In order to better feedback the dynamic displacement information of the heavy-load hydraulic cylinder of the test-bed in the performance test of the large-scale damping damper,so as to truly simulate the vibration excitation signal required by the large-scale damping damper,and then to obtain better characteristic parameters,comparing with the traditional variable step size LMS algorithm,according to the actual vibration loading condition of the heavy-load hydraulic cylinder of the test-bed,under the condition that the expected reference signal was unknown,the original signal was fitted by wavelet transform,and the nonlinear function expression between error and step size was established by MATLAB,and a LMS adaptive filter based on wavelet transform was proposed.Finally,a measurement and control program was written by LabVIEW.The experimental results show that,compared with the traditional variable step size LMS algorithm,the adaptive filter based on wavelet transform has faster convergence speed,smaller steady-state error and stronger tracking ability.In practical application,it has better filtering effect than other types of filters.