洪水预报智能模型在中国半干旱半湿润区的应用对比

为探究不同洪水预报智能模型在我国半干旱半湿润区的应用效果,选用决策树、多层感知器、随机森林和支持向量机4种模型在陕西省3个典型流域进行逐时洪水预报;选择相关系数、纳什效率系数、均方根误差、平均绝对误差和相对误差等评价指标,比较不同预见期下4种模型在半干旱半湿润典型流域洪水预报的适用性。结果表明:在短预见期预报中,4种模型在半湿润区典型流域均可获得较高的预报结果,在半干旱典型流域模拟精度相对偏低,仅支持向量机模型满足预报要求;随着预见期延长,不同模型性能变化差异大,支持向量机模型整体稳定,在小流域实时洪水预报中具有明显优势;随机森林模型与决策树模型精度随预见期延长而缓慢下降,前者适应性更好;多层感知器模型精度随预见期延长而骤减,模型稳定性较差。     

印尼某火电厂Ａ２—Ａ４区人工边坡失稳机制研究

以印尼某火电厂涉外项目Ａ２—Ａ４区的问题人工边坡为研究对象,通过现场地勘、实际变形失稳过程、数值模拟等方法综合研究了该地人工边坡失稳机制。在此基础上,探讨了边滑坡加固措施的作用效果及合理性。所得主要结论:Ａ２—Ａ４区人工边坡变形失稳机制主要为软弱地层（煤层）切穿暴露所造成的组合滑移以及坡体沿土岩分界面所造成局部变形,变形破坏模式接近圆弧滑动,人工边坡中的软弱夹层是导致边坡失稳的主控因素;Ａ２、Ａ４区综合考虑桩锚＋坡面锚杆的方式进行加固,煤层和全风化泥岩及上部土层内的组合变形没有发展贯通趋势,边坡整体安全。Ａ３区在加固作用下存在沿二级坡脚＋煤层＋全风化泥岩＋上部土层内越顶组合滑移风险以及基坑开挖造成基坑侧壁坍滑风险,需要引起重视并加强监测。     

Semi-automated user agents in distributed change management

Flexible integration of distributed design and manufacturing activities is one of the key issues in applying just-in-time principles along the logistics chain. The potential of AI based approaches in this field is clear, but the inherent asynchrony, concurrency, and inconsistency of distributed operations has reduced enthusiasm so far.This paper discusses coordination in the context of geographically distributed manufacturing of one-of-a-kind products. Cooperation of partners is based on semi-automated user agents which operate as an integrative layer between more dedicated tools and systems. The management of order changes in elevator manufacture is used as a case example.The work reported here is part of the DIMUN project (DIMUN, 1989) in the RACE program.     

一种基于网络安全风险评估的入侵检测方法

针对传统的入侵检测系统存在报警数量大、误报率高等缺陷,提出了一种基于网络安全风险评估的入侵检测方法,该方法基于入侵检测结果,引入抗体浓度随入侵强度动态变化这一人工免疫理论的最新研究成果进行网络安全风险的计算,然后根据当前网络面临的实时安全风险动态设置报警策略。实验结果表明,该方法能够实时、定量地计算主机和网络所面临的风险,并极大地降低报警数量和误报率。     

入侵检测主流技术及发展动态

入侵检测技术是网络安全的主要技术和网络研究的热点,入侵检测方法包括基于数据挖掘、粗糙集、模式识别、支持向量机和人工免疫等主要技术,详细分析了各种检测方法在入侵检测应用中的优缺点。通过回顾研究人员近期的研究成果,提出了该技术的主要发展方向,为进一步研究提供参考。     

放电故障紫外检测系统及其状态识别方法

紫外成像检测技术是一种检测高压设备故障的新方法;在此技术上研究设计了一种放电故障紫外检测系统,该系统能对实时采集的紫外图像进行相关的技术处理,同时可对故障点进行准确定位;根据数据建立了设备状态模型并提出一种基于人工神经网络的状态识别方法,应用MATLAB工具进行神经网络的设计和模拟,仿真结果达到预期效果;该系统可以准确地诊断设备的故障和运行状况,有较好运用前景和理论意义。     

气动人工肌肉手臂的神经网络Smith预估控制

针对含时滞d的1关节气动人工肌肉(PAM)手臂,用三层递归神经网络(RNN),建立PAM手臂包含时滞的模型(即非线性Smith预估器),并超前d步预测PAM手臂的输出角度.将此超前d步的预测值作为反馈量,与设定值相比较得到的误差作为PID控制器输入量,实现Smith预估PID控制.同时每一步都用RNN模型当前时刻的输出值与PAM手臂当前时刻实际输出值之差的平方做为RNN权值的在线调整准则对RNN预测模型的权值进行在线调整,以自适应PAM手臂的不确定性和时变性.使用Matlab通过串口和研华亚当模块对实物PAM手臂进行控制,控制效果表明所提出的Smith预估PID控制算法比常规PID控制算法的性能有显著提高,证明所提出的算法是有效的和切实可行的.     

Distributed constraint optimization with MULBS: A case study on collaborative meeting scheduling

This paper introduces MULBS, a new DCOP (distributed constraint optimization problem) algorithm and also presents a DCOP formulation for scheduling of distributed meetings in collaborative environments. Scheduling in CSCWD can be seen as a DCOP where variables represent time slots and values are resources of a production system (machines, raw-materials, hardware components, etc.) or management system (meetings, project tasks, human resources, money, etc). Therefore, a DCOP algorithm must find a set of variable assignments that maximize an objective function taking constraints into account. However, it is well known that such problems are NP-complete and that more research must be done to obtain feasible and reliable computational approaches. Thus, DCOP emerges as a very promising technique: the search space is decomposed into smaller spaces and agents solve local problems, collaborating in order to achieve a global solution. We show with empirical experiments that MULBS outperforms some of the state-of-the-art algorithms for DCOP, guaranteeing high quality solutions using less computational resources for the distributed meeting scheduling task.     

人工神经网络技术在电工产品可靠性实验中的应用(理论部分)

本文简要介绍了人工神经网络，BP神经网络的结构、训练和学习规则以及人工神经网络非线性时间序列预测的应用概况。     

A Novel Artificially Humanized Anti-Cripto-1 Antibody Suppressing Cancer Cell Growth

Cripto-1 is a member of the EGF-CFC/FRL1/Cryptic family and is involved in embryonic development and carcinogenesis. We designed a novel anti-Cripto-1 artificial antibody and assessed the recognition to the antigen and the potential to suppress the growth of cancer stem cells. First, single chain antibody clones were isolated by bio-panning with the affinity to recombinant Cripto-1 protein from our original phage-display library. Then, the variable regions of heavy chain VH and light chain VL in each clone were fused to constant regions of heavy chain CH and light chain CL regions respectively. These fused genes were expressed in ExpiCHO-S cells to produce artificial humanized antibodies against Cripto-1. After evaluation of the expression levels, one clone was selected and the anti-Cripto-1 antibody was produced and purified. The purified antibody showed affinity to recombinant Cripto-1 at 1.1 pmol and immunoreactivity to cancer tissues and cell lines. The antibody was available to detect the immunoreactivity in tissue microarrays of malignant tumors as well as in Cripto-1 overexpressing cells. Simultaneously, the antibody exhibited the potential to suppress the growth of human colon cancer derived GEO cells overexpressing Cripto-1 with IC₅₀ at approximately 110 nM. The artificially humanized antibody is proposed to be a good candidate to target cancer cells overexpressing Cripto-1.