通用技术与设备

Y2000-62203-3446 0015881非线性辨识(含6篇论文)=FA18:Nonlinear identifi-cation[会,英]//1999 IEEE Proceedings of AmericanControl Conference,Vol.5 of 6.—3446～3473(NiD)本部分收录6篇论文,内容包括控制实验的最优设计,维纳模型闭环辨识的简接方法,内联系统非线性     

非线性信道的LMS Newton均衡算法的研究

本文使用Hammerstein模型和维纳模型代替Volterra级数模型来模拟非线性结构以降低运算复杂度,提出了一个由Hammerstein模型和维纳模型构建成的非线性信道传输系统的模型,由此模型给出并推导出了基于该信道模型的NCLMS算法、改进1型NCLMS Newton算法和改进2型NCLMS Newton算法.仿...     

哈默斯坦非线性时变系统的加权学习辨识方法

针对有限区间哈默斯坦(Hammerstein)非线性时变系统,该文提出一种加权迭代学习算法用以估计系统时变参数。首先将Hammerstein系统输入非线性部分进行多项式展开,采用迭代学习最小二乘算法辨识系统的时变参数。为了防止数据饱和,采用带遗忘因子的迭代学习最小二乘算法,进而引入权矩阵,采用加权迭代学习最小二乘算法改进系统跟踪误差,以提高辨识精度。该文分别给出3种算法的推导过程并进行仿真验证。结果表明,与迭代学习最小二乘算法和带遗忘因子迭代学习最小二乘算法相比,加权迭代学习最小二乘算法具有辨识精度高、跟踪误差小以及迭代次数少等优点。     

一类Wiener非线性时变系统的迭代学习辨识

针对Wiener非线性时变系统的参数辨识问题,该文提出一种基于重复轴的迭代学习算法来实现对时变甚至突变参数的估计.文中将维纳系统输出非线性部分的反函数进行多项式展开,进而构造了回归模型,未知参数及中间变量用其估计替代,分别给出了采用迭代学习梯度算法和迭代学习最小二乘算法实现时变参数辨识的方法.仿真结果表明,与带遗忘因子的递推算法和迭代学习梯度算法相比,迭代学习最小二乘算法更具有参数估计收敛速度快,辨识精度高,系统输出误差小等优势,验证了所提学习算法的有效性.     

基于操作轨迹模型的非线性预测控制算法

针对工业过程的非线性,本文首先提出了一种操作轨迹非线性模型的辨识方法:根据调度变量的操作轨迹,选取若干个典型工作点;在各个典型工作点,辨识各自的线性模型;根据测试数据以及过渡数据,得到全局插值非线性模型.在此基础上,本文进一步提出一种基于操作轨迹模型的非线性预测控制算法,并采用多步线性化方法进行问题求解.由于仅需要在典型工作点上进行测试,降低了全局建模的辨识成本,而且控制品质好,仿真结果表明了该算法的有效性.     

非均匀周期采样多率系统的一种辨识方法

本文利用提升技术,推导了非均匀采样多率系统的提升状态空间模型.对于状态可测量的多率系统,利用最小二乘原理,给出了模型参数矩阵辨识方法;对于状态不可测的未知参数多率系统,利用递阶辨识原理,在考虑提升模型的因果约束下,将提升系统分解为子系统进行辨识,形成了状态空间模型递阶辨识方法.仿真例子表明,本文提出的递阶辨识方法是有效的.     

非线性信道的均衡算法研究

本文使用Hammerstein模型和维纳模型代替Volterra级数模型来模拟非线性结构以降低运算复杂度,提出了一个由Hammerstein模型和维纳模型构建成的非线性信道传输系统的模型.基于该系统模型,分别提出并推导了三种非线性信道的均衡算法:NCRLS算法、NCKalman算法和NCRPEM算法,并对这三种新算法的性能进行了比较.仿真结果表明,在剩余均方误差方面三种算法中NCKalman算法最小,NCRPEM算法次之,NCRLS算法较差;在收敛速度方面NCRPEM算法收敛最快,NCRLS算法次之,NCKalman算法较差.     

基于递阶原理的非均匀采样非线性系统的模糊辨识

针对非均匀多采样率非线性系统的建模问题,提出了基于递阶原理的模糊辨识方法.首先,分析了非线性系统在输入信号非均匀周期刷新,输出信号周期采样的情况下,非线性系统可以通过提升技术,利用多个局部线性模型加权组合的模糊模型来描述.在此基础上,利用GK模糊聚类确定模糊模型前件结构,利用基于递阶原理的递推最小二乘辨识算法辨识模糊模型后件参数.同时,通过鞅定理对辨识算法的收敛性进行了研究.最后,通过仿真实例证明了本文方法的有效性.     

一类 Wiener 非线性时变系统的迭代学习辨识

针对Wiener非线性时变系统的参数辨识问题,该文提出一种基于重复轴的迭代学习算法来实现对时变甚至突变参数的估计。文中将维纳系统输出非线性部分的反函数进行多项式展开,进而构造了回归模型,未知参数及中间变量用其估计替代,分别给出了采用迭代学习梯度算法和迭代学习最小二乘算法实现时变参数辨识的方法。仿真结果表明,与带遗忘因子的递推算法和迭代学习梯度算法相比,迭代学习最小二乘算法更具有参数估计收敛速度快,辨识精度高,系统输出误差小等优势,验证了所提学习算法的有效性。     

归一化峰度及其在弱非线性系统盲辨识中的应用

本文主要研究信号的归一化峰度及其在弱非线性系统辨识中的应用策略问题.首先简要介绍了几类常见的无记忆/有记忆非线性模型及其表示方法;给出了信号的归一化峰度定义及重要性质;在此基础上,分别针对非线性系统的记忆效应和非线性阶数对系统输出信号归一化峰度的影响进行了理论推导和仿真分析,揭示了该参数随系统特性的变化规律,表明归一化峰度具备精确辨识弱非线性系统的潜力.最后,针对SFDR(无杂散动态范围)高达85dBFS(dB Full Scale)的弱非线性系统,本文提出了一种分步辨识的方法,并结合所提出的方法阐明了此规律对于弱非线性系统盲辨识和失真补偿的潜在应用价值及其精度优势.     

Use of trellis diagrams in state estimation in blocks for nonlinear discrete dynamic systems with aKth-order memory

A new suboptimum estimation scheme is proposed for nonlinear discrete dynamic systems with aKth-order memory. These systems are first represented by trellis diagrams, and then states are estimated by the Viterbi algorithm of information theory. The state and observation models of the proposed scheme can be nonlinear functions of the disturbance noise, observation noise, and present and past discrete values of the state, whereas the models of the classical estimation algorithms, such as the extended Kaiman filter, must be linear functions of the disturbance noise and observation noise. States are estimated in blocks, which results in an estimation scheme whose implementation requries a constant memory.     

基于双线性混合模型的高光谱图像非线性光谱解混

高光谱遥感图像的非线性光谱解混能弥补线性方法难以解释复杂场景中非线性混合效应的不足, 而双线性混合模型及算法是其研究的热点.提出了一种基于双线性混合模型几何特性的光谱解混算法.通过将模型中的非线性混合项表示为一个融合了共同非线性效应的额外端点的线性贡献, 使复杂的双线性混合模型求解转化为简单的线性解混问题.然后结合传统的线性解混算法直接迭代估计正确的丰度.模拟和真实遥感图像数据的实验结果表明, 与其它相关解混方法相比, 该算法能较好地克服共线性效应以及拟合优化过多参数对双线性混合模型求解造成的不利影响, 同时提高了解混的精度和速度.     

Wiener功率放大器的分离预失真方法

Hammerstein(以下简称H)系统能实现对Wiener功率放大器的预失真,文章从其结构特点出发,提出了分离预失真方法,将放大器中的记忆和非线性因素分离,并独立地得到各自的预失真模块,从而组成能对WienerPA进行线性化的H系统,克服了以往在对H系统各参数的辨识中,由于没有排除记忆和非线性因素的相互干扰而导致的算法较为复杂,收敛速度缓慢,收敛精度不高的问题。仿真表明,该方法在收敛速度和收敛精度上都有明显提高。     

Recursive Extended Least Squares Parameter Estimation for Wiener Nonlinear Systems with Moving Average Noises

Many control algorithms are based on the mathematical models of dynamic systems. System identification is used to determine the structures and parameters of dynamic systems. Some identification algorithms (e.g., the least squares algorithm) can be applied to estimate the parameters of linear regressive systems or linear-parameter systems with white noise disturbances. This paper derives two recursive extended least squares parameter estimation algorithms for Wiener nonlinear systems with moving average noises based on over-parameterization models. The simulation results indicate that the proposed algorithms are effective.     

一种基于折线逼近的对数似然比简化算法

针对16QAM信号的对数似然比计算,该文提出一种新颖的折线逼近简化算法,采用分段的折线逼近16QAM信号的对数似然比曲线,进而使用简单的线性运算替代原标准LLR算法中复杂的非线性运算,仿真结果表明,该算法可以理想地逼近标准LLR算法的计算结果,在BITCM系统中应用不会带来系统的性能折损。     

Maximum-likelihood detection of nonlinearly distorted multicarrier symbols by iterative decoding

This paper proposes a new method for decoding multicarrier symbols with severe nonlinear distortion. The first part evaluates mutual information expressions for practical nonlinear models and shows the performance bounds for commonly used receiver structures. Then, we derive the maximum-likelihood (ML) sequence estimator, which unfortunately has an exponential complexity due to the nonlinear distortion. This extremely large complexity can be reduced with a simple algorithm that iteratively estimates the nonlinear distortion, thereby reducing the exponential ML to the standard ML without nonlinear distortion. The proposed method can be used to reduce the peak-to-average power ratio of multicarrier signals by clipping the transmit sequence. It can also be used to correct any nonlinear distortion present in transmitter/receiver amplifiers that are operating close to saturation.     

Global identifiability of nonlinear models of biological systems

A prerequisite for well-posedness of parameter estimation of biological and physiological systems is a priori global identifiability, a property which concerns uniqueness of the solution for the unknown model parameters. Assessing a priori global identifiability is particularly difficult for nonlinear dynamic models. Various approaches have been proposed in the literature but no solution exists in the general case. In this paper, we present a new algorithm for testing global identifiability of nonlinear dynamic models, based on differential algebra. The characteristic set associated to the dynamic equations is calculated in an efficient way and computer algebra techniques are used to solve the resulting set of nonlinear algebraic equations. The algorithm is capable of handling many features arising in biological system models, including zero initial conditions and time-varying parameters. Examples of usage of the algorithm for analyzing a priori global identifiability of nonlinear models of biological and physiological systems are presented.     

基于并行BP神经网络的近红外光断层图像重建方法基础研究

提出一种基于并行BP神经网络的近红外光断层成像(Near-infrared optical tomography,NIR OT)图像重建算法,利用BP神经网络来表征生物组织内部光学参数的空间分布和边界光强之间的非线性映射关系.该方法将一个复杂的模型分解成简单的模型分别建立并行的神经网络.利用Femlab软件完成基于有限元的稳态扩散方程的两个简单模型的正向问题求解,根据提出的平均优化散射系数和正向问题训练的大量数据集合,建立并训练并行神经网络,通过对两个网络结果的分析,实现快速获得更复杂模型的光学参数的重构.算法能够快速识别特异组织的位置和准确反映热疗过程中生物组织的优化散射系数的变化趋势.     

MATLAB环境下非线性控制系统的建模与仿真

MATLAB软件在许多科学领域中成为计算机辅助设计、算法研究和应用开发的基本工具,在MATLAB/Simulink中对线性定常系统或者简单的非线性控制系统的建模与仿真比较简单方便,但对复杂非线性控制系统的建模与仿真实现困难。提出一种在MATLAB环境下利用m函数实现非线性控制系统的建模与仿真方法,该方法简单直观,维护性较好,具有可移植性。对复杂的非线性控制系统的建模与仿真,该方法可以明显提高仿真的效率。仿真实例验证了该方法的有效性和可行性。     

A Novel Nonlinear Joint Transmitter-Receiver Processing Algorithm for the Downlink of Multiuser MIMO Systems

This paper focuses on signal processing algorithms for the downlink of multiuser multiple-input multiple-output (MIMO) systems with multiple-antenna mobiles. A novel nonlinear joint transmitter-receiver processing algorithm is proposed based on the zero-forcing (ZF) criterion. In this algorithm, nonlinear Tomlinson-Harashima precoding (THP) is applied at the base station, whereas linear receiver processing and modulo operation are applied at each mobile. It is first shown that the proposed algorithm effectively decomposes the multiuser MIMO channel into parallel independent single-user MIMO channels, and then, the performance of each mobile can be separately optimized. Subsequently, closed-form expressions for the transmitter and receiver processing matrices are derived to optimize the asymptotic bit error rate (BER) of each mobile. When used on the downlink of multiuser MIMO systems with multiple-antenna mobiles, this algorithm achieves significantly better performance than the ZFcriterion-based nonlinear preprocessing algorithm designed for the multiuser MIMO systems with single-antenna mobiles, because it effectively utilizes the processing capabilities of the mobiles. Moreover, the proposed algorithm achieves a much higher sum capacity at a high signal-to-noise ratio (SNR) than the known block diagonalization technique due to the effective application of the nonlinear preprocessing at the transmitter. When the proposed algorithm is applied, it is found that better system performance can be achieved by suitably ordering the channel matrices of different mobiles, and a combined optimal diversity and best-first (CODBF) ordering method is proposed to perform the ordering. Simulation is used to show the advantages of the proposed algorithm and the CODBF ordering method.