Study On The Methods Of Ultra-Low Frequency Measurement And Active Control Of Precision Vibration Isolation System

The precision vibration isolation system(PVIS)is fundamental guarantee for ultraprecision manufacturing and measurement equipment.With the accuracy of instruments and facilities into(sub-)nanometer level,the PVIS's high-performance vibration isolation,especially ultra-low frequency active vibration isolation technologies are becoming much more stringent.Inertial sensors such as acceleration sensor or geophone are essential to make active control a reality,but suffering limited low-frequency bandwidth and poor signal-tonoise ratio.It sets barriers to the performance improvement of the PVIS.Therefore,a remarkable challenge must be faced during the designs of the ultra-low frequency measurement and control system of the PVIS.Supporting by the national major scientific research project,this dissertation would pay more attention on the performance influences from geophone's the low-frequency noise and dynamic.Relying on the sensor characteristic compensation and control algorithm design,the aim is to solve the contradiction between the requirements of environmental micro-vibration isolation and the ultra-low frequency measurement and control of the PVIS.Thus,the performance of PVIS is improved,meeting the increasing requirements of vibration isolation,and providing an ultra-stable environment for ultra-precision manufacturing and measurement equipment.Stepping by the analysis of sub-components to system-level,the passive vibration isolation mechanism to the system dynamic characteristics of the PVIS is studied,clarifying the control expectation.Since the modelling of active control components,such as geophone and voice coil motor,and giving insight of insufficient low-frequency measurement capability,we present the general principle of zero-pole cancellation for geophone's compensation,and the resulted noise is also considered and modeled.From these results,the compensation way of analog-circuit combined with digital filter design is determined,which provides a valuable theoretical basis for the designs of the subsequent controllers.In the design of the ultra-low frequency feedback controller,it weighs the feedback stability influences of sensor dynamic and noise-induced perturbation,and shows the impacts of sensor's low-frequency dynamic on feedback stability,therefore,the ultra-low frequency bandwidth expansion strategy of the feedback sensor is proposed.The bandwidth cut-off frequency of the geophone is extended from 4.5 Hz to 0.12 Hz,without any losses of the high-frequency performance,however a distortion coming from noise may results in a degraded vibration isolation performance,a multi-sensor fusion-based LQR regulator is developed,which restrains the noise-induced perturbation and also reserves the performance of vibration isolation.While in feedforward controller,the SNR-considered compensation rule of geophone is presented to obtain the 0.3Hz aimed frequency,since that the noise governs the low-frequency isolation performance.According to the proposed rule,the predicting influence of feedforward performance from sensor modelling and parameter identification error is analyzed,and the implementation of the feedforward controller employs a pole-fixed adaptive algorithm.It is also mended with error and reference signal power spectrum shaping technologies,improving the performance of the algorithm and also reducing the impact of low-frequency noise.At last,the frameworks of the software and hardware for the PVIS's prototype were designed and the proposed ultra-low frequency measurement method and control algorithms were evaluated.The experimental results had good consistency with the simulation analysises: the bandwidth-extended sensor maintained accurate amplitude and phase responses,and acceptable noise levels in the ultra-low frequencies;the feedback controller algorithm achieved 40 dB vibration damping at the PVIS's natural frequency,and the feedforward controller further enhanced the vibration attenuation 10 dB at 5Hz;any noise-caused degradation of control performance was not visible at low-frequencies,verifying the effectiveness of the proposed ultra-low frequency measurement and control method.