| The passive vibration control is widely used in many areas and it can reject the vibration of high-frequency well. But the rejection of low-frequency vibration is not ideal. Active vibration control can attenuate low-frequency disturbances significantly, which overcomes the disadvantage of passive vibration control. Active vibration control can be seen as a special problem of disturbance attenuation. It is viable to deal the active vibration control problem with the method of disturbance rejection. Many feedback methods use the state (or the estimation of state), and the estimation of state increases the order of the system. Acceleration signal (continuous systems:state-derivative, discrete systems:state-difference) of vibration system can be directly obtained from the accelerometer, which is the state-derivative of the system. Acceleration response are widely used to evaluate the level of vibration in vibration system. What is more, acceleration signal can be used directly in feedback without designing the state observer. A common class of low-frequency perturbations in vibration system can be abstracted as a sinusoidal signal or the mix of several sinusoidal signals. These disturbances will affect the performance of vibration system or even damage the system if they are not rejected efficiency.Considering the two features of vibration system, one of the work of this study is about the estimation and compensation for sinusoidal disturbances, the other study is to research the vibration control law based on state-derivative feedback. The simulation on Maglev Isolation System verified the theory of this paper.In order to observe a class of disturbance which can be expressed as power series, this article considers the non-linear disturbance-observation (NDOBC) based control, by selecting the non-linear weighting function to improve the speed and accuracy of the disturbance observer. Then the state-derivative feedback controller with disturbance compensator is designed based on the estimation of disturbance.For the interference with unknown frequency and amplitude, we can obtain the accelerate signal by reasonable assumptions, then propose the adaptive asymptotically disturbances observer. Then the state-derivative feedback controller with disturbance compensator is designed. And the algorithm proposed in this thesis is verified by the simulation of Maglev Isolation System.In literature, all of the algorithms based on state-derivative are proposed in continuous system. There are no algorithms of state-derivative feedback pole placement in discrete systems. So, the theorems and algorithms of pole placement by state-difference in discrete system have been provided. This is a good complement to the continuous-time domain algorithm. We first transfer the general system to the canonical form, then discuss the pole placement method of canonical form, at last the algorithm of the general system is obtained by a linear transformation. Finally, simulation of the examples and the experiment of Maglev Isolation System demonstrate the effectiveness of the algorithm. |