模型误差界的频域辨识方法分析

根据型误差界的确定是系统辨识一个和组成部分。本文应用 截尾傅氏变换和Ｈ∞辨 识理论，对模型误差界的常规频域辨识方法进行了分析，给出了传递函数的频率响误差界和总体误差界，对ＢＡＲＹＡＲＤＴＦＦＵＷ Ｒ ＨＩＦＡ界估计方法做了修正。     

基于小波变换估计频域模型误差界

讨论鲁棒辨识问题，基于离散小波变换，提出分段频带逼近，估计频域模型界。首先介绍离散小波变换，然后给出分段频带逼近算法，仿真结果验证所提方法。     

递推鲁棒频域辨识算法误差界计算的改进*

文［１］研究了带未建模动态系统的频域辨识问题，辨识的结果为传递函数在单位圆上的有限个点估计及其误差界。本文指出文［１］定理的错误，并对其进行了更正，数字仿真表明，更正的误差界比文［１］中的误差界更精确。     

基于频域分析的高阶加时滞模型辨识方法

工业生产中含有时滞环节的高阶过程对象,由于控制器结构复杂,采用直接设计控制器和传统模型降阶的方法实现起来比较困难,同时噪声会对控制器产生干扰,可能导致这些方法得不到满意的结果。针对这一问题,提出了一种高阶加时滞模型的频域辨识方法,通过对控制回路中过程对象的输入和输出信号进行频谱分析和拉普拉斯变换,得出过程对象在重要频率段上的频率响应特性,采用最小二乘法从幅值和相位两方面拟合成一个一阶或二阶加纯滞后传递函数,最终获得对象模型。通过MATLAB仿真结果验证了所提方法的有效性和精确性。     

多输入多输出变量带误差模型的最坏情况频域辨识

本文将单输入单输出(SISO)变量带误差(EIV)模型的频域最坏情况辨识方法推广应用于多输入多输出 (MIMO)情况. 类似于SISO情况, 多输入多输出变量带误差(MIMO EIV)模型的辨识模型集合由估计的系统名义模型及 其最坏情况误差界描述. 所估计的系统名义模型表征为正规右图符号, 其最坏情况误差界具有可能的更少保守性, 可利 用EIV 模型的先验信息和后验信息由v-gap度量量化得到. 因此, 这种模型集合非常适合于后期利用Vinnicombe提出 的H1回路成形法设计鲁棒控制器. 最后, 利用一数值仿真实例验证所提出辨识方法的有效性.     

有界误差模型的一种结构辨识方法

针对具有未知但有界(UBB)误差的线性回归模型辨识问题,提出了一种新的鲁棒结构选择方法.该方法以重复递推椭球外界算法所得椭球轴信息阵的行列式相对值最大作为模型定阶准则.不同于以往对噪声独立性、常方差或鞅差特性的假设,该方法假设噪声是渐近独立的.文中证明了该方法的强相容性.     

流水线ADC组合误差分析与辨识模型设计

本文详细研究了影响流水线ADC动态性能的主要误差来源及其不同组合和耦合方式的影响结果与权重,获得了流水线拼接输出数字码中的杂谐波分量与组合误差之间的对应关系。在此基础上,设计了用于表征流水线ADC非线性特性的分立式维纳模型。对14位三级流水的AD6645芯片进行测试,结果表明流水线ADC器件的实际输入输出特征与该辨识模型的性质吻合。得到的误差分析结论与经过检验的模型结构可为数字后补偿方法中非线性逆模型的选取提供参考依据。     

油田射孔器材检测装置加热系统的频域模型辨识

油田射孔器材检测装置的加热系统是多介质的复合传热过程 ,很难建立精确数学模型 ,本文通过实测系统的输入信号和输出响应 ,利用理论分析、数据优化及数值的积分法 ,辨识出该系统传递函数形式的近似数学模型。该模型在仿真研究和实际应用中取得了良好的效果。     

基于Hopfield网络的系统参数频域辨识方法

本文综合利用频谱分析技术和 Hopfield网络的神经计算能力 ,提出一种基于 Hopfield网络的系统参数频域辨识方法 ,并进行了仿真验算 ,结果表明了方法实现的可行性及在克服噪声影响方面的有效性。     

On Exact Error Bounds for View-Dependent Simplification

In this article we present an analytical closed‐form expression to ensure exact error bounds for view‐dependent simplification which is of importance for several algorithms. The present work contains proofs and solutions for the general 2D case and particular 3D cases. Most preceding works rely on coarse heuristics, that might fail and/or restrict movements or object representations. We introduce the notion of validity regions as the complete set of possible simplifications respecting a given error bound between the object and its simplification. The approach handles arbitrary polygonal viewcells which allow for free movement in the interior. We show how to compute these regions for mesh points and faces. Since the validity region of a face accounts for all its points, properties like silhouette preservation and textures are gracefully handled. This is not the case if the error is controlled only at the face's vertices or edges.     

LTV Model Reduction With Upper Error Bounds

A new approach for computing upper error bounds for reduced-order models of linear time-varying systems is presented. It is based on a transformation technique of the Hankel singular values using positive-real, odd incremented functions. By applying such time-varying functions, the singular values to be removed can be forced to become equal and constant, so that they can be reduced. Two variations of this method are proposed: one for finite-time horizons and the other for infinite-time problems including periodic systems.      

On Upper Bounds for the Decoding Error Probability of Convolutional Codes

A new approach to upper bounding the decoding error probability of convolutional codes is proposed. Its idea is, instead of evaluating the individual contribution of each fundamental path, to compare it with contribution of another (lighter) fundamental path. This allows us to (1) take into account the dependence between different fundamental paths based on the code tree structure; (2) represent the decoding error probability through the contribution of the first fundamental paths and a correction factor; (3) get much more accurate estimates.     

On the Assumption of Spherical Symmetry and Sparseness for the Frequency-Domain Speech Model

A new independent component analysis (ICA) formulation called independent vector analysis (IVA) was proposed in order to solve the permutation problem in convolutive blind source separation (BSS). Instead of running ICA in each frequency bin separately and correcting the disorder with an additional algorithmic scheme afterwards, IVA exploited the dependency among the frequency components of a source and dealt with them as a multivariate source by modeling it with sparse and spherically, or radially, symmetric joint probability density functions (pdfs). In this paper, we compare the speech separation performances of IVA by using a group of l^p-norm-invariant sparse pdfs where the value of and the sparseness can be controlled. Also, we derive an IVA algorithm from a nonparametric perspective with the constraint of spherical symmetry and high dimensionality. Simulation results confirm the efficiency of assuming sparseness and spherical symmetry for the speech model in the frequency domain.     

On Long Time Error Bounds for the Wave Equation on Second Order Form

Temporal error bounds for the wave equation expressed on second order form are investigated. We show that, with appropriate choices of boundary conditions, the time and space derivatives of the error are bounded even for long times. No long time bound on the error itself is obtained, although numerical experiments indicate that a bound exists.     

Experimental Estimation of Model Error Bounds Based on Modified Stochastic Approximation

A modification of the estimation algorithm stochastic approximation is presented. With assumptions to the statistical distribution of the training data it becomes possible, to estimate not only the mean value but also well directed deviating values of the data distribution. Thus, detailed error models can be identified by means of parameter-linear formulation of the new algorithm. By definition of suitable probabilities, these parametric error models are estimating soft error bounds. That way, an experimental identification method is provided that is able to support a robust controller design. The method was applied at an industrial robot, which is controlled by feedback linearisation. Based on a dynamic model realised by a neural network, the presented approach is utilised for the robust design of the stabilising decentral controllers.     

Computable Error Bounds for Coefficients Perturbation Methods

The coefficients perturbation method (CPM) is a numerical technique for solving ordinary differential equations (ODE) associated with initial or boundary conditions. The basic principle of CPM is to find the exact solution of an approximation problem obtained from the original one by perturbing the coefficients of the ODE, as well as the conditions associated to it. In this paper we shall develop formulae for calculating tight error bounds for CPM when this technique is applied to second order linear ODEs. Unlike results reported in the literature, ours do not require any a priori information concerning the exact error function or its derivative. The results of this paper apply in particular to the Tau Method and to any approximation procedure equivalent to it. The convergence of the derived bounds is also discussed, and illustrated numerically. Received April 5, 2002; revised June 11, 2002 Published online: December 12, 2002