多变量CARMA模型的结构辨识

本文提出了多变量受控的自回归滑动平均(CARMA)模型结构辨识的新方法.根据模型参 数的递推增广最小二乘法(RELS)估计,给出了确定模型的阶、子阶和时滞的F检验判决器, 且可得到节省参数模型.本方法推广和改进了Bokor和Kevizky的结构辨识方法, 数值模拟例子证明了所提出方法的有效性.     

基于偏差消除最小二乘估计和Durbin方法的两阶段ARMAX参数辨识

针对带有外生变量的自回归移动平均模型(Autoregressive moving average with exogenous variable, ARMAX)的参数辨识问题提出一种两阶段辨识方法. 首先通过偏差消除最小二乘方法辨识带有外生变量的自回归部分(Autoregressive part with exogenous variable, ARX),然后采用Durbin方法将移动平均部分(Moving average, MA)的参数辨识问题转换成一个长自回归模型(Long autoregressive, LAR)的参数辨识问题, 并利用MA与等价LAR的参数对应关系直接得到MA参数, 最后利用辨识出的MA参数计算出噪声方差. 与扩展最小二乘法的数值仿真比较验证了这种两阶段辨识方法的有效性.     

时间序列的AR模型自动辨识机

本文提出了基于自回归(AR)模型对时间序列统一建模的方法,同流行的基于自回归滑动平均(ARMA)模型对时间序列建模的方法相比,显著地减少了计算量,并且提出了 AR 模型自动辨识机,它可自动地决定时间序列所服从的AR 模型的阶和参数,因而有明显的实用价值,以太阳黑子数目的时间序列建模为例,说明了所提出的辨识的有效性。     

自回归滑动平均建模中观测噪声方差估计的新方法

目前的自回归滑动平均(ARMA)建模方法由于只利用了观测数据的高阶自协方差构建Yule-Walker方程,而没有利用观测数据的低阶自协方差信息,导致观测噪声方差的估计精度不高,并且在自回归(AR)阶次p小于或等于滑动平均(MA)阶次q时无法估计出观测噪声方差.为此,本文提出了一种单独估计观测噪声方差的新方法,即先将ARMA模型近似为一高阶AR模型,再构建从观测数据1阶自协方差开始的Yule-Walker方程.由于充分利用了观测数据的统计信息,有利于提高观测噪声方差的估计精度,为后续的AR和MA参数估计精度的提高奠定了基础,也解决了p小于或等于q时观测噪声方差无法估计的问题,仿真和实验结果验证了该方法的有效性.     

应用DDS方法辨识机器人振动的模态参数

本论文应用DDS方法将机器人振动系统的运动微分方程转换到一个自回归滑动平均(ARMA)模型.根据 ARMA 模型的性质,给出了ARMA模型的特征参数与机器人振动模型之间的关系,并通过Householder-OUSEHOLDER变换的最小二乘法.对机器人振动系统的模态参数进一步识别.     

基于Nios的聚类向量ARMA轴温探测器——列车轴温探测系统的信号处理

阐述了在向量ARMA(AutoRegressive Moving Average,自回归滑动平均)模型中的聚类算法,介绍了使用Altera DSP Builder和SOPC Builder实现聚类向量自回归滑动平均的卡尔曼最优平滑滤波器的设计方法,给出了聚类向量ARMA卡尔曼最优平滑滤波器的设计框图,比较了32阶FIR（Finite Impulse Response ,有限冲击响应）滤波器与聚类向量ARMA卡尔曼最优平滑滤波器的滤波效果,同时还利用16阶DITFFT（时间抽取快速Fourier变换）展     

动态纹理背景的建模

针对室外条件下动态纹理背景,采用自回归运动平均(ARMA)模型建立背景模型,并引入快速增量主元分析(IPCA)算法对模型进行降维,并辨识其中参数,实现最大似然估计.运用增量主元分析算法,不需要估算协方差矩阵,直接可以递增地得到特征向量和奇异值,计算出样本序列的主要元素.完成参数辨识后,ARMA模型可以合成无限长度的预测图像序列.最后,仿真实验证明了算法的有效性.     

基于时序模型系数的半导体批次过程故障诊断

提出了一种基于时间序列模型系数的半导体批次过程故障诊断方法。首先将带有批间控制器的半导体批次过程推导成自回归移动平均(ARMA)模型,接着利用参数重置递推扩展最小二乘法(PRELS)辨识半导体批次过程的ARMA模型,得到模型系数矩阵,然后通过多变量统计过程控制(MSPC)方法监测模型的系数矩阵,确定故障发生的批次,最后采用支持向量机(SVM)对过程参数故障进行隔离。与传统基于输入输出数据的故障诊断方法相比,基于模型系数的方法可有效提高诊断的准确率。仿真结果验证了方法的有效性。     

多变量CAR模型自动辨识机

本文提出了多变量受控的自回归(CAR)模型自动辨识机,已用BASIC语言软件包在Z—80机上实现,它由三部分组成:(ⅰ)递推最小二乘法参数估计器,(ⅱ)模型阶的F检验判决器;(ⅲ)节省参数模型的时滞和阶的F检验判决器,数值换拟例子证明了所提出的方法的有效性。     

基于ARMAX模型的子空间辨识算法

针对于子空间辨识算法辨识闭环系统时,由于输入信号与不可测噪声是相关的,往往会得到有偏估计的问题.提出一种采用自回归滑动平均模型(ARMAX)的闭环子空间辨识方法,通过扩展最小二乘方法(ELs)估计ARMAX模型中的马尔科夫(Markov)参数,使用预测的子空间辨识方法(PBSID)获取系统参数矩阵,避免了采用高阶自回归模型(ARX)所导致的过大的估计方差等问题.算法实例验证结果表明,改进方法能够获得较好的闭环系统一致性估计,辨识精度较高,有非常良好的应用前景.     

Construction of the Minimum Phase Systems and the Problem of Stabilization

For the affine system with vector control, the problem of stabilization of the zero equilibrium position can be solved by stabilizing the zero output value, provided that the affine system is a minimum phase one. The minimum phase systems enable essential extension of the set of stabilizing feedbacks. In order to make use of this approach in the problem of stabilization of the stationary point of the multidimensional affine system, one needs a virtual vector output such that the system is a minimum phase one. Existence of conditions for these virtual vector outputs and a method for their determination were presented. Virtual output-based stabilization is attained by constructing the Lyapunov function for a closed-loop system.     

基于可决系数的自适应关联规则挖掘算法

针对以频繁项集产生?规则产生为核心的两阶段关联规则挖掘,存在需要人工以先验知识指定最小支持度和最小置信度阈值的缺陷。本文提出以支持数和置信度为依据,采用曲线拟合技术,根据可决系数自动确定曲线的次数及对应多项式的算法AARM_BR(Adaptation Association Rule Mining Based on Determination Coefficient R²),从而确定支持度和置信度阈值。在标准数据集Trolley和Groceries上进行关联规则挖掘实验,结果表明本算法更具有数据依赖性,在用户不具备先验知识的情况下,无须人为指定多项式阶次、支持度和置信度阈值的优点。     

Index based optimal anatomy of a metamorphic manipulator for a given task

This paper introduces an approach for the determination of the best anatomy of a metamorphic manipulator for a given task at a given location. The location of the task is determined by maximizing the performance of a current industrial fixed anatomy robot. Two types of tasks are considered: a point to point task and a path following task, where in the first case the approximated minimum of the manipulability index is formed along the task points and in the second case the approximated minimum of the manipulator velocity ratio is formed along the line segments. These indexes are maximized in order to determine the best anatomy for the task. The proposed approach is tested and the results show that the determined best anatomy for each type of task acquired higher performance than the respective one achieved by the fixed anatomy manipulator.     

A geometrical workspace calculation method for cable-driven parallel manipulators on minimum tension condition

In this paper, a geometrical expression to delineate the sum of tensions (i.e. minimum 1-norm tensions) of 2-DoF planar cable-driven parallel manipulators (CDPMs) is proposed and proofed, which can be used to reduce calculation time of workspace determination. Furthermore, this paper also presents a systematic analysis on the relation between cable tension and workspace by means of convex analysis. In order to obtain wrench-feasible workspace, a unified and feasible algorithm is adopted in simulation examples of spatial CDPMs with different configurations, which demonstrate that the proposed method is valid and straightforward to calculate workspace.     

Predictions of the minimum relative depth of die cavity and the minimum amount of preforming in the radial extrusion of tubular components

In this study, both the minimum relative depth of die cavity and the minimum amount required in preforming are determined to produce a defect-free product of tubular component in the preforming technique of radial extrusion. This scheme is based on the least amount of material dissipation in radial extrusion of tubular component. A finite element based code is utilized to investigate the effects of various amount of relative deformation and different relative depths of die cavity on the material flow characteristics resulting in the movements of the defects locations. An algorithm is thus developed for the determination of minimum relative depth of die cavity and the minimum amount required in preforming, and the abductive network is also applied to synthesize the data sets obtained from the numerical simulations through the algorithm proposed in this study. Consequently, a prediction model is established for organizing the minimum relative depth of die cavity and the related minimum amount of preforming of the tubular component that the locations of induced defects are just coincided with the inner surface of the tube at the end stroke of deformation.     

基于数据驱动多输出ARMAX建模的高炉十字测温中心温度在线估计

高炉（Blast furnace,BF）炼铁中,十字测温作为炉顶温度和煤气流分布监测的最主要手段,对高炉的安全、稳定和高效运行起着重要作用.然而,由于高炉炉顶中心部位温度较高,造成十字测温装置中心位置传感器极易损坏,并且更换周期长,因而无法及时判断炉顶煤气流分布.针对这一实际工程问题,本文基于时间序列建模思想,集成采用多输出自回归移动平均（Multi-output autoregressive moving average,M-ARMAX）建模、因子分析、Pearson相关分析、基于赤池信息准则（Akaike information criterion,AIC）与模型拟合优度联合定阶等混合技术,提出一种模型结构简单、精度较高且易于工程实现的十字测温中心温度在线估计方法.首先,提出利用因子分析与Pearson相关分析相结合的稳健特征选择方法选取多输出建模输入变量.然后,采用样本均值消去法预处理采集的高炉样本数据,使其成为离散随机数.基于离散随机数,建立算法简单、易于工程实现的M-ARMAX温度模型:为了克服传统基于AIC阶数确定造成模型阶次高、结构复杂的问题,提出在AIC准则基础上进一步引入模型拟合优度来选取模型最小阶,可保证模型估计精度的同时降低模型阶次;同时,采用可快速收敛的递推最小二乘算法辨识M-ARMAX模型参数,并用残差分析方法检验模型.工业试验和比较分析表明:建立的M-ARMAX模型能够根据实时数据同时对十字测温装置多个中心温度点进行准确和稳定估计,且模型估计误差符合高斯白噪声特性.     

Mathematical programming for the scheduling problem with alternate process plans in FMS

The purpose of this research is to consider the operational scheduling problem and the determination of production routing with alternate process plans simultaneously such that the advantages of routing flexibility are expected. The problem is formulated by using 0–1 integer programming regarding to the performance measure of either the minimum of the absolute deviation of meeting due date or the minimum of total completion time. The approach of mathematical programming generates the optimal schedule, rather than near optimal schedule or a better schedule, to meet the selected criterion. Two mathematical formulation models developed are presented and an example is shown in this paper referring to the either one of performance criteria.     

基于Walker星座的导航卫星自主导航的GDOP最小值分析

卫星自主定轨时选择测量卫星的几何构型对于定轨精度有着重要的影响.目前地面接收机与导航卫星的几何构型对定位精度的影响主要由GDOP（几何衰减因子）来衡量,GDOP值的下限限制了在一定的测量精度下用户定位的精度范围.通过构建Walker构型的导航星座将地面接收机GDOP取最小值时的边界确定,推广应用到自主导航环境下,利用均匀采样和遗传算法分别独立得到GDOP的最小值,同时用仿真数据验证了这个值的正确性.     

Robot Self—Location by Line Correspondences

This paper thoroughly investigates the problem of robot self-location by line correspondences.The original cotributions are three-fold:(1) Obtain the necessary and sufficient condition to determine linearly the robot‘s pose by two line correspondences.(2) Show that if the space lines are vertical ones,it is impossible to determine linearly the robot‘s pose no matter how many line correspondences we have,and the minimum number of line correspondences is 3 to determine uniquely (but non-linearly)the robot‘s pose.(3) Show that if the space lines are horizontal ones,the minimum number of line correspondences is 3 for linear determination and 2 for non-linear determination of the robot‘s pose.     

基于两步测量方法及其最少观测次数的任意量子纯态估计

量子状态层析所需要的完备观测次数d~2(d=2~n)随着状态的量子位数n的增加呈指数增长,这使得对高维量子态的层析变得十分困难.本文提出一种基于两步测量的量子态估计方法,可以对任意量子纯态的估计提供最少的观测次数.本文证明:当选择泡利观测算符,采用本文所提出的量子态估计方法对d=2n维希尔伯特空间中的任意n量子位纯态进行重构时,如果为本征态,那么所需最少观测次数memin仅为memin=n;对于包含l(2 6 l 6 d)个非零本征值的叠加态,重构所需最少观测次数msmin满足msmin=d+2l..3,此数目远小于压缩传感理论给出的量子态重构所需测量配置数目O(rd log d),以及目前已发表论文给出的纯态唯一确定所需最少观测次数4d..5.同时给出最少观测次数对应的最优观测算符集的构建方案,并通过仿真实验对本文所提出的量子态估计方法进行验证,实验中重构保真度均达到97%以上.