工业过程多变量时滞系统的预测函数控制

为提高工业过程中多变量、时滞和强耦合系统的动态特性和耦合能力,文章将预测函数控制与解耦控制相结合,给出了一种新的多变量时滞系统预测函数控制算法.该法首先给出了解耦控制算法,通过对解耦后的对象幅频和相频特性分析,获得对象的一阶加纯滞后模型;然后根据预测函数控制原理对解耦后的多个单变量时滞对象设计控制器;最后仿真结果表明,该控制方法具有在线计算量小、跟踪快速和精度高等特点,能够实现复杂多变量系统的更有效控制.     

基于辅助变量法的系统参数辨识

重点介绍了系统参数辨识中辅助变量法的基本原理,并通过某导弹天线伺服机构数学模型的辨识实例,具体说明了基于辅助变量法的参数辨识在Matlab中的实现方法.结果表明基于辅助变量法的参数辨识比最小二乘法具有更高的精度,尤其当噪声为有色噪声和噪声模型结构不好确定时,辅助变量法更加有效。     

时滞动态补偿方法在泵车臂架液控系统中的应用研究

针对泵车臂架电液控制系统的非线性时滞问题,提出一种适用于泵车臂架结构的动态时滞补偿方法。该方法以泵车臂架末端振动位移历史数据为基础,通过时间序列法对此后某段时间内的泵车臂架的振动姿态提前预测,对系统中的非线性时滞进行动态补偿,为泵车臂架末端位置控制系统提供可靠的反馈数据,为液压阀控缸的位置和速度控制系统提供实时的参考输入。外场试验验证了该时滞补偿算法在提高泵车臂架操控平顺性和阀控缸响应速度方面的有效性。     

临床纵向数据缺失的多重填补及其敏感性分析

目的：以芪蛭通络胶囊治疗缺血性脑卒中临床试验为例来指导临床纵向数据缺失的多重填补及其敏感性分析,并强调敏感性分析的重要性。方法：应用SAS中的PROC MI过程实现临床纵向数据缺失的多重填补及其敏感性分析。 结果：实例分析中,多重填补后,芪蛭通络组的下肢运动评分改善值优于安慰剂组（P值均小于0.01）,而两种非随机缺失机制假定下的敏感性分析结果与随机缺失下的结果一致。结论：多重填补结合使用敏感性分析可以确保多重填补结果的稳健性,建议临床研究者对数据填补后还要进行敏感性分析。     

基于高斯过程的倾转旋翼机器人模型辨识

针对一种新型倾转旋翼水空跨域机器人,采用一种非参数系统辨识算法,目的是在数据量较少情况下,实现对跨域机器人的动力学建模。该算法基于多输出高斯过程,在无需机器人动力学模型的先验知识情况下,可通过采集试验数据,获得机器人的模型。为降低高斯过程回归的计算复杂性,采用局部高斯过程回归,对机器人的动力学模型进行辨识。为验证该辨识方法的有效性,在空气中以倾转四旋翼模型为试验平台,利用遥控获取的机器人数据进行验证。在机器人相同控制输入量下,对比模型预测值和实际测量值,证明该机器人动力学辨识方法的有效性和局部高斯过程回归的快速性。     

多变量电液伺服系统的时序辨识

本文提出了用多变量CAR时序模型来辨识电液伺服系统的阶和参数。文中用辅助变量法和LILC定阶准则结合来估计电液位置和力偶系统的阶和参数,根据估计参数的渐近正态性确定模型的子阶和时滞。用辨识模型仿真和实测阶跃响应比较,证明所得到的辨识模型是正确的。该方法可用于任何多变量机械系统的模型辨识。     

基于神经网络的盾构液压推进系统的参数辨识

为了提高盾构液压推进系统速度控制的精度,减小建立数学模型与实际系统之间的偏差,提出基于时延神经网络和系统参数上下界值的动态系统参数辨识方法。在建立盾构液压推进系统速度控制模型的基础上,利用神经网络在待辨识参数的范围内对系统的不定参数进行辨识,搜索出更逼近实际盾构推进系统的一组参数,满足对系统输入实际信号时能够准确地得到实际系统的输出。利用盾构模拟实验台的另一组采样数据对系统辨识结果进行验证,结果表明：该辨识方法得到的模型能够逼近真实的物理系统。     

基于RELS方法的激光焊接过程非线性辨识

激光焊接过程是一个很复杂的非线性系统,用机理建模法很难得到系统的数学模型,通常采用基于输入输出数据的系统辨识方法。考虑到实际系统通常伴随着有色噪声,本文采用递推增广最小二乘法(RELS)辨识。这种方法考虑了噪声模型,可以得到模型参数的无偏最小方差估计。试验数据和模型仿真的拟合效果表明,递推增广最小二乘法是激光焊接过程非线性辨识的一种行之有效的方法。     

基于神经网络的盾构液压推进系统的参数辨识（英文）

为了提高盾构液压推进系统速度控制的精度,减小建立数学模型与实际系统之间的偏差,提出基于时延神经网络和系统参数上下界值的动态系统参数辨识方法。在建立盾构液压推进系统速度控制模型的基础上,利用神经网络在待辨识参数的范围内对系统的不定参数进行辨识,搜索出更逼近实际盾构推进系统的一组参数,满足对系统输入实际信号时能够准确地得到实际系统的输出。利用盾构模拟实验台的另一组采样数据对系统辨识结果进行验证,结果表明:该辨识方法得到的模型能够逼近真实的物理系统。     

基于实验模型的气动人工肌肉位置控制研究

针对目前气动人工肌肉数学模型建模过程较为复杂的问题,提出一种气动人工肌肉的实验模型辨识方法。以气动人工肌肉-质量系统为实验对象,对其输入气压与气动肌肉收缩量之间的关系模型进行实验辨识研究。采用欠阻尼二阶系统来描述气动人工肌肉充气与排气过程的动态响应,并利用实验数据对模型参数进行辨识。最后基于辨识得到的模型进行气动人工肌肉的位置控制,实验结果验证了模型的有效性。     

Stochastic approach to experimental analysis of cylindrical lapping process

Lapping is a very complicated and random process resulting from the variation of abrasive grains by its sizes and shapes and from the numerous variables which have an effect on the process quality. Thus it needs to be analyzed by experiment rather than by theory to obtain the relative effects of variables quantitatively.In this study, the cylindrical lapping experiment designed by Taguchi's L₈ orthogonal array was performed and analyzed by Yates' ANOVA table. As a result, effective variables and interaction effects were identified and discussed. Also the optimal variable combination to obtain the largest percentage improvement of surface roughness was selected and confirmatory experiments were performed.     

Optimal temperature variable selection by grouping approach for thermal error modeling and compensation

One of the difficult issues in a thermal error compensation scheme is to select appropriate temperature variables as well as to obtain accurate thermal error component models. In this research, an optimization method is presented to overcome this difficulty. The optimization objective function is formulated by a modified model adequacy criterion based on the Mallows' C_p statistic. A new search method is developed for discrete search domains with non-directional or unknown-order variables. The search process includes correlation grouping, representative searching, group searching and variable searching. It not only ensures optimal results but also reduces computational time greatly. One modeling example is presented. The optimal model is found with a 0.982 R²-value using four temperature variables selected from 46 candidates of temperature variables. The largest error residual is reduced down to 2.2 microns from 20.0 microns. The comparison of modeling results from the proposed approach and three other modeling methods is addressed as well.     

基于USB接口的超声TOFD检测系统研制及其应用

为了定量检测厚壁焊接结构中的缺陷,对基于USB接口的便携式超声衍射时差法(TOFD,timc of flight diffraction)的缺陷检测及定位系统进行了研究.为了便于缺陷的定量及定性识别,引入了信号互相关算法,实现了检测图像的校正.利用该检测系统进行了锅炉压力容器焊缝的检测.结果表明,该系统所获得的D扫描图...     

A new approach to the modeling of oblique cutting processes

A new upper-bound model that incorporates force equilibrium parallel to the cutting edge is proposed for oblique cutting operations. The energy approach is framed in terms of the normal shear angle and two new fundamental variables that characterize the energy requirements of the oblique cutting process. SLIP is a kinematic variable and is defined as the ratio of the shear velocity imparted to the chip on the shear plane parallel to the cutting edge, to the incoming velocity in the same direction. RATIO is a force-based variable and is defined as the ratio of the friction force on the rake face to the resultant shear force in the shear plane. Calibration of the model for either real time identification purposes or for process planning/optimization requires experimental force data but no shear angle data, making it very suitable for the analysis of cutting operations with non-straight cutting edges. The relationships between SLIP, chip flow angle, and RATIO are shown to be remarkably consistent over a very wide range of inclination and normal rake angles. Finally, the authors demonstrate the success of the model using existing experimental data.     

基于RANSAC和卡尔曼滤波的窄焊缝识别北大核心

针对高效鲁棒的焊缝识别对于提高焊接质量和效率具有重要的作用,提出了一种基于线结构光的薄壁件窄焊缝快速识别方法。首先,对激光照射的焊缝图像进行去噪和增强;其次,分别利用Ransac算法和最小二乘法对焊缝和激光条纹进行直线拟合;接着,以二直线交点为中心建立ROI,连接ROI中的连通体获得修正后的交点;最后,对修正后的交点加以平移,以斜率为状态变量进行卡尔曼滤波,以修正后的交点为中心建立ROI并计算ROI中各点斜率,寻找具有最接近滤波后的斜率的点,将其反向平移后的坐标作为交点的最终值。实验证明该算法具有较高的鲁棒性和较快的处理速度,可满足窄焊缝的实时识别需求。     

基于FIR分解的轴承故障快速诊断方法研究

为实现轴承故障的快速准确诊断,以互相关和互信息为基础构造一种针对轴承的快速故障诊断方法。该方法首先运用有限长单位冲激响应（Finite Impulse Response,简称FIR）滤波器对各单一故障（包括内圈、外圈、滚珠、保持架）振动信号进行分解,降低信号分解过程中因模态混叠造成的干扰,以力学分析建立的各故障振动模型为参考,对分解后的子信号采用互相关分析法,选出表征故障特征的子信号,计算子信号透露的信息量——互信息,用于构造故障特征矩阵,最后由K最近邻分类算法（K-Nearest Neighbor,简称KNN算法）的识别结果验证该算法对实现轴承故障快速识别具有优势。     

基于Sobol’法的机床关键几何误差元素辨识方法研究

针对机床几何误差元素多、误差测量与辨识过程繁琐等问题,利用Sobol’全局灵敏度分析方法对空间误差模型中的几何误差元素进行灵敏度分析,筛选出影响较大的几何误差元素,从而降低误差测量与辨识过程的复杂度,简化空间误差模型。以螺旋理论为建模基础,建立机床空间误差模型;对所有几何误差元素进行Sobol序列抽样并通过蒙特卡洛估计法求解灵敏度,计算各误差元素的一阶灵敏度值及全局灵敏度值,从21个误差项中筛选出对机床空间误差影响较大的12项;将简化模型与完备模型进行对比,空间误差元素简化率为48%,其预测精度大于80%,说明了误差元素筛选的有效性,为机床空间误差建模、误差元素辨识以及空间误差补偿工作的简化提供参考。     

The statistical modeling of surface roughness in high-speed flat end milling

Surface roughness is one of the most important requirements in machining process. The surface roughness value is a result of the tool wear. When tool wear increase, the surface roughness also increases. The determination of the sufficient cutting parameters is a very important process obtained by means of both minimum surface roughness values and long tool life. The statistical models were developed to predict the surface roughness.This paper presents the development of a statistical model for surface roughness estimation in a high-speed flat end milling process under wet cutting conditions, using machining variables such as spindle speed, feed rate, depth of cut, and step over. First- and second-order models were developed using experimental results of a rotatable central composite design, and assessed by means of various statistical tests. The highest coefficient of correlation (R_adj²) (88%) was obtained with a 10-parameter second-order model. Meanwhile, a time trend was observed in residual values between model predictions and experimental data, reflecting the probable effect of the tool wear on surface roughness. Thus, in order to enhance the estimation capability of the model, another independent variable was included into the model to account for the effect of the tool wear, and the total operating time of the tool was selected as the most suitable variable for this purpose. By inserting this new variable as a linear term into the model, R_adj² was increased to 94% and a good fit was observed between the model predictions and supplementary experimental data.In this study, it was observed that, the order of significance of the main variables is as X₅>X₃>X₄>X₁>X₂ (total machining time, depth of cut, step over, spindle speed and feed rate, respectively).     

RANDOM-FUZZY SAFETY ANALYSIS FOR AN AERO ENGINE TURBINE DISK

A numerical simulation method is presented for the random-fuzzy safety analysis of an aero engine disk. Based on the equivalent transformation from a fuzzy variable to a random variable, the equivalent random Probability Density Functions(PDFs) are got from their corresponding Fuzzy Possibility Distributions(FPDs) for the fuzzy variables. In that case the perfect numerical simulation method for the random uncertainty is employed to solve the fuzzy uncertainty. For the complex structure such as the aero engine disk with implicit relationship between the input basic variable and the response variable, the equivalent PDFs of the input basic variables are delivered simultaneously to the response variable by an empirical PDF, which is simulated by Finite Element Method(FEM). Then, in view of the fuzzy application requirement occurring in engineering usually, the reliability definition and calculation are discussed for the aero engine disk with multiple fuzzy failure modes. On the other hand, through the inverse transformation of the fuzzy variable to the random variable, the FPDs of the response variables can be calculated from their empirical PDFs as well.     

PMSM分数阶滑模变结构MRAS仿真

针对永磁同步电机(PMSM)模型参考自适应控制系统(MRAS)中存在的抖振及转速估计精度低的问题,提出了一种分数阶滑模变结构MRAS转速辨识方法。首先建立了PMSM的参考模型和可调模型,利用参考模型和可调模型的电流输出误差来构造分数阶滑模面,其次设计了电机速度观测器,用Lyapunov稳定性理论证明了其可行性。最后利用MATLAB/SIMULINK软件搭建仿真模型进行仿真。仿真结果验证了PMSM分数阶滑模变结构速度观测器的有效性,具有更加精确地对转速进行跟踪的能力,有较好的鲁棒性。