Study Of Slide Carriage Active Vibration Control Of Ultraprecision Machine Tool

With the development of advanced technology, the machining accuracy of parts is needed higher and higher in every field, especially national defence, aviation and space fields. Vibration is one of the key factors affecting the machining quality in precision and ultra-precision machining that influences national defence remarkably. In order to enhance the motion accuracy, aerostatic guideway is usually used in ultra-precision machine tool. But the vibration of slide carriage worsens the surface of workpiece directly through tool post. Therefore, the study of effective elimination of vibration is very important for enhancing machining accuracy. The thesis analyzes the vibration of slide carriage supported by air film, establishes it's mathematical model of two-degree-of-freedom active vibration control system which active air bearing is used as actuator, discusses output feedback control and state feedback control. Through the design of feedback gain matrix, the system can be uncoupled and the anticipant characteristics can be obtained. When considering the characteristics of actuator, the mechanical and mathematic model of system would be changed and the characteristic of system would be different somewhat. In spite of the mass of moving part of actuator can be neglected, stiffness coefficient and damping coefficient of actuator air film cannot be ignored due to correspondence to that of guideway air film. Active vibration control system of slide carriage is set up. The characteristic of slide carriage is analyzed, the configuration of vibration control system discussed, noise eliminating using wavelet analysis studied and the effect of delays in the feedback loop analyzed. For carrying out transverse vibration control in the process of slide carriage movement, active noncontact air-actuator based on PZT is developed. Adopting air bearing not only obtains noncontact but also reduces the influence of manufacturing error of parts due to average effect of air film. Unitary structure of elastic hinge ensures that the air-bearing pad can redraw in time to obtain enough frequency response. Twenty nanometers order micro-displacement control is obtained using PZT. Through the analysis of active vibration control based on air-actuator, it is pointed out the characteristics of system is affected not only by the object controlled, but also by the structure of actuator, the characteristics of air bearing film and the characteristics of PZT. In the control strategy of slide vibration, the PID controller is used as main controller, which combined with fuzzy control and neural network control is discussed. Fuzzy controller is a developed recently controller, which advantage is that mastering precise mathematical model of the controlled object is not needed. Two-dimension fuzzy controller is most widely used, but it has not satisfactory stable characteristics because of the lack of fuzzy integral control function. If PID control and fuzzy control work together, both good flexibility and adaptability of fuzzy control and high precision of PID control can be obtained. The parameters' adjustment is the key of PID controller design. For the traditional PID controller, the parameters are always unchanged in the control process after being confirmed. It is not suitable to complicated time-varying system. Artificial neural network is a new interdisciplinary subject developing rapidly in recent years. With the characteristics of approaching any nonlinear function, neural network can carry out good combination of PID control parameters on line through the learning of performance of the system. Finally, many experimental studies are carried out on the developed system. The results show the feasibility of active vibration control based on air-bearing actuators and the effectiveness of fuzzy-PID control and NN-PID control. From the comparison of experimental results, it is showed that NN-PID control is better than PID control and fuzzy-PID control. This is mainly because neural network can adjust and carry out good combination of PID control parameters on line through the learning of performance of the system.