No Model Fuzzy Control Method Study Electro-hydraulic Vibration Table

Electro-hydraulic shaker force control system has the advantages of highpower quality，large output, etc.. It can track a given excitation waveform anddetect accurately fatigue life of the object, which have been widely used in thevibration test. High-frequency vibration can not only greatly shorten the testcycle, but also save a lot of manpower and material resourse. Andelectro-hydraulic force control servo system is a non-linear and it has thetime-varying parameters and external load disturbance, which makes the systemuncertain, so it is difficult to establish an precise mathematical model in practice.And the conventional PID controller is difficult to obtain good control effect.Therefore, a suitable control method should be seeked to reduce the influence ofuncertain parameters and increasing the bandwidth of systems, which has theimportant meaning.Model-Free Adaptive Control（MFAC）does not depend on mathematicalmodel of the system, For nonlinear, uncertainty and time delay of the system,MFAC has good control performance and adopts the method of dynamiclinearization to design model-free adaptive control law. Although itsconvergence rate is slower，control accuracy is higher. MFAC has betteradaptability and robustness in practice and widely used in the uncertain system.For force tracking control problem of electro-hydraulic shaker force controlsystem，a fuzzy MFAC control is proposed to reduce the influence of uncertainparameter. Firstly，based on the strong adaptability of MFAC in large inertia anduncertainty system，fuzzy control is introduced to solve the problem that MFAChas the slow tracking speed, combining the advantages of two control methodsto improve control performance of the system. Secondly, a simulation research ismade to the electric hydraulic servo control system from different perspective byusing Simulink with AMESim, The results show that compared with basicMFAC, the Fuzzy-MFAC has faster response speed, and it can eliminate theoutput error of the specimen, improve the accuracy of the output power andmake a better adaptability for the system. Finally, it is proved that the control method is feasible in engineering through the semi-physical real-time simulationexperiments of Dspace.