Study On Model Predictive Control Technology Of The Dynamic Load In Resonance Material Fatigue Tester

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.