Modeling, Identification And Control Technology Research On Electric Shaking Table

As a kind of vibration equipment, the shaking table has been widely used in aerospace, vehicle transportation, building structure, industrial automation and many other engineering fields. It’s mainly used to achieve the characteristic information such as structure strength, reliability and stability of the mechanical and electronic parts or assemblies. The electric shaking table is favored for its advantages of wide frequency band and waveform easy to control. The main technical indicator of shaking table is the high-precision repetition for the desired periodic vibration output force signal. For the research of this problem, this paper proposes a control strategy based on the three parameter controller to achieve high-precision vibration output force tracking control for electric shaking table driven by AC servo motor. The research in this paper is supported by the Guangdong Province Strategic Emerging Industries Projects(No.2011A091101001, Industrial robot core technology research and industrialization of typical products).Based on the dynamic model of the electric shaking table driven by AC servo motor and experiment platform for model parameter identification, the recursive least square is used to obtain the dynamic model parameters. Then, the output force tracking control strategy is proposed. This strategy contains feedforward and feedback control of the output force, displacement, velocity. Besides, the real-coded genetic algorithm and adaptive genetic algorithm is used to optimize the three parameters of the feedforward controller gain. The comparison and simulation with the result is carried out, and the results prove that the convergence rate of adaptive genetic algorithm is faster, and the control effect of controller gain is more ideal.The Kalman filter is adopted to estimate the speed of the three parameters feedback signal, to overcome the speed signal quantization noise brought by the direct displacement difference. Based on the kinematics equations of acceleration and displacement signals, the discrete linear Kalman filter model is established for the speed estimation. The stability of the model is proved, and the variance of process noise and observation noise are derived and calculated. The method of steady state Kalman filter gain is presented instead of time-varying gain, so as to improve the running efficiency of the program.An electric vibration experiment device is designed to the study of vibration output force tracking control, and the controller is embedded computer CX9020 made by German Beckhoff company. The real-time module is developed by international standard IEC61131-3 programmable controller language, and the minimum sampling period of the control system is up to 50 us. A linear displacement sensor and acceleration sensor as the sensing module are used to the displacement and acceleration signal acquisition of the vibration platform. The experiment research on speed estimation is conducted, and the results prove that Kalman filter can effectively eliminate the quantization noise. At the same time, the experiment research of vibration output force tracking control based on the three parameter controller is carried out, and the experimental results testify the correctness and effectiveness of the above theory and algorithm.