Study On Active Control Theory And Technology Of Self-Excited Vibration In Milling

Thin-walled workpieces are widely used in national defense,space exploration,energy engineering,etc.They have a low stiffness,strong coupling,easy to produce deformation and vibration characteristics in the milling process,which limit the productive capacity of machine tools,and lead to inferior workpieces and high processing cost.In order to suppress the occurrence of vibration,it is necessary to study the mechanism and characteristics of the vibration,and then design an effective active controller.In order to solve the deformation problem of the thin-walled parts,it is necessary to carry out a lot of experimental research and analysis,and then design an efficient compensation scheme.This dissertation focuses on deformation compensation and developing active control techniques for self-excited vi-bration mitigation to enhance manufacturing efficiency.The main research contents and contributions are as follows.Aiming at the nonlinear milling machining dynamics system with input constraints,an active model predictive control method is developed to mitigate the machining vibration for milling process with input constraints.This method extends the closed-loop stability region of the system.In particular,the self-excited vibration system is transformed into a linear time-varying system with input constraints by finite Fourier series approximation and dis-cretization of the perturbation system.Then,the linear polyhedron technique is combined with the rolling optimization method to solve the optimal state-feed back control gain.The method is applied to the milling self-excited vibration suppression process with input satura-tion,which realizes the maximum of the stable region in the stable lobe diagram composed of "spindle speed-axial cutting depth",and improve the maximal material removal rate and the processing efficiency.Based on the periodic time-varying characteristics of the dynamical system itself,the periodic dynamic cutting force in the milling process can be approximated by Fourier series,and the reliance on accurate measurement of the cutting cycle is removed.The mitigation of self-excited vibration often requires low computational complexity and large active con-trol force in the high spindle speed and large axial cutting depth range.So,we design a semi-global input saturation controller by constructing monotonic nondecreasing function and backstepping method,and establish a closed-loop system.To guarantee the feasibility of the proposed approach,moderate stable conditions of the closed-loop system are after-wards derived by LaSalle-Yoshizawa theorem as well.Thus,the method can be expected to improve the efficiency of milling processes.There is often a low radial immersion machining in the process of milling,i.e.the radial cutting thickness is less than the milling cutter diameter,and the amplitude of the periodic dynamic cutting force varies greatly,so the method that the periodic cutting force variation matrix is covered by an uncertainty set is proposed.In order to ensure the stability of the thin-walled workpiece milling process within the specified spindle speed and axial cutting depth range,the time-delay uncertainty of the self-excited vibration system is covered by the Pade approximation and the uncertainty set.Thus,a linear system with structure uncertainty is established,and then the robust active control problem with uncertainty sets is solved by using a D-K iteration ? synthesis method.The simulation results show that the proposed control strategy substantially enlarges the stable region in the stability lobe diagram,and the milling efficiency and quality can be improved effectively.Aiming at deformation problem that is often encountered in the real production of thin-walled workpieces,a purely data-driven Sparse Bayesian learning-based method is develoed to predict and compensation the deformation.In particular,the marginal maximum likeli-hood estimation based on the sparse Bayesian function is used to estimate the short-term prediction deformation,and then the z-direction displacement of the milling spindle is com-pensated by the z-axis motor of the spindle,the thin-walled workpieces also achieve uniform milling.Then,a dual-mode predictive controller is designed to effectively enlarge the re-gion of initial state attraction of the vibration system enlarge with input constraint,which can better suppress the external disturbance and has high practical value.We also built a thin-walled workpieces milling deformation compensation and vibration active controller platform,the experimental results show that the closed-loop system based on the proposed Bayesian learning and MPC improves the surface quality of the thin-walled workpiece ma-chining.In order to study the effect of the active control algorithm based on the magnetic bearing on the self-excited vibration mitigation of the milling process,this paper first introduces the five-axis CNC milling machining center prototype with active magnetic bearing milling spindle.The maximum spindle speed of the prototype is 40000rpm and the rated power is 20kW.Since the customized active magnetic bearing milling spindle is an open-loop unstable system,the axial and radial levitation experiments of the electromagnetic spindle are investigated,the corresponding analysis and improvement proposals are put forward,which provides the equipment and technical basis for the active suppression of the self-excited vibration by using the active magnetic bearing in the prototype machine.Finally,it is the conclusions of the dissertation and the prospects for further research work and development direction of the suppression of self-excited vibration and the deformation of thin-walled workpiece.