奥氏体不锈钢表面改性层耐蚀性实验研究：Ⅰ.孔蚀和均匀腐蚀性能

用等离子体源离子渗氮技术在３８０℃处理１Ｃｒ１８Ｎｉ９Ｔｉ奥氏体不锈钢,获得了高氮面心相（δＮ）表面改性层。γＮ相层在０．５％－０．６％ＮａＣｌ溶液中具有显著优于原始不锈钢的耐孔蚀性能;在０．５ｍｏｌ／Ｌ Ｈ２ＳＯ４溶液中具有相当于原始不锈钢的耐均匀腐蚀性能。γＮ相的氮浓度影响其耐孔蚀性能,但对其耐均匀腐蚀性能则没有明显影响,γＮ相的室温浸泡与电化学测试取得了规律相同的实验结果。     

总承包项目设备材料采购价格控制要素的探讨

本文通过对总承包项目设备材料采购过程中几个影响最终采购价格的环节的分析,提出了影响采购价格的控制要素.     

一种求解机组组合问题的内点半定规划GPU并行算法

针对内点法求解机组组合问题的半定规划(SDP)模型时大规模线性方程组计算时间太长的问题,提出一种基于图形处理器(GPU)的Krylov子空间并行算法.该算法采用预条件处理的拟最小残差法(QMR法),并以矩阵分块技术为基础,在CSR存储格式下使用GPU实现Incomplete Cholesky并行预处理矩阵的计算.通过对不同规模线性方程组的计算分析表明,与传统的Cholesky直接法相比,QMR并行算法具有速度和存储优势,可获得良好的并行加速比.10～100机6个系统的仿真结果也表明,该SDP并行内点法在减少计算时间的同时可求得近似最优解.     

大亚湾核电站调试锅炉工程总体布置和系统设计

本文介绍了大亚湾核电站调试锅炉工程总体布置和系统设计以及实施效果。     

通用小型汽油机工况法燃油消耗率测试方法解析

GB/T 37692-2019《非道路移动机械用小型点燃式发动机工况法燃油消耗率限值与测量方法》标准于2019年6月4日发布,将于2020年1月1日正式实施。本文根据标准内容,从试验条件、仪器设备、试验工况的选择、试验流程和数据计算等要点,对通用小型汽油机工况法燃油消耗率测试方法要求进行了解读,为检测机构和企业进行通用小型汽油机油耗测试提供参考。     

基于CUDA GPU的中期冲突探测加速算法研究

中期冲突探测(MTCD)是空中交通管制的重要决策支持工具。针对Prandini概率型中期冲突探测算法存在计算速度慢、探测精度不够高的缺点,提出基于GPU CUDA的加速算法。利用CUDA线程并行处理能力和GPU的计算能力,对算法进行了重新设计与优化。通过仿真实验,证明了在NVIDIA Geforce 8800GTS显卡上可得到50倍以上的加速性能,可以满足大型空管自动化系统20 min内中期冲突探测的实时性和精度要求。     

One‐dimensional analysis of thin‐walled closed beams having general cross‐sections

This paper presents a new one‐dimensional theory for static and dynamic analysis of thin‐walled closed beams with general cross‐sections. Existing one‐dimensional approaches are useful only for beams with special cross‐sections. Coupled deformations of torsion, warping and distortion are considered in the present work and a new approach to determine sectional warping and distortion shapes is proposed. One‐dimensional C⁰ beam elements based on the present theory are employed for numerical analysis. The effectiveness of the present theory is demonstrated in the analysis of thin‐walled beams having pillar sections of automobiles and excavators. Copyright © 2000 John Wiley & Sons, Ltd.     

Economic evaluation of independent power producers viewed from total generating costs of electric power systems

In Japan, electric power utilities purchase electricity from independent power producers (IPPs) through competitive bidding; the IPP evaluation is based on the avoided costs of corresponding generators of utilities. In this evaluation, however, nonpricing factors such as power flow constraints in the electric power system cannot be considered. In this paper, we propose a new approach to evaluate economic impacts of IPPs, the IPP electricity being priced on the basis of total generating costs of the electric power utilities. Such a price to purchase the electricity of IPPs is referred to as the “break‐even cost.” The main results are summarized as follows: 1) Break‐even cost depends on not only the power flow constraints in the system, but also various IPP factors such as generating pattern, location, and capacity, 2) Break‐even cost for the base‐type IPP is higher than the avoided cost of corresponding utility generators, because IPPs located on demand sides can reduce the transmission power loss in the electric power system, 3) Break‐even cost is affected by available capacity of utility generators, especially for the peak‐type IPP generating only during peak demand periods. © 2000 Scripta Technica, Electr Eng Jpn, 133(2): 20–30, 2000     

基于MOOC+SPOC的《数字图像处理》课程改革研究

将传统教学方法、现代化教学新方法有机的结合起来,把互联网大时代下的现代化教学工具与传统教学课堂相辅相成的结合,发挥各自的优势,提高了教学效果,让学生收获更多的知识,培养出更多元化的优秀人才.将基于MOOC+SPOC的教学模式引入到传统数字图像处理理论教学过程中,有效解决了传统教学环节中存在的问题,为学生提供了更加灵活、...     

图数据库中基于GPU的图分析计算方法

现有的图数据库对于在线分析操作大多采用基于CPU的分布式图计算引擎（如GraphX）,但CPU核心数量有限的不足会导致计算效率低下,同时集群间的同步也会产生额外的通信开销。通过使用图形处理单元（GPU）对图计算进行加速,设计并实现图处理系统RockGraph。该系统能够根据用户需求从图数据库中提取出包含核心信息的子图,经过数据格式转换后,利用JNI工具调用动态链接库,采用超显存GPU图计算框架进行在线分析,并将计算结果写回图数据库。实验结果表明,与基于CPU的分布式图计算系统相比,RockGraph的图分析效率可提高3倍~5倍。