Integrated Planning of Production and Engineering Process Improvement

The authors present a planning model to allocate equipment capacity between production that generates immediate revenue and engineering process improvement activity that results in increased future output. The benefit of engineering activity is modeled as a concave function of the total number of engineering lots processed to date, while the production facility is represented by a nonlinear clearing function capturing the nonlinear relationship between resource utilization and cycle time. We analyze the model to develop structural results and illustrate its behavior with numerical examples.      

A production planning methodology for semiconductor manufacturingbased on iterative simulation and linear programming calculations

We introduce a methodology for automated production planning of semiconductor manufacturing based on iterative linear programming (LP) optimization and discrete-event simulation calculations. The LP formulation incorporates epoch dependent parameters for flow times from lot release up to each operation on each manufacturing route. LP-derived release schedules are used as input to the simulation model, from which statistics on flow times are collected and used to reformulate the LP model for a revised planning calculation. Iteration continues until satisfactory agreement between simulation and LP models is obtained. We demonstrate in experiments on an industry data set that a relatively small number of iterations is required to develop a production plan correctly characterizing future flow times as a function of factory load and product mix. The methodology makes possible automated production planning of semiconductor manufacturing on an engineering work station     

基于HBF神经网络的自适应观测器

传统的RBF(Radial Basis Function)神经元基函数通常把高斯类型与单一宽度作为每个神经元的激活函数,这些特性限制了网络神经元的性能,特别是在处理复杂的非线性建模问题上.为了克服这个限制,本文应用了具有类似RBF网络,但激活函数不同-超基函数HBF(Hyper Basis Function)的网络.结合RBF网络,分析了HBF网络的结构、基函数形式及基函数对网络的影响,利用决策树算法计算了网络中心.在此基础上,提出了一种基于HBF神经网络的自适应观测器设计方法,并通过引入Lyapunov函数,证明了这种观测器设计方法的稳定性;最后通过仿真验证了这种HBF神经网络观测器能很好地观测系统的状态值.     

非线性动态网络的离散算法及仿真

在非线性网络响应分析中,采用Volterra级数法可以导出与线性系统传递函数相似的非线性传递函数,能使非线性系统用线性化和系统化方法达到精确分析.文中给出了非线性网络响应的Volterra级数解的连续算式,为解决连续算式计算麻烦的问题,提出用方波脉冲技术处理用Volterra级数表示法描述的非线性网络响应与激励之间关系的一组广义卷积积分的迭加计算,从而得到非线性网络响应求解的Volterra级数解的离散算式.仿真表明该算法求出的非线性网络响应与真实模型曲线十分逼近,证明了它的有效性.     

Nonlinear models for the intermodulation analysis of FET mixers

An accurate, detailed analysis program has been developed for intermodulation distortion (IMD) simulation of FET mixers. This program is very efficient at calculating the IMD from multiple RF inputs. We have proposed a simplified nonlinear model for IMD analysis of FET gate mixers. The accuracy of the simplified model has been verified experimentally using two different MESFET mixers and one HEMT mixer at X band. All the tests show good agreement between measured results and the calculated results for second- and third-order IMD. The simplified model is based on modeling the derivative of the device transconductance by a sum of a Gaussian function and a linear function of the gate voltage. Drain bias dependence is ignored. The advantage of this model is that it can be used for both MESFET and HEMT mixers, and its fitting parameters can be easily determined from a nonlinear characterization of the devices at low frequencies     

Blind identification of quadratic nonlinear models using neuralnetworks with higher order cumulants

A novel approach to blindly estimate kernels of any discrete- and finite-extent quadratic models in higher order cumulants domain based on artificial neural networks is proposed in this paper. The input signal is assumed an unobservable independently identically, distributed random sequence which is viable for engineering practice. Because of the properties of the third-order cumulant functions, identifiability of the nonlinear model holds, even when the model output measurement is corrupted by a Gaussian random disturbance. The proposed approach enables a nonlinear relationship between model kernels and model output cumulants to be established by means of neural networks. The approximation ability of the neural network with the weights-decoupled extended Kalman filter training algorithm is then used to estimate the model parameters. Theoretical statements and simulation examples together with practical application to the train vibration signals modeling corroborate that the developed methodology is capable of providing a very promising way to identify truncated Volterra models blindly     

RBFN restoration of nonlinearly degraded images

We investigate a technique for image restoration using nonlinear networks based on radial basis functions. The technique is also based on the concept of training or learning by examples. When trained properly, these networks are used as spatially invariant feedforward nonlinear filters that can perform restoration of images degraded by nonlinear degradation mechanisms. We examine a number of network structures including the Gaussian radial basis function network (RBFN) and some extensions of it, as well as a number of training algorithms including the stochastic gradient (SG) algorithm that we have proposed earlier. We also propose a modified structure based on the Gaussian-mixture model and a learning algorithm for the modified network. Experimental results indicate that the radial basis function network and its extensions can be very useful in restoring images degraded by nonlinear distortion and noise.     

A model selection method for nonlinear system identification based FMRI effective connectivity analysis

In this paper a model selection algorithm for a nonlinear system identification method is proposed to study functional magnetic resonance imaging (fMRI) effective connectivity. Unlike most other methods, this method does not need a pre-defined structure/model for effective connectivity analysis. Instead, it relies on selecting significant nonlinear or linear covariates for the differential equations to describe the mapping relationship between brain output (fMRI response) and input (experiment design). These covariates, as well as their coefficients, are estimated based on a least angle regression (LARS) method. In the implementation of the LARS method, Akaike's information criterion corrected (AICc) algorithm and the leave-one-out (LOO) cross-validation method were employed and compared for model selection. Simulation comparison between the dynamic causal model (DCM), nonlinear identification method, and model selection method for modelling the single-input-single-output (SISO) and multiple-input multiple-output (MIMO) systems were conducted. Results show that the LARS model selection method is faster than DCM and achieves a compact and economic nonlinear model simultaneously. To verify the efficacy of the proposed approach, an analysis of the dorsal and ventral visual pathway networks was carried out based on three real datasets. The results show that LARS can be used for model selection in an fMRI effective connectivity study with phase-encoded, standard block, and random block designs. It is also shown that the LOO cross-validation method for nonlinear model selection has less residual sum squares than the AICc algorithm for the study.     

Decomposing bent functions

In a recent paper , it was shown that the restrictions of bent functions to subspaces of codimension 1 and 2 are highly nonlinear. Here, we present an extensive study of the restrictions of bent functions to affine subspaces. We propose several methods which are mainly based on properties of the derivatives and of the dual of a given bent function. We solve an open problem due to Hou . We especially describe the connection, for a bent function, between the Fourier spectra of its restrictions and the decompositions of its dual. Most notably, we show that the Fourier spectra of the restrictions of a bent function to the subspaces of codimension 2 can be explicitly derived from the Hamming weights of the second derivatives of the dual function. The last part of the paper is devoted to some infinite classes of bent functions which cannot be decomposed into four bent functions.     

Optimal preventive maintenance scheduling in semiconductor manufacturing

Preventive maintenance (PM) scheduling is a very challenging task in semiconductor manufacturing due to the complexity of highly integrated fab tools and systems, the interdependence between PM tasks, and the balancing of work-in-process (WIP) with demand/throughput requirements. In this paper, we propose a two-level hierarchical modeling framework. At the higher level is a model for long-term planning, and at the lower level is a model for short-term PM scheduling. Solving the lower level problem is the focus of this paper. We develop mixed-integer programming (MIP) models for scheduling all due PM tasks for a group of tools, over a planning horizon. Interdependence among different PM tasks, production planning data such as projected WIP levels, manpower constraints, and associated PM time windows and costs, are incorporated in the model. Results of a simulation study comparing the performance of the model-based PM schedule with that of a baseline reference schedule are also presented.     

自适应神经模糊推理系统（ANFIS）及其仿真

以一个非线性模型为研究对象,通过对自适应神经模糊推理系统（ANFIS）建模机理的研究建立了非线性实例模糊模型,借助MATLAB中ANFIS的功能讨论隶属度函数的数目、ANFIS输出、训练误差等对自适应神经模糊推理系统（ANFIS）的影响,取得了良好的效果。结果表明利用ANFIS进行非线性系统建模和辨识是可行的,其辩识精度很高。     

Linear and nonlinear ARMA model parameter estimation using anartificial neural network

Addresses parametric system identification of linear and nonlinear dynamic systems by analysis of the input and output signals. Specifically, the authors investigate the relationship between estimation of the system using a feedforward neural network model and estimation of the system by use of linear and nonlinear autoregressive moving-average (ARMA) models. By utilizing a neural network model incorporating a polynomial activation function, the authors show the equivalence of the artificial neural network to the linear and nonlinear ARMA models. They compare the parameterization of the estimated system using the neural network and ARMA approaches by utilizing data generated by means of computer simulations. Specifically, the authors show that the parameters of a simulated ARMA system can be obtained from the neural network analysis of the simulated data or by conventional least squares ARMA analysis. The feasibility of applying neural networks with polynomial activation functions to the analysis of experimental data is explored by application to measurements of heart rate (HR) and instantaneous lung volume (ILV) fluctuations     

一种改进窗函数非线性组合法的旁瓣抑制性能研究

为解决在雷达脉冲压缩过程中通过海明窗、汉宁窗等加权方式降低旁瓣而引起主瓣展宽、分辨率下降的问题,提出了一种以常用窗函数为基窗函数的改进窗函数非线性组合法。首先,相较于基窗函数,基于窗函数加权脉冲压缩的原理,证明窗函数线性组合法在主瓣展宽不变时可获得更好的旁瓣抑制能力;并在此基础上进一步证明改进的窗函数非线性组合法具有更优的旁瓣抑制能力。然后,通过对改进的窗函数非线性组合法在三角窗、汉宁窗、海明窗、布莱克曼窗等作为基窗函数时的脉冲压缩性能仿真与分析,验证了该方法使用不同基窗函数均可在主瓣展宽不变、信噪比损失增加低于0.5 dB的前提下,将峰值旁瓣相比凯塞窗降低4 dB以上、相比几种常用窗函数降低7 dB以上。最后,通过多目标的仿真实验验证,相比于不同的基窗函数,改进窗函数非线性组合法具有更优的弱小目标检测能力。     

基于神经网络方法的半导体生产工艺优化

以提高生产成品率为目标,利用神经网络的非线性和容错性,对半导体芯片生产过程进行了分析和优化,具体内容如下：（１）使用神经网络方法建立模型,确定生产线上工艺参数和成品率之间的映射关系,构造以工艺参数为输入,成品率为输出的多维函数曲面．（２）对上述多维函数曲面进行搜索,搜索成品率最高的最优点,以该最优点的工艺参数值为依据确定工艺参数的规范值．（３）对工艺参数规范进行优化,在实际生产工艺中反复实践,直至达到提高成品率的目的．生产实践证明,神经网络的分析结果是合理的．根据神经网络分析提出的优化建议,有效地提高了工     

基于模糊函数的双基地雷达信号检测分析

在双基地雷达系统中,基于纽曼-皮尔逊准则的似然比检测不但与发射信号的波形有关,而且还与双基地雷达系统的几何布站及目标的空间位置有关.首先以单基地雷达信号检测过程为基础,推导了双基地雷达系统的信号检测统计量模型,该模型是双基地雷达系统模糊函数的理论基础.然后给出了双基地雷达系统目标时延和多普勒频移参量的数学形式,与单基地...     

基于FastICA的混合音频信号盲分离

独立成分分析（ICA）作为一种有效的盲源分离技术已成为信号处理领域的热点,它以非高斯源信号为研究对象,在统计独立的假设下,对多路观测到的混合信号进行盲信号分离。为了提高算法的收敛速度和稳态精度,介绍了独立成分分析的基本原理,以及利用FastICA算法进行信号分离的理论依据,引入了改进的非线性函数,运用Matlab进行仿真比较3种非线性函数下的分离性能和改进的非线性函数在不同θ下的分离性能,结果表明在综合因素的考虑下,该改进函数在实现混合音频信号盲分离方面比改进前更有效。     

Blick hinter die Kulissen der Ausgleichsenergie — Technisch betrachtet

This paper describes the IT-system for the technical clearing of balancing energy based on the implemented balance group model. The model is adapted to the specific needs of the balancing energy clearing center for the control area of Austrian Power Grid (APG). The required tasks were realized with a complex IT-system. The system is fully web based and administrates the organization and data management of the different balance groups and their relationship among them. Based on this information it determines the balancing energy and the related fees for each individual balancing group.     

基于深度动态贝叶斯网络的服务功能链故障诊断算法

针对5G端到端网络切片场景下底层物理节点出现故障会导致运行在其上的多条服务功能链出现性能异常的问题,该文提出一种基于深度动态贝叶斯网络(DDBN)的服务功能链故障诊断算法。首先根据网络虚拟化环境下故障的多层传播关系,构建故障与症状的依赖图模型,并采用在物理节点监测其上多个虚拟网络功能相关性能数据的方式收集症状。其次,考虑到基于软件定义网络(SDN)和网络功能虚拟化(NFV)的架构下网络症状观测数据的多样性以及物理节点和虚拟网络功能的空间相关性,引入深度信念网络对观测数据特征进行提取,使用加入动量项的自适应学习率算法对模型进行微调以加快收敛速度。最后,利用故障传播的时间相关性,引入动态贝叶斯网络对故障根源进行实时诊断。仿真结果表明,该算法能够有效地诊断故障根源且具有良好的诊断准确度。     

融合非线性加速退化模型与失效率模型的产品寿命预测方法

针对高可靠性产品寿命数据少、获取成本高的问题,基于充分利用产品在研制、加速试验等不同环境下的退化数据、失效数据等可靠性数据的思想,提出了一种融合非线性加速退化模型和失效率模型的产品寿命预测方法.首先,根据退化数据对非线性退化过程进行分析,估计退化过程的参数;然后,根据加速退化数据及相应的加速退化模型估计加速退化模型的参数,从而得到退化参数与应力之间的关系.进一步,利用比例风险模型融合产品的寿命数据和未失效截尾数据,并基于此计算产品的可靠度函数、预测产品的寿命.实例应用验证了所提方法的有效性,同时说明了所提方法的应用价值.     

Modeling of Mammalian Skeletal Muscle

A model of mammalian skeletal muscle is developed from experimental results from a rat gracilis anticus muscle at 17.5°C. The approach consists of factoring pertinent variables of muscle contraction (length, force, velocity, and duration of stimulation) into a series of physically realizable functions. The analysis indicates that for lengths less than 120 percent of rest length, mammalian skeletal muscle can be modeled as a nonlinear force generator, a function of length and time, bridged by a nonlinear viscous-like element, a function of time, length, and velocity, in series with a nonlinear elastic element, a function of length. A close correlation exists between the results obtained during an analog computer simulation of the model and those from a typical rat gracilis anticus muscle.