UD分解与偏差补偿结合用于变量带误差模型辨识

本文提出一种基于UD(upper-diagonal)分解与偏差补偿结合的辨识方法,用于变量带误差(errors-in-variables,EIV)模型辨识.考虑单输入单输出(single input and single output,SISO)线性动态系统,当输入和输出含有零均值、方差未知的高斯测量白噪声时,该类系统的模型参数估计是一种典型的EIV模型辨识问题.为了获得这种EIV模型参数的无偏估计,本文先推导出最小二乘模型参数估计偏差量与输入输出噪声方差以及最小二乘损失函数与输入输出噪声方差的关系,然后采用UD分解方法递推获得模型参数估计值,再利用输入输出噪声方差估计值补偿模型参数估计偏差,以此获得模型参数的无偏估计.本文还讨论了算法实现过程中遇到的一些问题及修补方法,并通过仿真例验证了所提辨识方法的有效性.     

Monte Carlo evaluation of methods for pulse transfer function estimation

This paper describes the results of a Monte Carlo evaluation made of the methods proposed in current literature for the estimation of the pulse transfer function of a linear, time-invariant dynamic system with feedback. Considered are two basic methods for estimating the coefficients of a pulse transfer function, given only the normal operating input and output of the system obscured by noise and over a limited period of time. The most commonly proposed method is a linear method in which a set of simultaneous linear equations is formed from the sampled data and the coefficients obtained by a matrix inversion. The other method is an eigenvector method proposed by Levin which uses the eigenvector associated with the smallest eigenvalue of a matrix formed from the sampled data. This paper presents a set of examples designed to compare linear and eigenvector estimation methods and to verify experimentally the theoretical results and approximations given by Levin. The comparison shows that the eigenvector method generally gives estimates with equal or smaller rms errorsqrt{Variance+(Bias)^2}than the linear method. The eigenvector estimates had bias magnitudes which were consistently less than their standard deviations; the linear estimates did not, and thus had rms errors which often consisted largely of the bias. The approximate covariance matrix given by Levin for the coefficients estimated with the eigenvector method is found to be reasonably accurate.     

Linear system identification using numerical computation of Laplace transforms

A method for system identification using sampled values of the initial transient step or impulse response is described. A polynomial fit of the sampled values is made using Lagrange interpolation and the Laplace transform of the output observed is determined. Then the coefficients of the numerator and denominator polynomials of the system transfer function are determined by minimizing the square of the difference between the observed and calculated values of the Laplace transform of the output variable at a number of discrete points. This process is considerably simplified by the use of tables of coefficients for the numerical calculation of Laplace transforms.     

Estimation of the autoregressive parameters of a mixed autoregressive moving-average time series

The problem of estimating the autoregressive parameters of a mixed autoregressive moving-average (ARMA) time series (of known order) using the output data alone is treated. This problem is equivalent to the estimation of the denominator terms of the scalar transfer function of a stationary, linear discrete time system excited by an unobserved unenrrelated sequence input by employing only the observations of the scalar output. The solution of this problem solves the problem of the identification of the dynamics of a white-noise excited continuous-time linear stationary system using sampled data. The latter problem was suggested by Bartlett in 1946. The problem treated here has appeared before in the engineering literature. The earlier treatment yielded biased parameter estimates. An asymptotically unbiased estimator of the autoregressive parameters is obtained as the solution of a modified set of Yule-Walker equations. The asymptotic estimator covariance matrix behaves like a least-squares parameter estimate of an observation set with unknown error covariances. The estimators are also shown to be unbiased in the presence of additive independent observation noise of arbitrary finite correlation time. An example illustrates the performance of the estimating procedures.     

Time delay estimation for MIMO dynamical systems – With time-frequency domain analysis

In MIMO dynamical systems, the time delay estimation (TDE) problem between each output and input is often confounded due to the dynamic interaction between the inputs and the outputs. In this paper, analyses are given in the time, frequency and time-frequency domains, from which a novel TDE method using combined continuous wavelet transform (CWT) and cross correlation is conceived. In the proposed method, a series of time delays over scales (frequencies) are calculated and an unbiased estimation is deduced from them, by calculating and handling the cross correlation between the CWT coefficients of system input and output data. The TDE method in closed loop case is also studied. Numerical examples with simulation and experimental data verify the feasibility and effectiveness of the proposed method.     

Recursive identification of errors-in-variables Wiener systems

This paper considers the recursive identification of errors-in-variables (EIV) Wiener systems composed of a linear dynamic system followed by a static nonlinearity. Both the system input and output are observed with additive noises being ARMA processes with unknown coefficients. By a stochastic approximation incorporated with the deconvolution kernel functions, the recursive algorithms are proposed for estimating the coefficients of the linear subsystem and for the values of the nonlinear function. All the estimates are proved to converge to the true values with probability one. A simulation example is given to verify the theoretical analysis.     

一类有输入噪声扰动的逆系统无偏参数辨识算法研究

对逆系统建模时,原系统的输出作为逆系统参数辨识时的输入.由于原系统输出存在测量噪声,且噪声方差未知,采用普通最小二乘法辨识,无法得到逆系统参数的一致无偏估计.为此,本文研究了一种有输入扰动的的逆系统无偏参数辨识算法,该算法先通过小波变换估计输入信号噪声的方差,再由估计得到的方差,通过偏差消除的递推最小_乘法,对逆系统的参数进行无偏辨识.该算法降低了对输入辨识信号为白噪声的要求,具有较强的实用性.由于采用递推运算,该算法也可以用于逆系统参数的在线辨识.最后,通过实验验证了该算法的有效性.     

Unbiased parameter estimation by means of autocorrelation functions

A new method for obtaining an on-line unbiased estimate of the parameter vector of linear systems in the presence of additive uncorrelated output noise is described. The method, which is based on the equation error, is developed for continuous variables and then formulated both for continuous and discrete representations. Instead of the conventional least squares approach, the autocorrelation function of the error serves as the cost function to be minimized. It is shown that for comparatively wide-band noise, this cost function yields unique and practically unbiased estimates. On-line identification from system input and output is feasible. Bias suppression as a function of the time-delay in the cost function is demonstrated by numerical examples. Convergence of the estimates is demonstrated by means of simulated continuous and discrete examples.     

基于瞬时频率响应函数的间歇过程时段划分

多时段特性是间歇过程的本质特性之一,对间歇过程实现有效的时段划分是故障监测的基础。传统的时段划分方法大多针对过程的输入输出数据,对输入输出数据突变较为敏感。本文提出一种基于瞬时频率响应函数的间歇过程时段划分方法,该方法基于系统的瞬时动态特性,用瞬时频率响应函数替代输入输出数据进行时段划分,利用小波变换估计系统的瞬时频率响应函数进行核主元分析降维,通过模糊C均值聚类对降维后频率响应函数进行聚类划分时段。实验结果表明,本文所提出的方法能够实现对间歇过程的时段划分,并具有较高的鲁棒性。     

Optimal smoothing of non-linear dynamic systems via Monte Carlo Markov chains

We consider the smoothing problem of estimating a sequence of state vectors given a nonlinear state space model with additive white Gaussian noise, and measurements of the system output. The system output may also be nonlinearly related to the system state. Often, obtaining the minimum variance state estimates conditioned on output data is not analytically intractable. To tackle this difficulty, a Markov chain Monte Carlo technique is presented. The proposal density for this method efficiently draws samples from the Laplace approximation of the posterior distribution of the state sequence given the measurement sequence. This proposal density is combined with the Metropolis-Hastings algorithm to generate realizations of the state sequence that converges to the proper posterior distribution. The minimum variance estimate and confidence intervals are approximated using these realizations. Simulations of a fed-batch bioreactor model are used to demonstrate that the proposed method can obtain significantly better estimates than the iterated Kalman-Bucy smoother.     

Errors-in-variables identification of dynamic systems excited by arbitrary non-white input

This work deals with the identification of dynamic systems from noisy input–output observations, where the noise-free input is not parameterized. The basic assumptions made are (1) the dynamic system can be modeled by a (discrete- or continuous-time) rational transfer function model, (2) the temporal input–output disturbances are mutually independent, identically distributed noises, and (3) the input power spectrum is non-white (not necessarily rational) and is modeled nonparametrically. The system identifiability is guaranteed by exploiting the non-white spectrum property of the noise-free input. A frequency domain identification strategy is developed to estimate consistently the plant model parameters and the input–output noise variances. The uncertainty bound of the estimates is calculated and compared to the Cramér–Rao lower bound. The efficiency of the proposed algorithm is illustrated on numerical examples.     

含间隙非线性的Wiener-Hammerstein系统复合补偿控制

针对含非对称间隙环节的Wiener-Hammerstein系统提出了一种新的由输出反馈和间隙动态逆补偿构成的复合控制方案.首先应用参数化分段线性表达式设计了未知参数的整体估计模型,可同时估计线性参数和间隙的特征参数,然后提出了一种新的误差有界的间隙动态逆模型,该模型可使得驱动信号能在间隙的不同线性段之间快速切换,在此基础上设计了鲁棒补偿控制律,同时对输入线性环节采用输出反馈控制构建了复合控制器,通过李亚普诺夫方法证明了闭环系统的稳定性.带减速器的单电机伺服系统模型的仿真结果表明该方法在跟踪精度良好的同时可使系统动态响应满足要求.     

基于遗传规划的哈默斯坦模型的辨识

该文基于遗传规划提出了一种辨识哈默斯坦模型的新方法。哈默斯坦模型由静态非线性模块和动态线性模块串联而成,因此系统辨识的目标是要找到非线性和线性模块的最优数学模型。该文通过遗传规划确定非线性模块的函数结构,并结合遗传算法确定模型的未知参数,适应度值的计算采用了最小信息量准则(A IC),以平衡模型的复杂度和精确度。该方法不需要对模型的先验知识有详细了解,就能达到较好的辨识效果,并且能够克服观测噪声的污染,获得参数的无偏估计。仿真结果说明了该方法的有效性。     

Identification of dynamic errors-in-variables models: Approaches based on two-dimensional ARMA modeling of the data

In this paper we propose a parametric and a non-parametric identification algorithm for dynamic errors-in-variables model. We show that the two-dimensional process composed of the input-output data admits a finite order ARMA representation. The non-parametric method uses the ARMA structure to compute a consistent estimate of the joint spectrum of the input and the output. A Frisch scheme is then employed to extract an estimate of the joint spectrum of the noise free input-output data, which in turn is used to estimate the transfer function of the system. The parametric method exploits the ARMA structure to give estimates of the system parameters. The performances of the algorithms are illustrated using the results obtained from a numerical simulation study.     

Bilinear system identification by Volterra kernels estimation

This note considers the identification of bilinear discrete-time dynamic systems from sequences of input and noise corrupted output measurements. In contrast to other approaches, the proposed algorithm is simple and does not require knowledge of the noise statistics. It is also shown that the obtained estimates are unbiased and consistent, which is not shown in the previous papers.     

Identification of linear systems using input-output cumulants

A new identification method is proposed for estimating the parameters of a discrete-time linear dynamic system excited by non-gaussian inputs using kth order (k > 2) cumulants of input and output signals contaminated by additive (possibly coloured) gaussian noise. The parameter estimates obtained by this method are proved to be consistent under weak conditions. This method has an on-line algorithm for computing the parameter estimates, just like the least-squares method. A simulation example is included to demonstrate the effectiveness of this method.     

Transfer functions for infinite-dimensional systems

In this paper, we study three definitions of the transfer function for an infinite-dimensional system. The first one defines the transfer function as the expression C(sI−A)⁻¹B+D. In the second definition, the transfer function is defined as the quotient of the Laplace transform of the output and input, with initial condition zero. In the third definition, we introduce the transfer function as the quotient of the input and output, when the input and output are exponentials. We show that these definitions always agree on the right-half plane bounded to the left by the growth bound of the underlying semigroup, but that they may differ elsewhere.     

Parametric identification of linear systems using Laplace transform

Proposed was a method for identification of the transfer function coefficients of the stationary one-input one-output linear system using the Laplace images of the input and output measurements. The measurement processes are considered over the given set of values of the transformation parameter. The a posteriori density of distribution of the vector of identified parameters was determined with regard for the autocovariance functions of the measurement errors. A generalization to the multiple-input multiple-output linear system with perturbations and measurements of the state coordinates was described.     

Observer design for discrete systems with unknown exogenous inputs

A design procedure is developed for determining optimal discrete observers for estimating system states and unknown exogenous system inputs. This procedure is based on augmenting a standard system observer with an input model. The augmented model is then transformed into the discrete z-domain to determine relevant input/output transfer function matrices. The transfer function matrices are used to develop transfer function relationships between unknown exogenous inputs and the observer estimate of these inputs. It is shown that the optimal observer gains can be determined by implementing the observer as a Fisher filter. An example of the procedure is demonstrated with a third-order point-mass tracking filter