mutton,wool,wines,and cashmere are 197.91,115.61,693.66,76.84,and 5.39m²/kg,respectively.Virtual land has an accumulation amplification effect and increases with the product line.The order of virtual lands is:industrial products > animal products > crop products > crops.3) It is also found that virtual land of the broad-leaved forest is much larger than that of the coniferous and mixed forest and that virtual land of the natural forest is larger than that of the artificial forest.

Spatial structure of urban network in China based on the headquarters and branches of express enterprises

SUN Yang, ZHANG Luocheng, YAO Shimou

2017, 34 (5): 591-597. DOI: 10.7523/j.issn.2095-6134.2017.05.009

Abstract ( 264 ) PDF (KB) ( 3 )

In this work we start from the perspective of the space of flows by collecting inter-provincial headquarters and branches of express enterprises and structure 300×300 matrix architecture.The spatial structure of the urban network is analyzed,and different characters as the effects of the space of the flow among cities at different levels are revealed.The results are showed as follows.1) Nationwide Beijing-Guangzhou-Shanghai city flow presents a triangular structure.2) The linkages between Shanghai-Changchun and Shanghai-Harbin are obvious.The connections between Hohhot-Beijing and Baotou-Beijing should be improved.3) The central cities of Zhengzhou,Wuhan,and Changsha are obvious compared with other cities,but in the contact area of the cities the network is not prominent and the degree of concentration is low.4) The urban linkages between Chengdu and Beijing-Tianjin and between Chengdu and the Yangtze River Delta city group core are close.

Relationship between energy consumption and economic growth based on grey relational model:a case study of Xinjiang

WANG Xingmin, WANG Qiang, DONG Jiefang

2017, 34 (5): 598-609. DOI: 10.7523/j.issn.2095-6134.2017.05.010

Abstract ( 265 ) PDF (KB) ( 0 )

Deep understanding of the relationship between energy consumption and economic growth is of vital importance for exploring the energy saving space in Xinjiang. Based on the data related to the energy consumption and economic growth in Xinjiang from 2005 to 2014 and by using the grey relational analysis method, the correlation between energy consumption in various industries with economic growth and the correlation of energy consumption types with economic growth in Xinjiang are analyzed. As indicated by the results of the analysis, firstly, the relational degrees between energy consumption and economic growth differ significantly among industries. Specifically, the industrial sector and construction industry present the same highest relational degree. Secondly, among the industrial sectors, the manufacturing sector energy consumption has the greatest relevance with economic growth and accounts for the largest proportion. Lastly, the relational degrees with economic growth also differ significantly among various energy consumption types. The coal energy consumption has the highest relational degree with economic growth, while petroleum energy consumption has the lowest relational degree. These illustrate two points. On one hand, the economic growth in Xinjiang has a factor input economic growth mode which strongly depends on energy. On the other hand, the structure of energy consumption in Xinjiang is not reasonable to some degree under the consideration of requirements of the national strategies. Based on this study, some policy implications regarding the industrial energy consumption, energy consumption types, and economic growth in Xinjiang are proposed.

Characterization of water imbibition in sandstones studied using nuclear magnetic resonance

ZHANG Qian, DONG Yanhui, TONG Shaoqing

2017, 34 (5): 610-617. DOI: 10.7523/j.issn.2095-6134.2017.05.011

Abstract ( 318 ) PDF (KB) ( 0 )

Characterization of water imbibition in porous media is important in a variety of engineering and geological contexts.This work is aimed at investigating sorption characteristics of sandstones and how they are related to the pore structure of samples and boundary conditions.Over the past decades,NMR techniques have been proven to be reliable for visualizing the water distribution and quantifying the dynamic transport process of water in porous media.In this study,Carr-Purcell-Meiboom-Gill (CPMG) measurement was used to obtain the T₂ distribution which showed the change in water content in pores of different sizes with time.Moreover,one-dimensional magnetic resonance imaging (MRI) profile measurement was used to obtain the water content profiles during water imbibition into artificial sandstone samples in a quantitative and non-destructive manner.In order to understand the characteristics of water absorption and factors influencing the process,results of imbibition in samples of different permeability were compared by performing one-dimensional and three-dimensional NMR experiments.Sorptivity based on Fickian diffusion was adopted to make a quantitative analysis of sandstone samples. The results showed that sample of high permeability had a greater amount of water absorption with less time.In addition,the cumulative water absorption varied linearly with t^1/2 as expected.Therefore the sorptivity was derived directly to be 0.036 8 cm/min^1/2 with the correlation coefficient r=0.996 7.However,the increase of water content slowed down after the wetting front reached the top of the sample,which could be explained by the retarding effect of evaporation on the sample surface.

Field observation of the evolution of spatial distribution of seawater-freshwater interface

ZHOU Xiaoni, SHI Jiansheng, MA Rong, MIAO Qingzhuang, LIU Shaoyu, LIU Pengfei, ZHANG Wei, WANG Zhe

2017, 34 (5): 618-624. DOI: 10.7523/j.issn.2095-6134.2017.05.012

Abstract ( 287 ) PDF (KB) ( 0 )

In traditional research people mainly employed the laboratory simulation experiments to characterize the spatial distribution of interactive interface between salt and freshwater.However,these approaches did not take into account the scale effects and heterogeneity of stratum structure.To address these problems,a typical profile was selected from the piedmont plain to the alluvial coastal plain in the south part of Shandong Province.The water table and samples were collected from this profile within a cycle period of flood-ebb fluctuations,and the multivariate statistical analysis method was used to describe the spatial distribution of this interactive interface.The results indicated that the interactive interface distributed as an inverse U-shape and the burial depth was between 0.3 to 1.5 m vertically with the central point at the depth of 0.9 m at X5 point,and its intrusion intensity variance with depths was 0.9 m > 0.6 m > 1.2 m > 1.5 m.Under the condition of gravity activity the interactive interface migrated deep,showing a tendency of quicker saline-water intrusion and slower desalination.The groundwater type changed as an iterative process from bicarbonate type to chloride type in this interactive interface.

Application of two-phase smoothed particle hydrodynamics method to dam break over movable bed

QIAO Cheng, PAN Huali, OU Guoqiang

2017, 34 (5): 625-632. DOI: 10.7523/j.issn.2095-6134.2017.05.013

Abstract ( 269 ) PDF (KB) ( 0 )

Smoothed particle hydrodynamics (SPH) method adopts discrete particles for discretion of computational domain,and field variables and their spatial derivatives are evaluated by summation interpolation based on particle and kernel functions,which makes this method more suitable for processing the large displacement,movable boundary,and free surface problems.In order to analyze the dam break problem over movable bed,an erosion condition based on Drunker-Prager is used.The particles satisfying this condition will move with the non-Newtonian characteristic,i.e.,Herschel-Bulkley-Papanastasiou rheology,while the particles which do not satisfy the condition will remain still.In order to improve the results of numerical modelling further,a two-phase SPH method which is based on the ratio of volume is applied when modelling the dynamics of mixture,and the calculation of viscosity of mixture is improved based on the dynamic characters of solid-fluid mixture.A dam-break problem over movable bed is analyzed based on the model mentioned above.Numerical results are compared with the experimental data,and the comparison verifies that the proposed formula improves the results of numerical modelling.

Investigation of normalization methods in speaker adaptation of deep neural network using i-vector

YANG Jianbin, ZHANG Weiqiang, LIU Jia

2017, 34 (5): 633-639. DOI: 10.7523/j.issn.2095-6134.2017.05.014

Abstract ( 406 ) PDF (KB) ( 0 )

The deep neural network (DNN) was a remarkable modeling technology for speech recognition in recent years and its performance was significantly better than that of the Gaussian mixture model,which was the mainstream modeling technology in speech recognition before.However,commendable adaptation of DNN has not been solved yet.In this work,we use the identity vector (i-vector) to adapt a deep neural network by putting i-vector and the regular speech features together as the input of DNN for both training and testing.Then we focus on the normalization method of i-vector using a new max-min linear normalization method.We get a 5.10%relative decrease in word error rate over the traditional length normalization method.

Analysis of phase precision for non-linear ECS auto-registration imaging algorithm

CHEN Lifu, WANG Siyu, YUAN Zhihui, WANG Jing, XING Xuemin, WU Hong

2017, 34 (5): 640-646. DOI: 10.7523/j.issn.2095-6134.2017.05.015

Abstract ( 362 ) PDF (KB) ( 0 )

To further improve DEM precision for InSAR system,4 factors affecting phase precision are analyzed for the non-linear ECS auto-registration imaging algorithm.The results are given as follows.The equivalent error of relevant velocity between radar and scatter is very small and can be ignored.When the system resolution is better than 0.1 m,quadratic approximation error in the non-linear ECS auto-registration generates 0.1 pixel registration error,and it cannot be ignored.The second-order Taylor approximation error in the non-linear scaling is very small and can be neglected.During interferometric motion compensation,the flat assumption error in the imaging causes locus error and phase error,which cause some impact and should be considered for high-precision InSAR system (DEM precision is higher than 0.5 m).The above error analyes will provide some guidance for further improving phase maintenance precision and DEM precision.

A software behavior model based on dynamic taint analysis

YIN Zhiyi, SHEN Jiahui, GUO Xiaobo, ZHA Daren

2017, 34 (5): 647-656. DOI: 10.7523/j.issn.2095-6134.2017.05.016

Abstract ( 252 ) PDF (KB) ( 0 )

Based on the fine-grained binary dynamic analysis platform,we propose a taint analysis method to construct the software behavior model using the system call arguments.Firstly,the method obtains the associations between the arguments,between an argument and a local variable,and between an argument and a foreign data through monitoring the applications running and tracking the taint propagation of system call arguments at the instruction level,and then the taint propagation chains between arguments are generated.Secondly,a software behavior model,which covers control-flow and data-flow,is built according to these chains and system call sequence.Finally,the experimental and analytical results demonstrate that this model can be used to detect stealthy non-control attacks.

Current Issue