一种面向无参考的顺轨干涉SAR海面相干时间计算方法

利用顺轨干涉ATI-SAR(Along-Track Interferometric Synthetic Aperture Radar)计算海面相干时间时,需要获取噪声引起的前后通道间的相关系数的下降.目前常用的做法是以陆地目标或双基线干涉SAR(Synthetic Aperture Radar)中基线长度为零的干涉数据作为参考.而现有的顺轨干涉SAR多为单基线干涉,且获取的海面图像中往往不包含陆地目标.针对这一问题,本文提出了一种面向无参考的顺轨干涉SAR海面相干时间计算方法.首先从SAR图像功率谱中估计噪声功率,计算图像的信噪比;然后根据信噪比计算出噪声引起的前后通道的相关系数的下降,进而计算出海面相干时间.将该方法运用于实际的机载顺轨干涉SAR数据,获取了海面相干时间.实验结果与仿真结果的对比表明,该方法能获取较为准确的海面相干时间,验证了本文方法的有效性.     

干涉条纹图估测基线的方法研究

根据SAR干涉测量原理，提出了利用干涉相干条纹估测基线的方法，并以欧洲空间局资源卫星1号和2号携带的SAR系统在中国新疆伽师试验区获取的串接式（Tan-dem）干涉测量数据为数据源，对提出的方法进行了验证，结果表明，该方法能有效地估测基线，并弥补不能有效获取卫星轨道参数时，对基线进行估测的缺陷，该方法尤其适用于差分干涉测量的研究。     

用SIR-C航天飞机双频极化干涉雷达估计植被高度的方法研究

基于重复轨道航天飞机SIR—C的L和C波段极化干涉雷达数据,提出了一种将三阶段反演方法(Three—Stage Inversion)和ESPRIT(Estimation of Signal Parameters via Rotational Invariance Techniques)算法相结合的新的相位基植被高度估计方法。首先,基于L波段极化干涉数据,用三阶段反演方法获取植被覆盖地表下地形有效散射中心的相位;接着,基于c波段极化干涉数据,用ESPRIT算法估计植被冠层有效散射中心的相位;最后,对二者的相位差进行相位到高度的转换得到植被的高度。根据研究区域的实测林相数据,对该算法估计的植被高度结果进行了比较和验证,结果表明：该方法具有足够高的植被高度估计精度,如果能得到P和X波段的双频极化干涉雷达数据,该方法估计植被高度的精度将会得到进一步的提高。     

白光干涉数据的三维去噪研究

介绍了白光干涉系统噪声来源及其对白光干涉图像的影响;扫描白光干涉过程中既引入了时间噪声,又引入了空间噪声,使干涉数据的噪声呈复杂的时空分布;只对干涉数据进行单分量的传统去噪方法无法完全去除噪声,且会造成图像数据的畸变;利用三维噪声理论分析了白光干涉系统的噪声特性,并对干涉数据进行时间和空间分量的噪声处理,在去除噪声的同时,又避免了原始数据发生时空畸变;为了实现实时处理,设计了相应的基于图形处理器(Graphic Processing Unit,GPU)的并行算法,解决了三维去噪与重建耗时严重的问题。     

基于眼动数据的ATM机界面可用性测试

目的 ATM机作为面向大众的公共自助设施,应具有良好的可用性与通用性。测试ATM机的通用性水平并发现可用性问题。方法将ATM机界面原型作为实验研究对象,使用眼动仪进行绩效测试,并结合系统可用性量表和事后访谈评估界面。结果实验分青年、老年两组对现有ATM机界面进行绩效测试,获取有效性与效率指标数据;根据实验所得眼动数据进行了轨迹图和热点图的对比分析;结合满意度问卷与用户访谈综合分析了现有ATM界面在不同年龄组的用户操作中反映出的可用性水平。结论目前ATM机界面的整体可用性较低,两组被试在完成率和效率上都不高;老年组被试的完成率及效率都较年轻组被试低且满意度也更低。     

二维纳米激光测量系统的研究

介绍了一种新型的纳米级精度的二维激光干涉测量系统．该系统以光学8倍频的耦合差动式干涉光路为基础，在获取大量测量数据的基础上，利用线性回归的方法对测量结果进行处理．系统结构设计简洁紧凑，符合阿贝原则和结构变形最小原则，与普通的迈克尔逊干涉仪相比具有光路布局对称性好，光程差倍增，抗干扰能力强等优点．通过与电容测微仪比对的方法对该系统进行检测．结果表明，该干涉测量系统的精度为10～12nm．     

芯片表面形貌检测方法及系统研制

传统检测方法由于测量范围小、测量时间长,不能满足在线、在机和大幅面的检测需求,本文提出了一种白光扫描干涉和原子力显微测量相结合的芯片表面形貌测量方法。在分析了白光扫描干涉测量技术和原子力显微测量技术特点的基础上,构建了两种方法互补的测量方案：首先采用白光扫描干涉测量技术对芯片进行大范围快速扫描,再通过原子力显微测量技术对关键区域进行精细扫描检测。基于此进行了测量系统的硬件设计与集成,测试结果表明,该系统能够在毫米级检测范围内实现纳米级表面形貌的精确测量。     

基于VC++的光谱分析软件设计与实现

 为了对红外干涉型光谱分析仪采集得到的干涉条纹进行分析,准确地反演相应的光谱分布函数,设计了基于迈克尔逊干涉结构的光谱分析算法,并通过VC++完成了对应的光谱分析软件．实验对830,940,1 064 nm三个常用近红外波长进行测试,分别与FTIR500型光谱仪的检测数据和MATLAB仿真数据进行对比．实验结果显示,本算法获得的光谱数据在主波长位置选择以及幅值探测上与FTIR500型光谱仪相近,而在噪声、杂波抑制方面优于FTIR500型光谱仪．数据处理速度略低于MATLAB,但光谱分布函数信噪比要高于MATLAB的数据处理结果,证明本系统具有一定的优越性．     

齿轮传动系统损伤检测与多故障分类研究

研究了数据挖掘的支持向量机的智能故障检测与诊断方法。通过对齿轮系统在不同的运转状态下的工作状况进行试验测试分析,获取了有关的测试信号,并对不同的故障振动特征信号进行了特征提取与分析研究。在此基础上将支持向量机引入到齿轮传动的损伤检测与诊断之中,建立了两分类和多分类分类器,研究了支持向量机的两分类和多类分类算法。通过分析处理、训练和测试仿真数据以及齿轮振动特征信号,对齿轮系统在各种不同转速下不同故障进行了预测、分类和诊断。研究表明, 支持向量机能够很好的区分不同运转状况下各种典型齿轮损伤与故障,低转速下识别率更高,为95%,特别是对各种复合类故障具有较高的识别精度、识别率在81%以上。它在齿轮故障诊断中具有较好诊断识别能力与发展前景,是一种有效地损伤检测与诊断新方法。     

非接触式激光测量真空状态下温度的方法

研究了一种利用激光干涉原理进行非接触式测温的方法,该方法能在不破坏样品表面的前提下,准确、方便、廉价地获取玻璃在真空中的真实温度.在相关实验中以Pilkington TEC系列钠钙玻璃为样品,在150～500℃之间同时对样品进行热电偶测温和激光干涉测温,实验数据表明,非接触式激光测温方法得到的温度数据与热电偶直接接触样品表面测得的温度数据在误差10℃范围内基本相同.证明了这种非接触式的方法可用于以玻璃为基底的太阳能电池生产领域.     

Tool collision detection in machining using spatial representation technique

The objective of this research work is to develop a tool collision detection system for turning operations. Spatial representation technique is used for object representations of cutting tool, workpiece, workholding device and machining envelope to detect intersection or interference between solids. The algorithm, developed using Turbo C++, is able to detect various types of collisions and to modify the machining envelope to avoid collision. The algorithm is also able to determine the cutting tool efficiency based on the machining envelope. The effectiveness of the method has been illustrated with an example.     

Process planning for rotational parts using the generalized travelling salesman problem

Process planning for rotational parts can be described as deciding on the order of machining the various features. However, this problem entails more details since each feature, hypothetically, can be machined by a left- or a right-hand tool, and in many cases, the parts can be chucked from either end. Another complicating factor is that some features require machining from both directions (with both left- and right-hand tools) and some can be machined by either tool. Thus, the process planning problem for rotational parts decides on the order of machining the various features, the chucking direction and the tool to be used for each feature. Every time the part is chucked or to a lesser extent when the tool direction is changed, the accuracy of the feature is compromised and productivity is reduced due to time spent on non-cutting operations. A good process plan, therefore, should consider the productivity as well as the overall accuracy of the part. This paper presents an algorithmic graph-theoretic approach towards determining the optimal process plan such that the part's overall accuracy is maximized, given the accuracy of the machine and the features tolerances. In particular, we describe the process-planning problem as a generalized travelling salesman problem and then use known procedures to solve the problem. This solution minimizes the overall impact of rechucking operations and tool changes on the process accuracy. Therefore, in addition to the improved precision, the process plan results in faster and more efficient operations.     

Optimal tolerance re-allocation for the generative process sequence

In generative process planning, the sequence of machining processes is decided according to the specifications of parts, such as tolerance values. However, in order to obtain the minimal manufacturing cost, the machining process sequence needs to be considered before tolerances are assigned. It is therefore difficult to assign optimal tolerances so that a minimum manufacturing cost is achieved. This paper presents an iterative approach for reallocation of tolerance within the given functional constraints to minimize the manufacturing cost. With the given values of tolerance and corresponding process sequences, which are derived from a handbook or a designer's experience as initial inputs, each iteration of tolerance re-allocation tries to improve the total cost by shifting tolerances along the different processes in the current sequence. The re-allocation problem is formulated as a mixed integer nonlinear programming problem. The Lagrange Multiplier method has been used to solve nonlinear programming, and an exhaustive search method has been adopted to guarantee the global optimum in solving the zero-one algorithm. A prototype system has been implemented in an object-oriented programming environment and a case study is presented to demonstrate the capability of the system.     

Optimization of ultrasonic machining for fabricating square holes using the Taguchi methods

Abstract

One of the solutions for making quality molded plastic products is to adopt post-machining to avoid the warpage problem. Considering many traditional and non-traditional machining techniques, we have suggested that ultrasonic machining may be suitable for fabricating square holes in plastics. Because many factors are involved in ultrasonic machining, the objective of this research was to find the optimal conditions for ultrasonic machining by the Taguchi methods. The plastic material utilized in the study was PMMA, and there were two quality characteristics adopted for the experiment. One was the area deviation percentage (ADP) of square holes created by ultrasonic machining, and the other was the average thrust force during the machining process. Four factors, which were tool material, tool type, abrasive, and feed rate were considered in the experiment. The optimal conditions for ultrasonic machining to minimize ADPs and to reduce the average thrust force, respectively, are reported in the paper. These conditions can be applied to machine PMMA using ultrasonic machining in future applications.     

Influence of various metal powder mixed dielectric on micro-EDM characteristics of Ti-6Al-4V

Ti-6Al-4V, an advanced engineering material is difficult-to-machine using conventional machining process due to its high strength. It has properties like low weight ratio, outstanding corrosion resistance along with high level of reliable performance in micro components. Micro-electro-discharge machining (Micro-EDM), a popular nontraditional machining process has been identified as the most appropriate machining process for such material. In this paper, the effect of various conducting powders such as copper, nickel and cobalt with different concentrations are mixed with deionized water dielectric, on various responses such as material removal rate (MRR), tool wear rate (TWR), overcut (OC) and taper has been presented. Also, principal component analysis (PCA) has been applied to select the optimal parametric combination of micro-EDM process to achieve optimal values of MRR, TWR, OC and taper during micro-through hole machining. The optimal process parametric setting obtained from the proposed approach is peak current (I_p) of 1.5 A and cobalt (Co) powder concentration of 4 g/L so as to obtain the desired responses. It is also observed from the SEM image that the machined profile and surface topography obtained through the multi-objective optimal parametric combination based on PCA is quite satisfactory and can be applied to achieve geometrically more accurate micro-through holes on Ti-6Al-4V.     

New criteria for tool orientation determination in five-axis sculptured surface machining

The problem of optimal tool orientation determination in five-axis flat-end milling of sculptured surfaces is examined in this paper. The optimal tool orientation avoids local and global gouging of the tool and maximises a specific criterion related to machining efficiency. Two new criteria are introduced in this paper to quantify the tool orientation quality at a cutter contact point: infinitesimal machining volume (IMV) and infinitesimal machining area (IMA). The IMV criterion is used to maximise the material removal at the cutter contact point. The IMA criterion attempts to identify tool orientations that would lead to minimised overall tool path length. Using one of these criteria, an optimisation problem can be formulated to determine the optimal tool orientation among feasible gouge-free orientations. It is shown that the commonly adopted criterion of machining strip width in the determination of the optimal tool orientation cannot contribute towards maximising the material removal and does not really result in minimum overall tool path length. Results from various case studies have indicated that the newly introduced criteria can be used to generate optimal tool orientations that significantly increase machining efficiency.     

Optimal cutter selection for complex three-axis NC mould machining

Proper cutter selection can reduce NC machining time greatly. Most researchers select cutters based on tool path generation, which is very time consuming, and only a very limited number of cutters can be selected. A new cutter selection methodology for complex mould machining based on an efficient interference detection algorithm is introduced. With this approach, the feasible regions for the candidate cutters are first identified without tool path generation. The machining times for different cutter combinations are then estimated based on the areas of the feasible regions and the cutter feed rates. The set of cutters that can machine the workpiece with minimum time is selected as the optimal candidate cutters. Since no tool path needs to be generated before cutter selection, the cutters can be selected efficiently, and the number of the cutters that can be used in NC machining can be quite large. The system has been tested with several industrial parts and it can select optimal cutters effectively and efficiently.     

Development of an integrated model for process planning and parameter selection for machining processes

The concept of ‘do it right the first time’ in the machining industry not only expects the best quality products but also at the best possible cost. The cost of machining depends on intelligent process planning and selection of machining parameters such as speed, feed, and depth of cut. The problem of machining parameter selection has received great attention by researchers and many techniques have been developed. A review of these techniques reveals that the selection of the machine and cutting tool is done before the process of cutting parameter selection and process sequencing, and often the selection is based on experience. The current research is an attempt to develop an integrated model (ExIMPro: Expert system based Integrated model for Machining Processes) which finds the sequence of operations with set of machines, tools, and other process parameters to minimise the cost of machining for a cylindrical part. This system consists of existing expert system Machining Parameter SELection (MPSEL) for machine and tool selection and a Microsoft Excel® and Visual Basic® based parameter selection model. The present model focuses on turning and cylindrical grinding operations but other processes can be incorporated with little modification to the software.     

Machining parameters optimization for satisfying the multiple objectives in machining of MMCs

The metal matrix composites (MMCs) have gained acceptance in an extensive range of applications owing to their high strength to mass ratio. Machining of such complex MMCs is often challenging. It is essential to optimize the controllable machining parameters to simultaneously attain manifold objectives. In the current work, response surface design is created for experiments, and Genetic algorithm (GA) combined with Principal Components Analysis (PCA) coupled Grey Relational Analysis (GRA) is employed to improve the straight turning process of MMCs. The procedure is demonstrated by machining aluminum-based MMC with 25% SiC particulates. The procedure aims at identifying optimal combination of machining parameters to obtain high surface quality at lower cutting force without increasing the specific power consumption. PCA is helpful in providing the individual uncorrelated quality characteristics called as quality indices that do not have any influence on other responses. Individual quality indices have been utilized to obtain the grey relational grade through GRA. GRA has been used to alter manifold quality indices into singular column of grey relational grade as a means to change the manifold objective problem into a sole objective problem. Then, GA has been used to get the optimal parameters combination. The novelty present in this work is the avoidance of correlation existing among the quality characteristics and combining of the GRA and GA. This is an endeavor to augment the performance and accuracy of GA to solve the optimization problem associated with the turning operation.     

A mixed-integer programming approach to the machine loading and process planning problem in a process layout environment

A mixed-integer programming (MIP) model for the machine loading and process planning problem has been developed for a process layout environment. This model relaxes the most commonly used assumption that each operation can be assigned to only one machine. In this model, the most important production cost components: machining cost, material handling (MH) cost, setup cost and machine idle cost are considered. In addition, by assigning penalty cost for long makespans, the optimal process routes for each part and optimal load for each machine in terms of minimum production cost can be obtained with lower makespan. An example problem is solved, followed by discussions with regard to the impact of batch splitting, setup cost and demand change on optimal solution.