Identification Of Acceleration Harmonic For An Electro-hydraulic Servo Shaking Table Based On Kalman Filter

Electro-hydraulic servo shaking table is a kind of important equipment for carrying outvibration test. Due to its strong load capacity, big exciting force and large displacementvibration, electro-hydraulic servo shaking table is widely used in engineering field. Researchon electro-hydraulic servo shaking table has great research significance and practical value.Harmonics which contribute to the phenomenon of distortion in acceleration response signalof the sine vibration test is studied to find out the way to build identification system based onthe Kalman filter.Through an overview of research on the shaking table and identification method, it isfound that harmonic identification mainly concentrate in the field of power system, andharmonic identification for electro-hydraulic servo shaking table is quite rare. At first, thepower mechanism characteristic of the electro-hydraulic servo shaking table is analyzed, andthe three-variable controller for shaking table is designed. Then sine vibration test is carriedout on the shaking table real-time test system based on xPC Target, during which the test datais obtained successfully. Because of the serious distortion shown from the response signal ofthe test, the total harmonic distortion degree (THD) is adopted as the measurement of theshaking table acceleration harmonic distortion. In order to cancel harmonic components,reduce the signal distortion degree, and improve the shaking table performance, it needs toknow about the harmonics information.On the basis of analyzing the spectrum characteristics of the acceleration response signal,it is assumed that the noise-free signal is a direct sum of each harmonic. Thus, three differentstate-space models for the acceleration response signal are constructed. After combining thethree state-space models with the standard Kalman filter, extended Kalman filter andunscented Kalman filter respectively, four kinds of acceleration response signal harmonicidentification system are built. By identifying a constructed signal, the validity of theseidentification systems are all confirmed.Then the harmonic identification experiment is carried out on the identification systemsusing the acceleration response signal data obtained from the sine vibration test. Throughanalyzing the result of harmonic identification experiment comparatively, the performance ofeach identification system is discussed. The optimal identification system with highidentification speed and great identification precision is obtained, which also have theadvantage of less computational complexity and better real-time performance, and the output of the system could be used in harmonic cancellation directly without reconstruction.