一种新颖的模糊形态联想记忆网络

模糊形态联想记忆网络FMAM具有较强的抗膨胀或腐蚀噪声能力，且可以模糊性解释。但抗混合噪声的能力很弱。而在实际中，随机噪声往往是混合型的，既有膨胀又有腐蚀噪声。为此提出了一种基于尺度空间的模糊形态联想记忆网络，并分析了其抗膨胀／腐蚀噪声和抗随机噪声的能力，它提高了自联想FMAM的抗随机噪声能力。通过仿真实验验证了该方法具有良好的性能。     

教学效果的BP神经网络评价

对教师教学效果的考察是需要多角度进行评价,无论是学生为教师打分,还是督导组的评议,给出的评价是带有一些主观因素,合理建立评价体系,将各个评价指标客观化,使用综合评价向量作为输入,经由BP神经网络输出得到合理的分数。实验仿真表明,通过训练的BP反向传播网络可模拟一个稳定的评价系统。     

对数字家庭网络新核心——家庭网关的分析

伴随互联网、物联网的发展,家庭设施呈数字化发展趋势,数字家庭概念的轮廓逐步清晰,包括家庭媒体设备间的数据分享、自动化家居的远程监控、对家庭网络门户及家庭内数据进行安全管理。随之衍生出对数字家庭网络新核心—家庭网关的需求。文章对家庭网关的形成、发展及未来可能性作了分析与预测,并结合国内外在家庭网关需求上的差异作了描述与分析。     

基于小波神经网络的电动机变频系统故障检测的研究

文章提出应用小波分析技术提取故障特征并以模糊神经网络进行故障预测.由此建立了基于小波与神经网络的电动机变频调速故障的智能检测系统,并用MATLAB进行仿真,证明了该方法对电动机变频系统故障的检测是有可行性的.     

Wireless LANs: From WarChalking to Open Access Networks

This work discusses the evolution of W-LANs from their current status of wireless termination of LAN services to a possible global infrastructure where the access networks become open to multiple operators and a vehicle of a win-win scenario, where both users and operators benefit from the new network architecture. The idea of Open Access Networks (OANs) can go beyond wireless HotSpots and be generalized to a generic shared access infrastructure that fosters service operators competition and drastically reduces the cost of last mile coverage.The general concept of Open Access Networks is detailed, highlighting its difference with the more traditional model of vertical integration of the access network into the global service. About the OANs development, it is shown how to support the quick and smooth evolution of the infrastructure toward a widespread and reliable communication support.Business models are discussed by mentioning the different actors, the market organization and the different organization forms.The final part of the paper is devoted to technical challenges such as access control, security, privacy, roaming, resource exploitation and service differentiation. As an example of how to tackle these problems, we discuss a pricing technique devoted to resource management and billing support.In addition we present a simulation on how the OAN concept can speed-up the deployment of broadband access in a real case.A preliminary version of this paper with the title “Global Growth of Open Access Networks: from WarChalking and Connection Sharing to Sustainable Business” was presented at WMASH 2003. Roberto Battiti received the Laurea degree from the University of Trento, Italy, in 1985 and the Ph.D. degree from the California Institute of Technology (Caltech), USA, in 1990. He has been a consultant in the area of parallel computing and pattern recognition and since 1991 he has been a faculty member at the University of Trento, where he is now full professor of Computer Networks. His main research interests are heuristic algorithms for optimization problems, in particular reactive search algorithms for maximum clique, maximum satisfiability, graph coloring, networks and massively parallel architectures, code assignment in wireless and cellular networks, protocols for pricing and Quality of Service in wireless networks.Prof. R. Battiti is currently Dean of the international Graduate School in Information and Communication Technologies at Trento, Deputy Dean at the Faculty of Science, member of the advisory committee for the future Telecommunications Plan of the Autonomous Province of Trento. Prof. R. Battiti is associate editor of various scientific journals. He is the author of more than fifty scientific publications, including special issues dedicated to experimental algorithmics and wireless on-demand networks.R. Battiti is a member of IEEE Computer Society and ACM Sigmobile. Renato Lo Cigno is Associate Professor at the Department of Computer Science and Telecommunications (DIT) of the University of Trento, Italy. He received a Dr. Ing. degree in Electronic Engineering from Politecnico di Torino in 1988. From 1989 to 2002 has been with the Telecommunication Research Group of the Electronics Department of Politecnico di Torino.From June 1998 to February 1999, he was at the CS Department at UCLA as Visiting Scholar under grant CNR 203.15.8. He is coauthor of more than 90 journal and conference papers in the area of communication networks and systems. His curent research interests are in performance evaluation of wired and wireless networks, modeling and simulation techniques, flow and congestion control, as well as network management and architecture. Renato Lo Cigno is member of the IEEE Communication Society and ACM Sigcomm. Mikalai Sabel received his master degree in radio physics from Belorussian State University in 2002. He is now pursuing his Ph.D. at the international Graduate School of Information and Communication Technologies at University of Trento, Italy. His research interest include wireless LANs, pricing and incentive engineering, IEEE802.11 MAC protocol. His current research is focused on optimization algorithms that can provide stability and maximum global welfare in a non-cooperative environment where singles tries to maximize their own benefit function. Experimentation of such schemes to pricing techniques and micro-payments in TLC networks provides the application environment of the fundamental research. Fredrik Orava is associate professor at KTH, Stockholm, Sweden. He conducts research into scalable (in terms of cost, capacity, number of users and devices etc.) communication system architectures and technologies. He has a MSc in engineering physics from Uppsala University and a PhD in computer systems also from Uppsala University. He previously held positions as researcher at the Swedish Institute for Computer Science (SICS); senior lecturer and manager of the telecommunications systems laboratory at the department for teleinformatics, KTH; acting professor in telecomunication systems at KTH; director for the Swedish Centre for Internet Technologies; and vice president of Dynarc AB with world wide responsibility for product management. Dynarc AB develops, manufactures and sells IP routers for optical IP networks. Bjorn Pehrson is a professor of telecommunication systems at KTH, the Royal Institute of Technology in Stockholm, since 1992, where he has also served as a department chairman and vice dean. He had his PhD from Uppsala University 1975 where he also served as senior lecturer and established a research group in computer and communication technology. During the period 1985-1992 he participated in the establishment of the Swedish Institute of Computer Science. Björn Pehrson research interests are currently focussed on open network topologies allowing different actors forming new value chains to build networks together based on different, sometimes very local business models. The results are applied in testbeds, especially in rural areas and in developing countries.This revised version was published online in August 2005 with a corrected cover date.     

Wavelet neural networks for modelling high pressure inactivation kinetics of Listeria monocytogenes in UHT whole milk

The aim of the present work was to investigate the applicability of a Wavelet Neural Network to describe the inactivation pattern of Listeria monocytogenes by high hydrostatic pressure in ultra high temperature (UHT) whole milk, and evaluate its performance against models used in predictive microbiology such as the re-parameterized Gompertz and modified Weibull equations. A comparative study with linear partial least squares regression (PLS-R) as well as neural network (NN) models demonstrated on the same dataset has been also considered. Milk was artificially inoculated with an initial population of the pathogen of ca. 10⁷ CFU/ml and exposed to a range of high pressures (350, 450, 550, 600 MPa) for up to 40 min at ambient temperature (ca. 25 °C). Typical survival curves were obtained including a shoulder, a log-linear and a tailing phase. Increasing the magnitude of the applied pressure resulted in increasing levels of inactivation. Modelling approaches provided good fit to experimental training data as inferred by the low values of the root mean squared error (RMSE) and the high values of regression coefficient (R²). Models were validated at 400 and 500 MPa with independent experimental data. First or second order polynomial models were employed to relate the inactivation parameters to pressure, whereas the wavelet network as well as the PLS and NN models were utilised as a one-step modelling approach. The prediction performance of the proposed learning-based network was better at both validation pressures. The development of accurate models to describe the survival curves of micro-organisms in high pressure treatment would be very important to the food industry for process optimisation, food safety and would eventually expand the applicability of this non-thermal process.     

基于BP神经网络的白酒探汽上甑方法研究

目的:针对目前白酒"探汽上甑"工艺在自动化酿造过程中存在探汽准确率低的问题,提出一种基于BP神经网络的探汽方法.方法:使用红外相机采集上甑蒸馏过程中酒醅表层图像获得酒蒸汽温度分布特征,按照上甑工艺特点将图像分类,结合图像预处理技术对图像温度特征提取,通过神经网络分类模型训练得到探汽模型.结果:通过实验验证,该方法的探汽...     

神经网络建模及其在食品包装机械中的应用

针对液压伺服系统的强非线性，提出了一种神经网络建模方法。在系统反馈输入信息的基础上，构造一样条函数和常数作为神经网络的输入量，以扩充网络输入量矩阵，使用传统BP算法离线训练网络，得到神经网络模型。应用于食品机械液压伺服系统的建模结果表明，所提出的神经网络模型训练过程收敛速度较快，模型精度高，可有效避免网络陷入局部最优值，而且不需要更多的系统信息。     

基于神经网络算法的复合材料板声发射源定位

使用Levenberg-Marquardt训练函数的神经网络算法在复合材料板上进行声发射定位,采用声波到达时间作为输入向量,训练得到了到达各个传感器的声波触发时间与板上坐标的良好的映射关系.考虑到波形衰减,用传统的固定门槛提取的到达时间容易与实际情况偏差较大,讨论了一种相对准确的适合工程使用的到达时间提取方法.     

BP神经网络在摩擦学设计计算中的应用

用双隐层ＢＰ人工神经网络,建立了丝杆螺母副的磨损率与滑动速度关系的数学模型。该模型可用于准确地计算丝杆螺母副和蜗轮蜗杆副的磨损率,可十分方便地用于摩擦学程序设计。采用Ｌ－Ｍ规则进行神经网络学习训练可使网络收敛快,误差小。风络输出结果与实验结果比较有极好的吻合性。该神经网络为工程设计人员,在摩擦学设计时提供有效的计算工具。