Study On Electromagnetic Actuation And Multi-physics Coupled Optimization Of High Bandwidth And Long Stroke Fine Pointing Mechanism

Fine Pointing Mechanism(FPM)is a critical component of beam Acquisition,Pointing,and Tracking(APT)subsystem for optical communication system,and is used to address the high-precision pointing task between communication terminals.Generally,FPMs are driven by voice coil actuators or piezoelectric actuators.However,limited by inherent drawbacks of these two kinds of actuators,the FPMs using voice coil actuators achieve longer strokes but lower bandwidths whereas the piezoelectrically driven FPMs allow higher bandwidths but short strokes.The ideal alternative to FPM actuator should combine the advantages of both voice coil actuator and piezoelectric actuator,which allows to achieve an FPM with both high bandwidth and long stroke.Linearized normal-stress electromagnetic actuator(LNSEA)maintains high force output ability as well as NSEA but with its nonlinearity greatly improved,and its stroke is longer than piezoelectric actuator's stroke.Hence,LNSEA is ideal to actuate FPM.However,using LNSEA to actuate FPM remains many key theoretical issues to address,and needs further developed.This dissertation addresses the key theoretical issues in using LNSEA to actuate High-Bandwidth and Long-Stroke Fine Pointing Mechanism(HLFPM),including topology research and modeling of LNSEA,matching of LNSEA and flexure support system,multi-physics coupled optimization of HLFPM and so on.It mainly aims to improve stroke and bandwidth of the HLFPM.The main works and achievements are listed as follows.Firstly,a new LNSEA named LNSEA-RTM which overcomes the drawbacks of the published LNSEA actuated FPMs is developed to drive the HLFPM,and an advanced linear model is proposed for the LNSEA-RTM.The LNSEA-RTM will not suffer from additional axial force problem,and has a compact structure and an armature with small rotational inertia.Based on Modified Nodal Approach,an controlled condition for simplifying the electromagnetic analysis of the LNSEA-RTM is obtained.With both coil leakage and permanent magnet leakage involved in the linear modeling,accuracy of the LNSEA-RTM's linear model is far more improved compared with the published linear model of the LNSEAs driving FPMs,which is validated by finite element simulations.Secondly,a nonlinear modeling method is proposed to model LNSEA,and a nonlinear model for the LNSEA-RTM,which overcomes the drawback that traditional linear model is unable to describe LNSEA's nonlinearity which occurs when displacement is large,is proposed.Before the nonlinear modeling,the dissertation at first proves that simplifying electromagnetic analysis and calculating air gap reluctance with linear method will not cause LNSEA's nonlinearity.Then,a fringing loss model of the LNSEA-RTM is presented and a nonlinear magnetic circuit is developed.After that,a Neural Network model for nonlinear permanent magnet leakage is proposed by using discrete data fitting idea.On this basis,a nonlinear model for the LNSEA-RTM is proposed.The obtained nonlinear model is certificated by both simulation and experiment.The research results indicate that the LNSEA-RTM's nonlinearity is caused by the variation of fringing loss and permanent magnet leakage with rotation angle.The nonlinear model is the basis of improving the HLFPM's stroke and maintaining its linearity simultaneously.Thirdly,a flexure support system is designed for the HLFPM based on the output characteristics of the LNSEA-RTM and performance requirements of the HLFPM.Then,the flexure support system is modeled by using analytical method and discrete data fitting method.Designing a flexure support system whose supporting characteristics matches with the output characteristics of the LNSEA-RTM can avoid decrease of the HLFPM's linearity which is caused by the LNSEA-RTM's nonlinearity,so that the HLFPM's stroke is expanded from the LNSEA-RTM's linear working domain to its nonlinear working domain.And hence,the HLFPM's stroke is improved and the HLFPM's linearity is guaranteed simultaneously.Fourthly,with the modeling results of the flexure support system and the linear and nonlinear models of the LNSEA-RTM,a multi-physics and multi-objective coupled optimization model which considers the stability,linearity,bandwidth,scanning stroke and fatigue life of the HLFPM is developed.To improve the HLFPM's stability,the pull-in mechanism is involved in the optimal modeling.Besides,the HLFPM's linearity is seen as an optimization objective which guarantees the HLFPM's linearity.To solve the HLFPM complicated optimization model,a heuristic searching algorithm driven by constraints classification is developed.Combing the algorithm with Genetic Algorithm,the optimization model is solved.The optimization result shows that the HLFMS's rotation angle varies linearly with actuating current.Designing a proper topology,the flexure support system's nonlinearity can cancel the LNSEA-RTM's nonlinearity through multi-physics coupled optimization,so that the HLFPM's stroke is expanded from the LNSEA-RTM's linear working domain to its nonlinear working domain.And hence,the HLFPM's stroke is improved and the HLFPM's linearity is guaranteed simultaneously.Finally,to validate the research works of the dissertation,control algorithm for the HLFPM is developed and the experimental performances of the HLFPM are analyzed.To obtain the identified model of the HLFPM system via system identification approach,the open-loop frequency response of the HLFPM system is measured.From the open-loop frequency response,it is found that the HLFPM system has a significant resonance peak.To suppress the resonance,a notch filter is designed.After that,designing high-bandwidth control algorithm for the HLFPM is performed by using series compensation method.Then,scanning stroke and bandwidth of the HLFPM are measured and are compared with that of some typical FPMs.The test and comparing results shows the HLFPM can achieve a small signal tracking bandwidth of 2.4 kHz and a scanning stroke of 20 mrad.The HLFPM's scanning stroke is 2.9 times of that of the published LNSEA actuated FPM which is designed by MIT,and more than 5 times of that of typical piezoelectric actuators actuated FPMs.The HLFPM's bandwidth is higher than that of typical voice coil actuators actuated FPMs and equals even exceeds that of piezoelectric actuators actuated FPMs.The test and comparing results validate the research works of the dissertation and indicate that the LNSEA-RTM is ideal to actuate the HLFPM.