模拟退火-BP神经网络法测定无机铬的形态

提出了用神经网络方法测定无机铬的形态。在Back Propagation(BP)算法的基础上,吸取了模拟退火算法的优点,解决了BP算法容易陷入局部极小点的问题。经实际样品分析,结果满意。     

Long-Term Changes in a Heat Exchanger Steel

The steel ASTM A213 P22 is used for superheater outlet header in power plants. During duty cycles lasting over tens of years the temperature is of the order of 545°C and the pressure 125 atm. The microscopic changes in these steels are hard to analyze under working conditions, but they are believed to be responsible for the appearance of creeps in such devices. Investigation of the microscopic modifications will help to predict future failures due to creeps, increasing the reliability and saving a lot of money.We report the investigation of such changes in ASTM A213 P22 steel in use for 30 years as a Superheater header at Haifa power station. The analysis has been carried out using Mössbauer Spectroscopy, X-ray diffraction and optical microscopy. As a result it seems that the morphology and phase change of the carbide phases M₃C/M₇C₃ and M₂₃C₆ are the most prominent changes leading to the material failures.     

On Numerical Solution of Shape Inverse Problems

The new method is proposed for the numerical solution of a class of shape inverse problems. The size and the location of a small opening in the domain of integration of an elliptic equation is identified on the basis of an observation. The observation includes the finite number of shape functionals. The approximation of the shape functionals by using the so-called topological derivatives is used to perform the learning process of an artificial neural network. The results of computations for 2D examples show, that the method allows to determine an approximation of the global solution to the inverse problem, sufficiently closed to the exact solution. The proposed method can be extended to the problems with an opening of general shape and to the identification problems of small inclusions. However, the mathematical theory of the proposed approach still requires futher research. In particular, the proof of global convergence of the method is an open problem.     

基于反向传播神经网络的阻抗断层图像重建新方法

电阻抗成像是通过对物体表面电压、电流的测量来重建物体内部阻抗分布或变化图像的一种新颖计算机断层成像技术。阻抗断层图像重建是一种病态的、非线性的逆问题。提出了一种全新的阻抗断层图像重建方法，它利用反向传播神经网络来表征物体内部阻抗变化位置与物体表面电压变化大小的非线性映射关系，从而可以根据对物体表面测量电压的变化先准确定位阻抗变化区域，再用线性近似方法重建阻抗变化图像，这种方法不仅具有一定的抗噪能力，而且成像精度和空间分辨率都大大好于逆投影方法。     

基于模糊神经网络的自适应PID控制器

提出了一种应用于自适应PID控制器的神经网络与模糊控制相结合的算法,该算法可以有效地解决普通PID控制器依赖于对象的数学模型的缺点,可实现控制系统的在线自适应调整,可满足实时控制的要求。仿真结果表明,基于模糊神经网络整定的PID控制器具有较好的自学习和自适应性,具有较快的响应速度。     

The homotopy categorical crossed module of a CW-complex

Crossed modules have longstanding uses in homotopy theory and the cohomology of groups. The corresponding notion in the setting of categorical groups, that is, categorical crosses modules, allowed the development of a low-dimensional categorical group cohomology. Now, its relevance is also shown here to homotopy types by associating, to any pointed CW-complex (X,∗), a categorical crossed module that algebraically represents the homotopy 3-type of X.     

基于BP网络分类器的交通标志识别

针对中国全部3大类116个交通标志,即禁令标志、指示标志、警告标志,用BP网络实现分类功能.实验中使用了3种测试集,即加高斯噪声、水平扭曲和日本交通标志实景图,对BP网络的分类性能进行了测试.实验结果表明:用BP网络实现交通标志粗分类功能的效果是比较理想的,对交通标志的颜色失真和形状失真均具有较好的容错性和鲁棒性.     

Thermoelectric properties of tin clathrates under hydrostatic pressure

We report the temperature dependence of electrical resistance (R) and thermopower (S) of clathrate Cs₈Sn₄₄ under high pressure up to 1.2 GPa. We observe a reversible gap widening, prominent relaxation effect of R, irreversible increase of |S| under high pressure. We also find that the power factor S²σ (σ: electrical conductivity) reaches a maximum at pressure of 0.3 GPa. Comparison of the experimental results with band structure calculations suggests that the intrinsic vacancy in the clathrate structure of Cs₈Sn₄₄ plays an important role in transport properties under high pressure. Measurements on Cs₈Zn₄Sn₄₂, a clathrate which has defects other than vacancies, are compared with Cs₈Sn₄₄. The results indicate that replacing Sn by Zn has similar effect as the intrinsic vacancy on S.     

A multi-target approach for isolated word recognition using artificial neural network

I.TntroductionStatisticalandneuralnetworkmcthodsforpatternclassiflcationusesignificantlydifTerentapproachesintrainingaclassificr.Inthestatistica1approach,thcformationofaclassifier1arge1ydependsonthestatisticsofthetrainingpatternsand,insomecases,theassumptionsaboutthedistributionofthepopulation.Theneuralnetworkmethodisnon-parametricandcanbeadaptivcinthetrainingprocessl'l.Becauseofitssimplicityandflexibility,theneuralnetworkhasbecnincreasing1yusedforpatternclassiflcation.Itisnowwe11knownthatan…     

带有边集限制的最均匀支撑树问题

本在无向网络中，建立了带有边集限制的最均匀支撑树问题的网络模型．中首先解决最均匀支撑树问题，并给出求无向网络中最均匀支撑树的多项式时间算法；然后，给出了求无向网络中带有边集限制的最小树多项式时间算法；最后，在已解决的两个问题的基础上解决了带有边集限制的最均匀支撑树问题．