In this work the stochastic approximation problem of perturbed linear systems was examined. Inspired by the averaging-based accelerated algorithm of Polyak and Juditsky(SIAM J. Control & Optimization,1992,30:838-855), we propose an averaging-based truncated algorithm. The almost sure convergence and asymptotic normality of the sequence defined by this algorithm are proved.

The characteristics of heat transfer and pressure drop of the steam condensation in plate evaporator were analyzed based on experiment platform of the steam condensation test. The properties of heat transfer of steam condensation in herringbone plate evaporator were predicted by using the two-phase flow condensation model. A correlation of steam condensation heat transfer coefficient in herringbone plate evaporator was fitted by the multiple linear regression analysis method. Pressure drop per unit length along herringbone plate increased with the velocity of steam. Several pressure drop models of steam condensation in herringbone plate evaporator were compared. The experimental correlation between steam condensation friction factor and the equivalent Reynolds number was derived.

It is well known that hyperglycemia results in oxidative stress in diabetes mellitus, which causes pancreatic β cell dysfunction. However, this cytoprotective effect of Nrf2 on hyperglycemia-induced oxidative stress in pancreatic β cell is not fully understood. In the study, reactive oxygen species (ROS) and insulin synthesis and secretion as well as nucleus NF-E2-related factor 2 (Nrf2) expression in NIT-1β cells were determined after treatment of 48 h with low glucose (LG, 5.5 mmol/L), high glucose (HG, 27.6 mmol/L), or LG plus H₂O₂, respectively. We found that HG significantly induced ROS generation and decreased insulin synthesis and secretion in NIT-1β cells. However, HG-induced oxidative damage was improved by nucleus NF-E2-related factor 2 (Nrf2) expression. The results suggested that activation of Nrf2 through promotion of nuclear accumulation plays an important role in protection against hyperglycemia-induced oxidative injury.

The traditional approaches were used to assess the impact of underground coal gasification (UCG) on groundwater quality. The results show that all of those methods cannot reflect the groundwater pollution from UCG effectively, while the site influence index and single pollution index approaches work well. Sulfate, chlorine, dissolved solids, and volatile phenol were main pollution components. In addition, we study the long-term effects of sulfate in UCG of 3 years on downgradient aquifer by using numerical simulation, and the results predict that the effect time would not be less than ten years.

As the forefront of China's reform and opening up, the Pan-Yangtze River Delta has become one of the most developed areas in China. Recently, it experienced manufacturing relocation due to severe shortage of resources as well as other relevant factors. Hence, issues of manufacturing industry shift in the Pan-Yangtze River Delta have increasingly become academic focus. Based on the perspective of "core-periphery" theory, the Pan-Yangtze River Delta is divided into three regions, "core-marginal-peripheral". Based on the output data of 28 industry departments in the Pan-Yangtze River from 2000 to 2012, we systematically show and decompose the increase of industry at city level in the regions and explain the increase by the regional effect, the structural effect, and the competitive effect. Then the process of industry shift between different regions is studied based on the shift-share analysis diagrams. Finally we explore the mechanism of industry shift using the "push-pull" model.

County industry development is a key issue to spatial optimization and "revitalization of the old industrial base in Northeast China". By constructing an industry development index system, we use the entropy method to calculate the comprehensive level scores for every county industry. Then by taking this achievement as a spatial analytic basic variable we use Markov chains to analyze the evolution of spatial pattern and hot spots in Jilin county industry. Finally we analyze the factors which lead to this evolution. The results show that the level of industrial development in the central cities was significantly higher than in the surrounding counties in Jilin. The gap of the industrial development level in Jilin province was expanded from 2000 to 2004. The industrial development level has increased and balanced from 2004 to 2012. The hot spot was focused on Changchun-Jilin area where there is a better industrial infrastructure, and the hot spot was stable over the years. Finally, we propound the path for raising the level of industrial development.

With the data of CO₂ emissions and population of Xinjiang from 1958 to 2011, we make empirical analysis on the causal relationship between the change in population factors and the CO₂ emissions and on the acting force by using Granger causality test and the ridge regression model. The results are shown in the following. 1)Before the reform and opening up, the bidirectional causal relationship presents between all the relative indexes of population and CO₂ emissions. However since the reform and opening up, the causal relationship between the countryside inhabitant cousumption and CO₂ emissions disappears. 2)The promoting function of population factors on CO₂ emissions is rather different in different stages of development. With the implementation of the reform and opening-up policy, the households and consumption become the main driving factos in place of employment structure and population gross.

The index for orchard in plain terrain (IOPT), which is fit for the information extraction of orchard in plain terrain with medium-resolution remote sensing images, was proposed. With the full use of the multi-temporal features of images, an extraction mode based on object-based information analysis method was established. We took Dangshan county as the study area. The experimental results show that the extraction accuracy of orchard information reached up to 89.5%. Using this method one can avoid "salt and pepper phenomenon" and monitor the distribution of orchard correctly and efficiently.

An assessment index system is built and it is composed of the hazards of risk source and the sensitivity and resiliency of risk receptors. The dynamic assessment of environmental risk in the Yangtze River basin from 2000 to 2011 was evaluated by using the "vertical-and-horizontal scatter degree" method and the "time ordered weighted averaging operator" method. The results are given in the following. 1)Temporally, the environmental risk of the study area decreased slightly in the initial stage of the study period, followed by an obvious increase consequently. 2)Spatially, great variations in risk value indicate that the Yangtze River basin can be divided into the high, relatively high, medium, and low risk regions with assistance of cluster analysis. The high risk zone includes Shanghai, Jiangsu, and Henan; the relatively high risk zone includes Anhui, Hunan, Hubei, and Zhejiang; the provinces in the central area of the Yangtze River basin, such as Sichuan, Guangxi, Chongqing, Yunnan, and so on, are in the medium risk zone; and the Tibet and Qinghai occupy the low risk zone. The temporal and spatial patterns of environmental risk for the Yangtze River basin provides important scientific basis for the environmental risk management.

Carbonate melts play important roles in the carbon cycle process in the earth interiors. However, the previous studies were limited to the low temperature and low pressure conditions, and mainly focused on K⁺ and/or Na⁺ bearing carbonate melts. In this study we used extensive first-principle molecular dynamics simulation to obtain the equation of state of CaCO₃ melts under the mantle condition. The results are listed in the following. CaCO₃ melts have significantly greater compressibility than aragonite or typical silicate melts and hence the densities rapidly increase with pressure. When the pressure exceeds 10 GPa, densities of CaCO₃ melts are higher than that of albite melt and the densities are even higher than that of diamond (at the pressure above 37 GPa). Densities of CaCO₃ melts are lower than that of crystalline phase throughout mantle condition, but the density difference between them significantly decreases when the pressare increases. The high compressibility of CaCO₃ melts may have important implications for uncovering distributions of carbonate melts in the mantle and for the formation of ultra-deep diamonds.

Carbon and oxygen isotopic determination of carbonate samples has been carried out in this work, and the samples were collected from the Upper Permian-Lower Triassic of Gaimao section in Huaxi region of Guiyang. The measured data were used to plot the corresponding evolution curves. Based on the geological settings and lithology, the environmental causes of carbon isotopic variation were discussed and the analysis of palaeoenvironment at that time in the study area was performed according to changes in the curve. The results show that the marine ecological environment in this area was extremely damaged in the earliest Triassic. However, a transient recovery occurred in the early Griesbachian. After that, the ecological environment stepped into a long-term anoxic downturn until the middle Smithian.

A new bi-phase modulation(BPM) pulse UWB transmitter manufactured by discrete components is designed. The transmitter consists of mono-polar sub-nanosecond pulse generator, bipolar pulse generator, and high speed microwave switch. The transmitter transmit positive and negative pulses when controlled by clock signal as well as data ‘1’ and data ‘0’, respectively. Experimental results show that the generated pulse width is below 400 ps, the bandwidth is more than 1.5 GHz, and the data rate is up to 50 Mbps. It can be used in the pulse UWB(ultra-wideband) communication systems.

A linear transceiver optimization design based on the regularized block diagonalization (RBD) algorithm is proposed for the multiuser multiple-input multiple-output (MU-MIMO) relay downlink system. Considering that the source can not obtain the channel state information (CSI) of the second hop and the user can not obtain the CSI of the first hop, we optimize the transceiver design for the source-relay link and the relay-multiuser link, respectively. We use RBD algorithm at the relay to reduce the multiuser interference and noise, and then joint the receiver matrix to design an improved iterative algorithm based on the minimum mean square error (MMSE) criterion. Simulation results show that the proposed algorithm converges to a fixed value in a few iterations and it has a good bit error rate (BER) performance within the low signal-to-noise ratio (SNR) region.

This paper focuses on the Linux kernel-level privilege escalation vulnerabilities. Based on vulnerability detection thoughts of imitating attack, we design and develop an automated privilege escalation vulnerabilities exploiting system KernelPET. It reveals the typical process of exploiting privilege escalation vulnerabilities, and provides support to vulnerabilities defense. KernelPET is developed with today's mainstream vulnerability databases: exploit-db, securityfocus, etc. We test nearly one hundred of privilege escalation vulnerabilities by simulated attack, select 30 classic Linux kernel privilege escalation vulnerabilities, and load them into KernelPET exploiting code libraries. The system is tested on different cores and releases of the Linux platform. Experimental results show that KernelPET runs in multi-class releases of Linux system with good results.

Aiming at the characteristics of the heavy-tailed distribution of network flows, we propose a large flow identification algorithm, CBF-SS(counting Bloom filter and space saving), on the basis of analyzing advantages and deficiencies of hashing and counting methods used for large flow identification. It has the capability of combining the advantages of hashing and counting methods efficiently. The algorithm CBF-SS uses the counting Bloom filter to filter mass of small flows at first. Then, CBF-SS uses the SS (space saving) counting method to our large flows. Both theoretical and experimental results show that CBF-SS is very space-saving and time-efficient and it performs much better than the SS algorithm in the precision of large flow identification.

By means of weight function and Laplace's integral transform, two new integral inequalities and their equivalent forms are given, and their constant factors are proved being the best possible. Some meaningful results are obtained by selecting the special parameter values. Three integral operators are defined, and a relation among their norms is established.

Two new zinc(II) complexes are prepared by reactions of 2-(2'-Hydroxyphenyl)-2-oxazoline(Hoz) with Zn(acac)₂ and ZnCl₂, respectively. The complexes are characterized by IR and X-ray crystallography. Complex 1 is dinuclear zinc structure bridged by the oxygen atoms from oz^- ligand and acac^- ligand as terminal chelating ligands. Complex 2 is tetranuclear zinc structure bridged by both μ₂-oxygen from oz^- ligand and μ₃-methoxide. Complex 1 crystalizes in P2₁ space group with crystallographic parameters: a=9.384 2 Å(19), b=13.584 0 Å(30), c=11.138 0 Å(20), and β=96.250°(30). Complex 2 crystalizes in P-1 space group with the crystallographic parameters: a=9.551 7 Å(19), b=11.123 4 Å(22), c=11.182 6 Å(22), α=102.820°(30), β=114.988°(30), and γ=100.806°(30). The UV absorption and fluorescence emission spectra are measured in the methanol solution and solid state of complex 1, showing strong luminescence.

Massive data of three-dimension urban model and the bottleneck of computer hardware are great challenge to real-time rendering and visualization. The efficiency of the visualization can be significantly improved by organizing and managing the massive data effectively. This work proposes a method for organizing the three-dimension model using quad-tree spatial index and dispatching the data using the technique of view-based level-of-details (LOD). Additionally, a method for reconstructing the mesh using the texture as the render unit rather than the single model is proposed to accelerate the rendering process of rough models. Experimental results show that the proposed method works efficiently with a stable frame-rate of 30 fps and satisfies the requirements of the visualization of massive model data and the real-time interaction.

To improve the generalization ability of BP network for prediction, a BP neural network optimized with QPSO is proposed. This model uses the QPSO improved by δ potential trough to optimize the initial values of weights and thresholds of BP network. Then the data of each year's GDP are selected in training and prediction. The experiments show that the QPSO-BP network optimized by using δ potential trough produces stable prediction results. Compared with the prediction models of PSO-BP and BP, the proposed model has a better generalization ability and a higher accuracy. In addition, the calculation results of the improved QPSO-BP optimization algorithm model have smaller relative errors and average errors compared with the results of the models in the literature.

Ground-based synthetic aperture radar (GB-SAR) system has the advantage of continuous monitoring for a certain area day and night regardless of the weather condition. GB-SAR has broad application prospects in fields of mine, landslide, bridge settlement, earthquake deformation, and so on. The geometric mapping 3D matching between GB-SAR image and terrain data obtained in above fields is the key technology to combination of the system and the practical application. In this work, the relationship model between GB-SAR system and terrain data and the relationship model between GB-SAR system and SAR image are established. A 3D matching algorithm for geometric mapping based on range-azimuth condition is proposed, and the error characteristic in geometric mapping 3D matching process is considered. Finally, the effectiveness of our model and geometric mapping 3D matching algorithm is verified by the measured data.

A big data-based hot-spot real-time traffic monitoring system is implemented. The system analyzes the passenger flow of each entry in real time, estimates the regular pattern of stay passengers in each hot-spot, and scientifically allocates resources for operation and administration. The period of the real-time processing in the system is shorter than 15 seconds, and the error rate of object detection is less than 7.5%. The operation results show that the system achieves the practical goals of real-time collection, analysis, and processing for big data.