Research On Theory And Experiment Of The Novel Active Vibration Isolation System

Active vibration control is to suppress vibration through driving actuator to the structure by certain control strategy based on structure vibration information detected. Active control technology has been a important novel vibration control method because of potential advantage of effective result and good adaptability, which may be used in aerospace, civil engineering, high speed vehicle isolation and others machine equipment vibration control. Piezoelectric materials suit active vibration control because of many unique advantages. So active control technology based on intelligent structures developed quickly and has been widely applied. Under this background, some theories and experiments of active isolation technology are investigated in this paper which a novel active isolation system based on piezoelectric intelligent structures has been studied mainly including mechanism, theory modeling, system identification, vibration control method and experiment validation.The concept of the novel active flexible isolation system is putted forward and the basic theory frame of the system is set up for the first time. The optimization model of the system is established through transfer functions of the system based on the isolation performance, the interior performance and the capability of the piezoelectric actuators.The usual modal of the piezoelectric intelligent structures has been developed. Started with discussion of piezoelectric constitutive equations, the element model of the system is established through Hamilton's principles and variational methods. Then the global equation of the intelligent composite structures is derived from element models. The system dynamic response is calculated through mode superposition method and the frequency response function matrix is gained from the system equation. The active flexible isolation system is modeled through beam element with electric degrees of freedom and space conversion matrix. And the theory of the active flexible isolation system is researched based on the frequency response function matrix and the system dynamic response.The identification of the active flexible isolation system is demonstrated through the frequency response function matrix and the time domain model. Some estimation models of the frequency response function matrix are discussed and some low modes are identified combined with the result from finite element method in the frequency domain, so the MIMO transfer function matrix model is gained. The time domain identification model composed of neural network and delay model is reached through the difference equation gained from the system model in the time domain. The identification has validated theory model and has become to be basis of control algorithms.The control strategies have been studied with the isolation performance, the interior performance and the capability of the piezoelectric for the first time. When disturbance to the system can be measured or be referenced, the adaptive feedforward control is studied and the adaptive feedforward control for MIMO system is developed. When disturbance to the system cannot be measured, the normal H-inf design model is obtained within the description of the system performance used by weighted functions. The neural network inverse model control and general neural network control are established combined with identification model of neural network. When disturbance has low power, simulation results demonstrate feedforward control and feedback strategy can satisfy the synthesis perfomances of the system.A complete set of experiment based on computer control system is designed and developed. The system is composed of the flexible base structure, the circuit board for signal procession, the power amplifier, the electric charge amplifier and the system software edited by MATLAB SIMULINK. And the identification of frequency response function, the neural network identification model, the adaptive feedforward control and the robust feedback control are demonstrated with experiments. The system can reduce low frequencies vibration with the 10dB effect and correspond to be a passive isolation at the high frequencies in the result. The adaptive feedforward control and robust feedback control can satisfy the synthesis request of the system performance and both can isolate vibration effectively.