水环境承载力与污染物总量分配研究进展

水环境承载力的大小具有明显的时间性、空间性、动态性、地区性等特点.它不但具有环境属性,还具有社会经济属性.它是污染物总量分配和总量控制的基础.总量控制是总量分配的目的.总量控制是通过控制某一区域污染源允许排放总量,并优化分配到各污染源,以实现水环境目标的环境规划管理措施.其本质是要寻求某一区域在环境质量目标与技术经济条件之间的最佳结合点.TMDL技术是这一领域最新技术成果的代表.     

Refinement of primary Si grains in Al–20%Si alloy slurry through serpentine channel pouring process

In this study, a serpentine channel pouring process was used to prepare the semi-solid Al–20%Si alloy slurry and refine primary Si grains in the alloy. The effects of the pouring temperature, number of curves in the serpentine channel, and material of the serpentine channel on the size of primary Si grains in the semi-solid Al–20%Si alloy slurry were investigated. The results showed that the pouring temperature, number of the curves, and material of the channel strongly affected the size and distribution of the primary Si grains. The pouring temperature exerted the strongest effect, followed by the number of the curves and then the material of the channel. Under experimental conditions of a four-curve copper channel and a pouring temperature of 701℃, primary Si grains in the semi-solid Al–20%Si alloy slurry were refined to the greatest extent, and the lath-like grains were changed into granular grains. Moreover, the equivalent grain diameter and the average shape coefficient of primary Si grains in the satisfactory semi-solid Al–20%Si alloy slurry were 24.4 μm and 0.89, respectively. Finally, the refinement mechanism and distribution rule of primary Si grains in the slurry prepared through the serpentine channel pouring process were analyzed and discussed.     

基于MPLS IP VPN机制的拥塞处理分析

简要地分析了目前IP VPN技术存在的弱点,对引入MPLS技术后带来的拥塞进行了分析和研究,给出了解决拥塞的几种可能途径,并进行了相应的评价和讨论。     

Effects on Structure and Abrasion Resistance of GCr15 Steel by Surface Gas-Phase RE Diffused Permeation with Laser Melting Solidification

The effects on abrasion resistance and the microstructure of GCr15 steel surface by the compound technology of permeating RE combined with laser melting modification was studied. The results show that after compound treatment, the abrasion resistance of samples has been improved significantly and the weight loss has been reduced to 14% of blank sample; the microstructure has been denser and more uniform than that of untreated; meanwhile, the grain has been refined and the concentration gradients of the elements permeated have been decreased obviously.     

Dynamic overlapped spectrum allocation based on potential game in cognitive radio networks

In order to make full use of wireless spectrum resources,the behavior of cognitive radio （CR）for dynamic spectrum allocation is analyzed based on the game theoretic framework.The traditional spectrum allocation schemes consider the spectrum allocation among independent frequency bands only,without taking into account mutually overlapped frequency bands.For this reason,an optimal allocation etiquette is defined to promote the cross characteristic of the frequency bands in a dynamic spectrum allocation model.New interference operator and interference temperature constraints are introduced in order to realize calculation of the interference,and the corresponding spectrum allocation scenario can be further formulated as a potential game.Based on the characteristic of dynamic selection using the game theory and the interference avoidance rule of interference temperature,the robustness of CR networks is increased and the scenario is more suitable for the dynamic changing of actual wireless communication and energy saving communication systems.Simulation results show that the signal to interference and noise ratio （SINR） level can be significantly improved through the optimal allocation of any available spectrum.The utilization rate of spectrum and throughput of overall CR networks are increased by fully utilizing the spectrum resources in the dynamic spectrum allocation model.     

基于支持向量机和多层感知器的多病类诊断基因选择方法

超高维基因空间超小样本的基因选择问题是基因芯片技术的挑战性课题之一,对于解决维数发难问题和获得诊断基因具有重要的理论和实际意义.针对DNA微阵列数据的超高维空间和超小样本特性,提出了实现二病类样本有效分类的基于支持向量机(SVM)和leave-one-out的递增基因选择方法,并针对多病类情况的基因选择问题,提出了在区分两两病类所选基因基础上的基于多层感知器(MLP)和leave-one-out/MLP leave-4-out的递减基因选择方法.对于真实MIT数据(7129个基因、两个病类、72个样本)和NCI数据(2308个基因、64个样本、4个病类),用文中方法选择出了11个和6个诊断基因,与传统SNR方法的比较结果表明所选出的这些诊断基因有很好的诊断性能(即推广能力),既可有效诊断相应疾病,同时极大降低芯片和诊断费用,表明了该方法的有效性.