【作者】 云艳；

【导师】 高红俐；

【作者基本信息】 浙江工业大学 ， 机械电子工程， 2012， 硕士

【摘要】 疲劳试验即测出各种结构材料在不同的交变载荷作用下的疲劳寿命,对于结构设计和产品可靠性保证具有十分重要的意义。谐振式高频疲劳试验机是基于共振原理的用于测定金属材料疲劳特性的试验设备,因其工作频率高、能量消耗低、试验时间短等优点被广泛地应用在高等院校科研和教学及工业生产等各部门。目前,高频疲劳试验系统存在的主要问题是因试验过程中试样刚度随裂纹扩展而逐渐减小,造成系统共振频率点的改变从而影响了工作台的共振振幅,因此需要通过调节试验机的工作载荷来实现系统的恒幅控制。基于高频疲劳试验系统的时变性,论文提出了模型预测控制在高频疲劳试验机控制系统中的应用实现研究,通过对被控对象的精确建模分析了系统的动态响应特性,以此为依据对试验机的动态载荷进行了模型预测控制。主要研究工作和取得成果如下:(1)建立了被控对象的双自由度线性定常及时变两个振动系统力学模型,通过理论分析计算得出了线性定常系统固有频率和共振振幅的表达式,并分析了试样刚度对两者的影响;将裂纹扩展试验得到的裂纹与时间的关系代入试样刚度模型,根据线性时变系统模型仿真得到了系统固有频率和共振振幅随时间的变化趋势。(2)依据预测控制中的动态矩阵控制(DMC)算法建立了谐振式高频疲劳试验机载荷控制系统的预测模型,根据仿真得到的系统试验过程不同阶段的振幅阶跃响应曲线,用MATLAB7.0分阶段设计了预测控制器并总结了控制参数的选取对控制器性能的影响。(3)开发了基于LabVIEW8.5的虚拟仪器实验平台,设计了控制系统的应用软件,实现了基于动态矩阵控制算法的预测控制器的在线应用,由实验结果对控制器的性能进行了评估。论文以电磁谐振式高频疲劳试验系统的动态特性为依据设计了预测控制器,并通过仿真分析和实验验证体现了预测控制的优越性。本文的创新点为首次将模型预测控制技术运用到高频疲劳试验机载荷控制系统中,并通过MATLAB7.0和LabVIEW8.5联合编程实现了动态矩阵控制算法的实际应用,方便了对被控对象输入输出的监测控制,提高了程序的运行效率。更多还原

【Abstract】 Fatigue test is used to measure the fatigue life of various structural materials under alternating load. It has great significance for structural design and assurance of product reliability. Resonant frequency fatigue testing machine is used to determine the fatigue characteristics of metallic materials based on the resonance principle. It has been widely used in research and teaching institutions of higher learning and industrial production sectors because of its high frequency, low energy consumption and short test time . Currently, the main problem of high-frequency fatigue test system is specimen stiffness gradually decreases with the crack during the test, which affects the system resonance amplitude due to the change of system resonance frequency. So this paper presents the Model Predictive Control applied in high-frequency fatigue test machine control system for its time-varying and designed the predictive controller through the accurate modeling of the controlled object and the analysis of system dynamic response characteristics. The main researches and achievements of this paper are listed as following:(1) The two degrees of freedom linear time-invariant and time-varying mechanical vibration system model of the controlled object were established, the system natural frequency and resonance amplitude were calculated based on theoretical analysis of linear time-invariant system work principle. The impact of both along the sample stiffness change was analyzed. The system natural frequency and resonance amplitude versus time relationship were simulated based on the relationship between the crack and time got from crack propagation test in time-varying system.(2) The predictive model of resonant frequency fatigue testing machine load control system was established based on of Dynamic Matrix Control algorithm of predictive control. Predictive controller in stages was designed with MATLAB7.0 based on system amplitude step response curves during the test at different stages got by simulation with MATLAB7.0. The selection of control parameters affecting the performance of the controller was summarized.(3) Experimental platform based on virtual instrument was developed with LabVIEW8.5. The control system application software was designed in order to achieve online application of the predictive controller based on DMC algorithm. The prediction model was amended and the controller performance was evaluated according to results.The core achievement of this paper is designing the predictive controller based on dynamic characteristics of high-frequency fatigue testing system, and showing the advantages of predictive control via simulation and experimental results. Innovation of this paper is the first application in dynamic load control system of high frequency fatigue tester with MPC technology and hybrid programming through MATLAB7.0 and LabVIEW8.5. The input and output of the controlled object are easily monitored and the operating efficiency of the program is improved meanwhile.更多还原