浅层模型剪枝对准确率及计算量变化探究

剪枝算法能显著降低模型的计算量,是一种在模型计算量和准确率间寻求平衡的方法。现有剪枝算法往往在稀疏性维度进行调整,以不同稀疏度的模型来满足不同计算量和准确率需求的任务。该文从浅层网络剪枝的粒度控制出发,提出了两种新的剪枝策略:减少浅层剪枝与增大浅层剪枝,在稀疏性相近的情况下分别达到进一步提高准确率和降低计算量的作用。通过该文的策略,可以从浅层模型的粒度控制维度去满足不同的任务需求。

星载微波辐射计交叉定标中陆面目标辐射亮温计算模型

从被动微波辐射计在轨辐射定标的角度出发,选取亚马逊雨林作为高温地面目标,利用DMSPF13-SSM/I数据和NCEP数据对研究区19—89GHz辐射特性进行分析。研究发现,利用辐射亮温稳定的热带雨林作为地面目标,对19—89GHz的频段,微波辐射亮温计算精度可以达到1K以内;对于6和10GHz这样的低频通道,撒哈拉沙漠是一个可供选择的地面目标,根据所建立的亮温计算模型,利用数值预报分析场NCEP数据作为输入,其亮温计算精度在2K以内。

基于负载均衡的并行FP-Growth算法

针对FP-Growth算法查找操作时间复杂度较高的问题,提出一种新的算法LBPFP。在PFP算法基础上,将哈希表加入链头表以实现项地址的快速访问,并设计基于前缀长度的计算量模型,优化并行流程,提升算法的执行效率。在webdocs.dat数据库上进行对比实验,结果表明,LBPFP算法比PFP、HPFP、DPFP算法具有更高的频繁项集挖掘效率。

Calculation Model of Equivalent Strength for Induced Crack Based on Double-K Fracture Theory and Its Optimizing Setting in RCC Arch Dam

By means of fracture testing on roller-compacted concrete （RCC） three-point bending beams with two different specimen sizes, the P-CMOD complete curve for RCC was gained. Furthermore, by applying double-K fracture theory, KiniⅠC,KunⅠC, as well as the critical effective crack length and the critical crack tip opening displacement, were evaluated. Based on the double-K fracture parameters above, the calculation model of equivalent strength for induced crack was established, thus the calculation method on its initiation, stable propagation and unstable fracture was ascertained. Moreover, the finite element simulation analysis of stress field in ShaPai arch dam and the on-site observational splaying points of induced crack at different altitudes validated the reliability of the model. Finally, crack inducer′s optimal setting in RCC arch dam was studied. It improves the design level of induced crack in RCC arch dam and satisfies the necessity of engineering practice.

Water Yield and Sediment Yield Simulations for Teba Catchment in Spain Using SWRRB Model: Ⅰ. Model Input and Simulation Experiment

Water yield and sediment yield in the Teba catchment, Spain, weresimulated using SWRRB (Simulator for Water Resources in Rural Basins)model. The model is composed of 198 mathematical equations. About 120items (variables) were input for the simulation, includingmeteorological and climatic factors, hydrologic factors, topographicfactors, parent materials, soils, vegetation, human activities, etc.The simulated results involved surface runoff, subsurface runoff,sediment, peak flow, evapotranspiration, soil water, total biomass,etc. Careful and thorough input data preparation and repeatedsimulation experiments are the key to get the accurate results. Inthis work in the simulation accuracy for annual water yieldprediction reached to 83.68/100.

A New Calculation of the In-Medium Quark Condensate at Finite Density and Temperature

The in-medium quark condensate is studied with an equivalent-mass approach in which one does not need to make assumptions on the derivatives of model parameters with respect to the quark current mass.It is shown that the condensate is generally a decreasing function of both the density and temperature with the decreasing speed depending on the confinement parameter.Specially,at given density,the condensate decreases on increasing temperature.The decreasing speed is comparatively small at lower temperature,and becomes very fast at higher temperature.

Holographic Gas as Dark Energy

We investigate the statistical nature of holographic gas, which may represent the quasi-particle excitations of a strongly correlated gravitational system. We find that the holographic entropy can be obtained by modifying degeneracy. We calculate thermodynamical quantities and investigate stability of the holographic gas. When applying to cosmology, we find that the holographic gas behaves as holographic dark energy, and the parameter c in holographic dark energy can be calculated from our model. Our model of holographic gas generally predicts c 〈 1, implying that the fate of our universe is phantom-like.

Seismic stability analysis of slopes with pre-existing slip surfaces

In analyzing seismic stability of a slope with upper bound limit analysis method, the slip surface is often assumed as a log-spiral or plane slip surface. However, due to the presence of a weak layer and unfavorable geological structural surface or a bedrock interface with overlying soft strata, the preexisting slip surface of the slope may be irregular and composed of a series of planes rather than strictly logspiral or plane shape. A computational model is developed for analyzing the seismic stability of slopes with pre-existing slip surfaces. This model is based on the upper bound limit analysis method and can consider the effect of anchor bolts. The soil or rock is deemed to follow the Mohr-Coulomb yield criterion. The slope is divided into multiple block elements along the slip surface. According to the displacement compatibility and the associated flow rule, a kinematic velocity field of the slope can be obtained computationally. The proposed model allows not only calculation of the rate of external work owing to the combined effect of self-weight and seismic loading, but also that of the energy dissipation rate caused by the slip surface, interfaces of block elements and anchorage effect of the anchors. Considering a direct relationship between the rate of external work and the energy dissipation rate, the expressions of yield acceleration and permanent displacement of anchored slopes can be derived. Finally, the validity of this proposed model is illustrated by analysis on three typical slopes. The results showed that the proposed model is more easily formulated and does not need to solve complex equations or time consuming iterations compared with previous methods based on the conditions of force equilibrium.

Functional Integrals and Quantum Fluctuations on Two-Dimensional Noncommutative Space-Time

The generalized Thirring model with impurity coupling is defined on two-dimensional noncommutativespace-time,a modified propagator and free energy are derived by means of functional integrals method.Moreover,quantum fluctuations and excitation energies are calculated on two-dimensional black hole and soliton background.

Fingerprint singular points extraction based on orientation tensor field and Laurent series

Singular point(SP)extraction is a key component in automatic fingerprint identification system(AFIS).A new method was proposed for fingerprint singular points extraction,based on orientation tensor field and Laurent series.First,fingerprint orientation flow field was obtained,using the gradient of fingerprint image.With these gradients,fingerprint orientation tensor field was calculated.Then,candidate SPs were detected by the cross-correlation energy in multi-scale Gaussian space.The energy was calculated between fingerprint orientation tensor field and Laurent polynomial model.As a global descriptor,the Laurent polynomial coefficients were allowed for rotational invariance.Furthermore,a support vector machine(SVM)classifier was trained to remove spurious SPs,using cross-correlation coefficient as a feature vector.Finally,experiments were performed on Singular Point Detection Competition 2010(SPD2010)database.Compared to the winner algorithm of SPD2010 which has best accuracy of 31.90%,the accuracy of proposed algorithm is 45.34%.The results show that the proposed method outperforms the state-of-the-art detection algorithms by large margin,and the detection is invariant to rotational transformations.

A Scheme for Simulation of Quantum Gates by Abelian Anyons

Anyons can be used to realize quantum computation, because they are two-level systems in two dimensions. In this paper, we propose a scheme to simulate single-qubit gates and CNOT gate using Abelian anyons in the Kitaev model. Two pairs of anyons （six spins） are used to realize single-qubit gates, while ten spins are needed for the CNOT gate. Based on these quantum gates, we show how to realize the Grover algorithm in a two-qubit system.

Simulation of Hail and Soil Type Effects on Crop Yield Losses in Kansas,USA

Computer simulation was used for predictive analysis of the effects of weather and soil type on crop yield in the U.S. crop insurance program. The Environmental Policy Integrated Climate （EPIC） model was modified to include hail weather events, which completed the modifications necessary to simulate the four most frequent causes of crop yield loss （hail, excessive wet, excessive cold, and excessive dry） associated with soil type in Kansas, USA. At the region level, per hectare yields were simulated for corn, wheat, soybean, and sorghum. We concluded that it was possible to predict crop yields through computer simulation with greater than 93% accuracy. The hail damage model test indicated EPIC could predict hail-soil-induced yield losses reasonably well （R^2 〉 0.6）. The investigation of soil type influence on dryland sorghum and wheat production indicated that Wymore silty clay loam soil and Kenorna silt loam produced the highest sorghum yields statistically; Kuma silt loam, Roxbury silt loam, Crete silty clay loam, and Woodson silt soils produced the second highest sorghum yields statistically; and Richfiled silt loam, Wells loam, and Canadian sandy loam produced the lowest sorghum yields. By contrast, wheat production showed less sensitivity to soil type variation. The less sensitive response of wheat yields to the soil type could be largely due to the unconsidered small-scale variability of soil features.

Dynamic Integrity Measurement Model Based on vTPM

With the development of cloud computing, virtualization technology has been widely used in our life. Meanwhile, it became one of the key targets for some attackers. The integrity measurement in virtual machine has become an urgent problem. Some of the existing virtualization platform integrity measurement mechanism introduces the trusted computing technology, according to a trusted chain that the Trusted Platform Module(TPM) established for trusted root to measure the integrity of process in static. But this single chain static measurement cannot ensure the dynamic credible in platform running. To solve the problem that the virtual trusted platform can not guarantee the dynamic credibility, this paper put forward Dynamic Integrity Measurement Model(DIMM) based on virtual Trusted Platform Module(v TPM) which had been implemented with typical virtual machine monitor Xen as an example. DIMM combined with virtual machine introspection and event capture technology to ensure the security of the entire user domain. Based on the framework, this paper put forward Self-modify dynamic measurement strategy which can effectively reduce the measurement frequency and improve the measurement performance. Finally, it is proved that the validity and feasibility of the proposed model with comparison experiments.

Visible calculation of mining index based on stope 3D surveying and block modeling

Aimed at the CMS laser scanning theory and characteristic, a combined actual situation of stope N4-5 of Fankou Lead-Zinc Mine and complementary monitoring of the stope were carried out by carefully choos- ing two measuring points. The cavity 3D visible model was created by large-scale mining industry soft- ware Surpac after changing the measured data. The stope mine design model, bottom structural model and backfill model of the south and north sides of the stope N4-5 were established according to the stope design data. On this basis, the stope block model was established, and then block attribute was estimated. The amount the ore remains, mullock, backfill and total mined ore were calculated through the solid model restrains. Finally, the stope mining dilution rate and loss rate reached 8.2% and 1.47%, respectively. The practice indicates that the mining index visible calculation method based on cavity 3D monitoring and stope block modeling can make up the deficiency of adopting the solid model to directly carry out the Boolean operation, The stope mining indexes obtained by this method are accurate and reliable, and can be used to guide the actual nroduction management and estimate the mining aualitv.

π-Conjugated Unit-Dependent Optical Properties of Linear Conjugated Oligomers

The optical properties of three linear conjugated oligomers （F-P, F-P-F, and P-F-P-F-P）, where phenothiazine （P） and fluorene （F） groups arrange alternately, are investigated. With the enhancement of the π-conjugated system, their absorption and emission bands both gradually red shift, and their two-photon properties are also improved. Meanwhile, their fluorescence dynamic traces are analyzed with continuous rate distribution model, exhibiting that their decay rates gradually accelerate and the rate distribution width become narrower. The quantum chemical calculation offers their molecular structures and transition mechanism, showing that the enhancement of π-conjugated system should be responsible for tile improvement of two-photon properties.

Route choice models based on altering OD

The condition and physical sense of actual dynamic user optimum are explained by analyzing a simple road network route choice. To match the practical application requirements, assignment network and simulation network are classified account for varying flowing loading. Instantaneous dynamic user optimum model should be applied to the former and actual dynamic user optimum model the latter respectively. The two model’s feasibility is studied as well. Considering the application in ATMS, the model is mainly used to analyze the altering OD problem. Moreover, it adds the method of route adapting into the object function selection to appraise elastic trip strategy and set up real means of route inducement.

Study on CO_2 Emission Reduction from Chinese Coal-Fired Power Plants Between 1993 and 2010

Based on the Chinese thermal coal and power generation data,such as ultimate analysis,proximate analysis,low heat value(LHV)on as received basis,power generation volume,thermal coal consumption volume and net coal consumption rate,several mathematical models for calculating CO 2 reduction by Chinese coal-fired power plants are established.Calculations of the CO 2 emission factor(CEF),the CO 2 emission volume and reduction volume are made according to these models.The calculation results reveal that between 1993 and 2010,the CO 2 emission volume reached 31.069 Gt,reduced by 0.439 Gt,averaging 28.83 Mt each year.

Computation model of proppant embedment depth based on dimensional analysis

Hydraulic fracturing is designed to form a high-conductivity fracture. The proppant will embed into the formation rock, especially in soft rock, owing to the interaction between proppant and fracture surface after fracture closure. Proppant embedment would reduce the fracture width and then lower the fracture conductivity. According to dimensional analysis, the rock is assumed to be an elastic material. Using the theory of elasticity to describe the stage of elastic deformation and analysis of the corresponding simplified embedding process, the study establish the static computation model of the two-dimensional infinite half plane and three-dimensional infinite half space model of the proppant embedment. According to laboratory results, the calculation model was modified, got an effective correction factor and analyzed the causes of errors, then discussed the factors which have impact on proppant embedment. The result calculated by the model in this paper can be reference of prop- pant optimization in on-site fracturing for a certainty degree.

Modeling of Earth Magnetic Field Disturbances Induced by Ferromagnetic Objects

This paper presents the methodology and results of modelling of Earth magnetic field disturbances induced by ferromagnetic objects. The modeling was carried out using the finite elements calculations, and the results were compared with the measurements. The special test stand, consisting of magnetovision scanner and Helmholtz coils, has been used. The measurement system is able to measure the distribution of planar magnetic induction vectors, and to differentiate the sources of measured disturbances.

Laminar Heat Transfer Modeling and Investigation of Thermal Dispersion in a Flat Plate Heat Exchanger： Estimating Heat Transfer Coefficient, Heat Flux and Nusselt Number

PHEs （plate heat exchangers） are among the most common thermal equipments in diverse industries particularly in oil and gas companies. This wide usage is obviously due to significant benefits of these heat exchangers over other types. In this article, a behavioral analysis of heat transfer in fiat plates of these heat exchangers in laminar flow situation through CFD （computational fluid dynamics） simulation using FLUENT 6.3.26 software is done. The study reveals results graphically based on fluid＇s behavior in co-current and counter current flows and discusses thermal indexes consisting of heat transfer coefficient, Nusslet and total heat flux in both conditions. Eventually, a comparison via the graphical results is presented between the two types of flow directions.