High-Frequency Feedforward Control And Its Parameters Tuning Approach For Piezo-Actuated Fast Tool Servo System

With the rapid development of science and industrial technology,the application of the non-rotary symmetric parts has become increasingly widespread.However,high-precision and efficiency machining of the non-rotary symmetric parts is a difficult issue in machining field.Piezo-actuated fast tool servo(FTS),which is an enabling and efficient technology for machining of the non-rotary symmetric parts,has been widespread concerned in academic circles and industry.Since the spindle speed and surface topography that can be machined are depended on the high-frequency performance of FTS,the high-frequency performance is the basis for FTS to machine of the non-rotary symmetric parts.However,the high-frequency performance suffers from the inherent rate-dependent hysteresis nonlinearity,which seriously degrades the tracking performance of FTS,and even makes the closed-loop system unstable.Moreover,FTS systems are prone to variations in their system parameters due to changes in external environment,installation method and loading,which seriously decrease its high-frequency performance.Therefore,it is necessary to compensate for the rate-dependent hysteresis and handle model uncertainty in order to achive high-frequency control of FTS.The purpose of this dissertation is to realize high-frequency control of FTS involving with the rate-dependent hysteresis and model uncertainty.The contents and achievements of this dissertation are listed as follows:The comprehensive inverse approach and rate-dependent model approach have been prosposed to realize high-frequency feedforward compensation of rate-dependent hysteresis.In comprehensive inverse approach,feedforward controller is composed by dynamic inversion and rate-dependent hysteresis inversion in which the rate-dependent hysteresis inversion is used to reduce dynamics compensation error and compensate for hysteresis simultaneously.In rate-dependent model approach,a new rate-dependent model based on exponential function is constructed to reduce phase lag caused by incompletely dynamics compensation.Experiment results show that the proposed approaches can effectively compensate for rate-dependent hysteresis to improve the high-frequency performance of piezoactuator system.To reduce the decline of high-frequency performance caused by model uncertainty,an iterative identification based open-loop feedforward tuning approach is proposed.By transforming feedforward controller into parametric Wiener model,the processes of system identification and controller design are translated into control-oriented system identification process.Due to the coupling between hysteresis and dynamics,feedforward tuning is decomposed into dynamic feedforward tuning and hysteresis feedforward tuning.Then an iterative identification algorithm is proposed to realize feedforward parameters tuning of piezoactuator system.Next,the conditions required to ensure algorithm convergence is given.After that,an adaptive feedforward control has been realized by updating controller parameters from input-output data.Experiment results show that,comparing with model-based feedforward,the proposed feedforward can effectively reduce the impact of model uncertainty on high-frequency performance of piezoactuator system.Aiming at realizing feedforward tuning in closed-loop system,an iterative learning identification based closed-loop feedforward tuning approach is proposed.From predicting tracking error of next iteration by current input-output data,the function between dynamics feedforward parameters and tracking error is established.Then an optimization problem can be established by minimize tracking error to estimate dynamics feedforward parameters.After that,hysteresis feedforward parameters can be obtained based on estimated dynamics feedforward parameters.To ensure the stability of tuning process,a new approximate stable inverse is proposed to handle the inversion problem of non-minimum phase system.Experiment results show that the proposed approach can realize feedforward tuning in closed-loop system,and thus achieve high-frequency control of piezoactuator system.Based on the requirements of machining of non-rotary symmetric parts,a large-stroke,high-frequency response FTS system is build.By parameter optimization of flexible hinge,the natural frequency and stroke of FTS reach 459 Hz and 550 ?m,respectively.Then turning experiments are carried out on the FTS system to verify the effectiveness of proposed control algorithm.Compared with the common used high-gain feedback control and P-I model based feedforward-feedback control in FTS machining,by using the proposed control algorithm,the machining contour error of sinusoidal surface is reduced by 69.16% and 25.42%,respectively;the machining contour error of ellipsoid surface is reduced by 92.35% and 69.16%,respectively.