基于随机森林算法的汽车轴承故障检测

针对传统机器学习算法受输入变量限制、且易出现过学习或欠学习,提出不受输入变量限制且存在大量数据缺失时有很好保持精确性的随机森林算法对汽车轴承故障进行检测。对采集到样本数据进行滤波处理,抑制信号中噪声;利用随机森林算法对采集到的时域信号进行分类标识,确定包含故障信息的信号序列;再将信号转换到频域,利用随机森林算法对频域内信号进行检测,确定出故障频率;最后采集试验数据对所提及算法进行验证,结果表明:相比于传统的机器学习算法,随机森林算法响应速度快,且准确率高。     

基于Popov超稳定理论的非线性电液伺服系统自适应控制研究

由于电液伺服系统死区非线性的存在，在正弦信号输入时，系统输出中存在高次谐波，使系统输出信号严重失真，对系统性能分析造成影响。本文基于Popov超稳定理论，设计了一种模型参考自适应控制（MRAC）系统，能消除死区对系统输出的影响，其稳定性分析显示，它能使跟踪误差迅速消失。从仿真结果卜看，它能有效跟踪参考模型输出，改善系统的响应特性。     

电子束扫描线圈高速驱动器研制

针对电子束高速偏转扫描的需要，研制了一种用于驱动电子束扫描线圈的高速驱动器。具体为通过降压、补偿、差分输入的方式设计驱动器输入级电路，提高电路对高速信号的响应及处理速度；通过射极跟随方式设计电流闭环调节电路，提高电路的闭环调节速度及稳定性；通过Cadence仿真软件对主回路结构仿真设计及参数优化。根据上述结果，研制出了电子束扫描线圈高速驱动器，并对其进行测试。结果表明，其驱动感性负载时，输出电流可高速闭环响应任意输入信号，工作频率达100 kHz，最大驱动电流±2 A。     

PID调节在PMAC运动控制器中的应用

文章讨论了一种通过调整基于PMAC的开放式数控系统的PID伺服控制器参数的调试方法。控制参数的调整与合理设定是数控系统获得理想控制效果的前提与关键环节,采用PEW IN TUNING软件给伺服电机输入阶跃信号和抛物线信号,通过分析系统响应曲线,对PID参数进行整定,可使系统获得良好的稳态特性和动态特性。     

滚动轴承故障信号处理方法与诊断试验研究

分析了滚动轴承故障振动信号的非线性、非平稳性特征,基于经验模态分解法（EMD）在处理此类信号中的优势,研究了滚动轴承故障信号的时频分析处理方法。通过EMD法将滚动轴承故障原始振动信号分解为多个平稳的IMF分量之和;选取前8个IMF能量值作为频域特征并结合时域特征构成故障振动信号特征集合,作为BP神经网络的输入;建立了滚动轴承故障诊断的BP神经网络模型,利用BP网络的自学习机制进行网络训练,得到了输入特征与故障模式之间的映射关系;通过对滚动轴承不同类别的故障诊断试验,验证了该方法的可行性。     

基于AMESim 的数字缸建模与仿真

本文简述了数字缸的工作原理,利用AMESim建立该液压系统的整体仿真模型,研究对系统动态特性影响的因素。通过仿真实验,得出不同负载、遮盖量和输入信号等因素对系统动态性的影响。仿真结果表明:负载对数字缸的位移和速度的影响不大;阀芯遮盖量为正时,响应速度快,但是泄露量大,影响系统精度;阀芯遮盖量为零时,响应速度快,泄露量小,但很难实现;阀芯遮盖量为负时,响应速度慢,实现比较容易;输入信号的频率越大,数字缸的速度越快,但正弦跟踪误差也越大。得出结论:根据负载合理选择数字液压缸型号;输入信号的频率和幅值要根据实际情况合理选用,减小数字液压缸的跟踪误差,保证稳定性和准确性;遮盖量尽量选择较小的负遮盖量,减少系统发热,提高能源利用率。     

基于MATLAB/Simulink的数字缸建模及性能仿真研究

建立了数字缸系统的方框图和传递函数模型,在给定参数下利用MATLAB/Simulink对系统性能进行仿真分析,得到了不同输入信号下系统的响应特性。仿真结果表明:该数字缸系统总体性能满足设计要求。     

永磁同步电动机变频调速的分析

针对永磁同步电动机与鼠龙式三相异步电动机变频调速的不同特点,本研究在传统的矢量控制系统中引入了模糊PI控制方法,利用模糊控制理论控制永磁同步电动机。在Matlab/Simulink仿真软件中设计了模糊PI控制模型,通过仿真分析可知,该控制方法相比三相异步电动机能更快地建立起旋转磁场,同时动态响应快,能实时抑制超调量,因此,采用模糊PI控制方法控制永磁同步电动机更具优越性。研制了一台样机,完成电机转速给定信号、转子位置信号、电压电流信号的采集,实现转速、电流双闭环控制算法,坐标变换及三相电流参考信号、PWM信号的产生,以及系统保护、显示等任务。验证了控制永磁同步电动机时,模糊PI控制下的转矩响应相比传统PI控制下的转矩响应更平缓。     

圆孔型2D伺服阀阀芯位移阶跃实验装置的设计

为了研究阀芯形式为圆孔型高低压孔2D伺服阀的阶跃响应时间,设计了摆轮式阶跃响应装置,并在此基础上完成对圆孔型2D伺服阀的阀芯位移的响应特性分析和稳定性分析的实验研究。实验结果表明:摆轮式机构可以给出一个理想的阶跃信号,当输入压力为21 MPa时,阶跃信号为0.9 ms,圆孔型2D伺服阀的阀芯位移阶跃响应时间为2 ms。     

基于响应面法的连杆中心质量研究

针对连杆毛坯杆部中心孔洞缺陷的问题,基于MATLAB平台,应用响应面法对模具结构参数进行了优化设计。首先利用中心复合设计抽取样本,然后对所得的数据进行数值模拟,作为响应面法训练的输入信号。以此来建立第一主应力为优化目标的响应面模型,并得到各参数组合的最优值。结果表明,第一主应力的误差在6%以内,验证了其响应面法的可靠性。     

Tool path generation and tolerance analysis for free-form surfaces

This paper focuses on developing algorithms that generate tool paths for free-form surfaces based on the accuracy of a desired manufactured part. A manufacturing part is represented by mathematical curves and surfaces. Using the mathematical representation of the manufacturing part, we generate reliable and near-optimal tool paths as well as cutter location data file for post-processing. This algorithm includes two components. First is the forward-step function that determines the maximum distance, called forward step, between two cutter contact (CC) points with a given tolerance. This function is independent of the surface type and is applicable to all continuous parametric surfaces that are twice differentiable. The second component is the side step function which determines the maximum distance, called side step, between two adjacent tool paths with a given scallop height. This algorithm reduces manufacturing and computing time as well as the CC points while keeping the given tolerance and scallop height in the tool paths. Several parts, for which the CC points are generated using the proposed algorithm, are machined using a three-axes milling machine. As part of the validation process, the tool paths generated during machining are analyzed to compare the machined part and the desired part.     

Developing an Empirical Relationship to Predict Tensile Strength of Friction Stir Welded AA2219 Aluminum Alloy

AA2219 aluminum alloy (Al-Cu-Mn alloy) has gathered wide acceptance in the fabrication of lightweight structures requiring a high strength-to-weight ratio and good corrosion resistance. Friction stir welding (FSW) process is an emerging solid state joining process in which the material that is being welded does not melt and recast. This process uses a nonconsumable tool to generate frictional heat in the abutting surfaces. The welding parameters such as tool rotational speed, welding speed, axial force, etc., and tool pin profile play a major role in deciding the joint strength. An attempt has been made to develop an empirical relationship between FSW variables to predict tensile strength of the friction stir welded AA2219 aluminum alloy. To obtain the desired strength, it is essential to have a complete control over the relevant process parameters to maximize the tensile strength on which the quality of a weldment is based. Therefore, it is very important to select and control the welding process parameter for obtaining maximum strength. To achieve this various prediction methods such as response surface method (RSM), analysis of variance (ANOVA), Student’s t-test, coefficient of determination, etc., can be applied to define the desired output variables through developing mathematical models to specify the relationship between the output parameters and input variables. Four factors, five levels central composite design have been used to minimize number of experimental conditions. The developed mathematical relationship can be effectively used to predict the tensile strength of FSW joints of AA2219 aluminum alloy at 95% confidence level.     

Delayed-reference control (DRC) applied to machining operations

This paper deals with a non-time based control for machining operations. Such a control can be implemented in most of the CNC centers in a practical way. Basically, in this control, which belongs to the category of non-time based controllers, the desired input reference is a function of the time and of a variable which plays the role of a time delay: x_d(t−T). Indeed, the proposed controller is called delayed-reference control (DRC). It is recalled that in a time-based control, the desired input reference is described as a function of time only, which is referred to as reference time: x_d(t). The calculation of this function is usually implemented off-line by means of the so called path-planning process. According to this traditional approach, during the task execution, at each instant, the control module is required to track an input reference. What is relevant is that such a reference can never be modified by any event or circumstance. The DRC control differs from time-based controllers in the sense that time delay is properly calculated on-line according to the measured force signals in such a way to improve the interaction during the cutting process. In fact, the DRC consists in an outer force feedback loop around an inner position feedback loop. The effectiveness of the controller has been proven by means of its implementation on a three axes CNC center for wood machining. Eventually, experimental results are presented.     

Study of ductile fracture and preform design of upsetting process using adaptive network fuzzy inference system

This paper combines adaptive-network-based inference system (ANFIS) and elasto-plastic finite element to predict the ductile fracture initiation and the preform shape of the upsetting process. From the hybrid-learning algorithm in ANFIS, it can efficiently construct rule database and optimal distribution of membership function to solve the punch stroke which causes the ductile fracture, and the preform shape which results a desired cylindrical workpiece after forming in the upsetting process. As a verification of this system, the punch stroke for ductile fracture initiation and the free boundary radius of the billet after forming are compared between ANFIS and FEM simulated results. In the ductile fracture prediction, it is proved that ANFIS can efficiently predict the ductile fracture initiation successfully for arbitrary friction coefficient and aspect ratio. In the preform shape prediction, the simulated cylindrical radius shows good coincidence with the desired radius after forming. From this forward and inverse investigation, the ANFIS is proved to supply a useful optimal soft computing approach in the forming category.     

Pre-compensation of servo contour errors using a model predictive control framework

Methods for pre-compensating contour errors in servo systems by adding components of the predicted contour error to the reference position command have recently been proposed in the literature. Such methods are very effective when the curvatures of the desired path are small but their performance degrades at locations of sharp curvature because they lack look-ahead capabilities. This paper presents an improved method for pre-compensating contour errors in servo systems by modifying reference position commands using a model predictive control framework. The pre-compensation value at any given location along the desired path is defined as a weighted average of contour errors within a prediction horizon, and the weights are selected to minimize the sum of squares of the estimated contour errors over the chosen prediction horizon. Constraint enforcement functionalities are also built into the proposed method to ensure that the pre-compensated reference commands stay within specified velocity and acceleration limits. Simulations and experiments are used to compare the performance of the proposed method to a recently proposed pre-compensation approach which lacks look-ahead and constraint enforcement capabilities. Significant improvements in contouring accuracy over the existing method are demonstrated.     

Development of a reference cutting force model for rough milling feedrate scheduling using FEM analysis

Recently developed feedrate scheduling systems regulate cutting forces at the desired level by changing the feedrate to reduce the machining time and to avoid undesirable situations. For effective scheduling, an optimized criterion is required to adjust the feedrate. In this study, a method to obtain the most appropriate reference cutting force for rough milling was developed. The reference cutting force was determined by considering the transverse rupture strength of the tool material and the area of the rupture surface. A finite element method analysis was performed to accurately calculate the area of the rupture surface. Using the analyzed results, the effect of various cutting parameters on the chipping phenomenon was determined. The calculation method for the reference cutting force considered the area of the rupture surface, the effect of the rake angle, and the axial depth of the cut. The reference cutting force calculated using the developed model was applied to feedrate scheduling for pocket machining. The experimental results clearly show that the reference cutting force obtained from the proposed method met the desired constraints that guarantee higher productivity without tool failure.     

基于快速蜜蜂试验法的304不锈钢激光焊工艺优化

基于工程师打样中确定合理工艺参数组合的工程背景,借鉴三元蜜蜂算法开发了用于工艺参数寻优的快速蜜蜂试验法.针对304不锈钢板激光焊中获得较高抗拉强度需求,比较了响应面法、基于响应面法拟合方程的蜜蜂算法优化求解和快速蜜蜂试验法3种方法下所获优化结果.结果表明,前两种方法所预测最优工艺参数获得抗拉强度均近似于688 MPa,而快速蜜蜂试验法求解的最优工艺参数获得抗拉强度734 MPa.快速蜜蜂试验法划分清晰及针对性强的优化特点,使其相比响应面法减少试验次数的同时避免了忽略不同区域中概率密度不同的问题.所开发的快速蜜蜂试验法能够作为一种学习成本低与重视结果突出试验点的试验方法,帮助工程师在打样过程中快速获得更优工艺参数组合.     

Linear analysis of texture–property relationships using process-based representations of Rodrigues space

Techniques for exploring texture–property–process relationships in various deformation processes are derived from a reduced-order model of texture evolution. The orientation distribution function (ODF) in polycrystals is represented over the Rodrigues space in a discrete form using finite element interpolation techniques. Linear programming algorithms are developed for retrieving ODFs with extremal or optimal properties from the complete ODF space. The relationship of optimal textures with processing is addressed by representing texture evolution in a space of reduced basis coefficients called the process plane. This involves generation of orthogonal basis functions for representing spatial variations of the ODF in a given process using proper orthogonal decomposition. These basis functions are found to work in interpolatory and extrapolatory processing modes and allow representation of texturing for deviations in the process variables. Optimization problems are posed in the reduced space for retrieving textures with desired properties. A graphical technique is discussed that allows identification of optimal processing paths for reaching desired textures in association with process plane databases.     

Three-dimensional finite element mesh generation from measured data of a stamping die

An efficient approach was proposed to generate a three-dimensional (3D) finite element mesh system from the measured data of a stamping die, and a computer software was also developed based on the proposed approach in the present study. The concept of a checker with regular grids was used as a template to sort the tremendous number of measured data into a regular pattern with fewer data. The Bezier and B-spline surfaces were then adopted to smooth the data to implicitly represent the stamping die surface, followed by a special procedure to construct the mesh from the treated data. In the developed software, the desired numbers of nodes and elements, aspect ratio, the maximum allowable curvature, and the allowable error can be specified to obtain an optimal mesh system from the measured data. The validity of the proposed approach and the developed software was demonstrated by three examples. The developed software will be of help to the designers of stamping dies.     

优化排样技术在专用数控机床中的应用研究

针对航空工业隔热橡胶实际生产下料问题，运用运筹学理论，建立了矩形件优化排样数学模型，并提出了一种矩形件优化排样算法，在此基础上开发了优化排样程序，能够自动快速的生成优化排样下料图，并将排样结果生成DXF文件输送到数控系统，数控系统根据所得到的图形信息进行加工。经实践证明，该数控系统能够有效地提高工件的加工效率和加工质量。