基于变神经网络的非线性最小方差预测控制器

提出基于变神经网络学习动态系统参数的最小方差预测控制器。其目的是通过在线学习 ,使控制器(MVPC)能适应被控对象参数变化和非确定性。提出的变神经网络由两部分组成 ,一部分是线性神经网络 (LNN) ,作为被控对象局部线性动态模型 ,另一部分是多层交叉回归神经网络 (DRNN) ,它近似为非线性动态模型。由于引进递推最小方差算法 ,本控制器运算速度相当快。仿真结果表明所提方法对非线性系统自适应控制是有效的     

利用初等变换求极大线性无关组

给出利用矩阵的初等变换求极大无关组的方法 ,并从理论上加以证明 .此法简单易行 ,且计算量小 .     

Fire Retardant Textiles: A Particular Reference to China and Kenya

ＴｅｘｔｉｌｅａｎｄＵｎｃｏｎｔｒｏｌｌｅｄＦｉｒｅｓ　　Ｔｅｘｔｉｌｅｓａｒｅｔｈｅｔｈｉｒｄｂａｓｉｃｒｅｑｕｉｒｅｍｅｎｔｆｏｒｈｕｍａｎｌｉｆｅａｆｔｅｒｆｏｏｄａｎｄｓｈｅｌｔｅｒ.Ｔｅｘｔｉｌｅｓａｒｅｕｓｅｄｆｏｒｔｈｅｒｍａｌｐｒｏｔｅｃｔｉｏｎａｇａｉｎｓｔｈｏｓｔｉｌｅｗｅａｔｈｅｒａｎｄｔｏｃｏｖｅｒｎａｋｅｄｎｅｓｓ.Ｔｅｘｔｉｌｅｓａｌｓｏｐｒｏｖｉｄｅｃｏｍ ｆｏｒｔｉｎｌｉｆｅｉｎｔｈｅｆｏｒｍｏｆｆｕｒｎｉｓｈｉｎｇｓ :ｕｐｈｏｌｓｔｅｒｙ ,ｃａｒｐｅｔｓ ,ｅｔｃ .…     

耗散准模腔场与激子相互作用的量子统计性质

研究了耗散准模腔场与激子相互作用的量子统计特性 ,给出了当腔场初始处于真空态而激子处于真空态与粒子数态大于 2的叠加态时的腔场与激子能量交换的表达式。研究结果表明 ,激子和腔场可以呈现亚泊松分布状态 ,激子与腔场之间的关联是经典的 ,不存在Cauchy Schwartz不等式的偏离现象。     

敏捷制造与中国制造业

阐述了敏捷制造的概念及其内涵，指出中国制造业实施敏捷制造战略要解决的主要问题及关键技术。     

苏云金杆菌库斯塔变种(HD-1)对菜青虫杀虫效果的研究

通过实验室饲料感染方法和大田喷杀对比试验研究，表明发酵培养基中添加花生饼可以提高HD—1的杀虫效果，且添加浓度与HD—1菌株毒力呈正相关，用添加了花生饼的固体浅盘发酵培养液比市售工业化生产的同类杀虫剂杀虫效果要高得多。     

利用单元状态的改变模拟采场破坏规律

采用能较好反映混凝土等脆性材料破坏特征的五参数Wilianm-Warnke强度准则破坏模型，利用有限单元法单元“处死”的功能模拟回采过程和相似材料模拟试验研究，初步总结出缓倾斜厚煤层悬移支架放顶煤顶板活动及矿压显现规律，为缓倾斜厚煤层悬移支架放顶煤制定较为合理的方案、安全技术措施等提供重要的依据和指导作用。     

Analysis of Pyrolysates for Phenol Formaldehyde Resin by Py-GC/MS

Pyrolysis of phenol formaldehyde resin has been investigated by Pyrolysis Gas Chromatography-Mass Spectroscopy at the different temperatures from 500℃ to 750℃. Its composition of pyrclysates has been analyzed. Several compounds, especially benzene, toluene, p-xylene could only be formed above 500-550℃. Howerver, peak intensities for some pbend derivatives were decreased at the higher temperature. During pyrolysis, for thermo-setting phenol formaldehyde resins, polymeric chain scissions take place as a successive removal of the monomer units from the polymeric chain. The chain scissions are followed by secondary reactions, which leads to a variety of compounds. Addition reactions can also take place among the double-bond compounds during pyrolysis.     

生产能力扩大的系统动态性研究

利用系统动力学的原理和建模方法描述了企业生产能力与企业物流、信息流与资金流所涉及的各种计划、生产、采购、仓储、销售及财务等职能的关系，利用因果环路图研究分析了生产能力与库存水平、交货期、市场份额、产品价格、成本、利润率、人力等多种因素的因果反馈关系，并构建了生产能力扩大的系统动力学模型来刻画其变化的复杂性与动态性，在两种市场需求（阶跃需求和泊松流随机需求）下模型的仿真结果表明，该模型能较好的反应生产能力扩大的动态行为。     

Detecting myocardial ischemia with 2-D CVIB imaging method--an in vivo animal experiment study

A 2-D cyclic variation of integrated backscatter (CVIB) imaging method was established for detecting myocardial ischemia. To demonstrate the feasibility and validity of this method, animal experiments were conducted. Acute myocardial ischemia was induced by occluding left anterior descending coronary artery in 10 anesthetized open-chest dogs. While scanning the normal hearts and the ischemic hearts with a B scanner, digital radiofrequency data were acquired by a real-time acquisition system in synchronism. The offline analysis to the radio-frequency signal with the 2-D CVIB imaging method was performed to verify the consistency between the imaging result and the design of the experiment. In addition, 4 dogs in experiment were treated with the heart pacemaker in order to investigate the influence of changing in heart rate on the detection of ischemic myocardium with the proposed method. The experimental result showed that the 2-D CVIB imaging method succeeded in detecting the ischemic myocardium and is a new non-invasive way for the cardiologists to both quantitatively and visually evaluate the contractile performance of the myocardium.