Adaptive Control of Hysteretic Smart Material Systems

Smart Materials;Hysteresis;Adaptive Control

Fan, Xiang ; Ralph Smith, Committee Chair,Fan, Xiang ; Ralph Smith ; Committee Chair

University:North Carolina State University

关键词: Smart Materials;    Hysteresis;    Adaptive Control;   

Others  :  https://repository.lib.ncsu.edu/bitstream/handle/1840.16/4648/etd.pdf?sequence=1&isAllowed=y

美国|英语

来源: null

【 摘 要 】

Smart materials exhibit nonlinearities and hysteresis when driven atfield levels necessary to meet stringent performance criteria inhigh performance applications. This requires models and controldesigns that effectively compensate for the nonlinear, hystereticfield-coupled material behavior. In this dissertation, weinvestigate model identification using the homogenized energy modeland adaptive control of hysteresis in smart hysteretic system, whilethe approaches are applicable to control of a wide class offerroelectric, ferromagnetic and ferroelastic materials, weillustrate the ideas through the example of controlling aferroelectric actuator.We pursue the problem of hysteresis control through twocomplimentary approaches: linear adaptive control using an inversecompensator and nonlinear adaptive control.Inverse control is a fundamental approach to accommodate hysteresiseffect by constructing a right inverse of the hysteresis. Due to theopen-loop nature of inverse control, the performance of the inversecompensation is susceptible to model uncertainties and to errorintroduced by inexact inverse algorithms. The objective of adaptivecontrol is to design a controller that can adjust its behavior totolerate uncertainties or time-varying parameters. We employ thehomogenized energy model to quantify the hysteresis. On the basis ofthe hysteresis model, we propose an adaptive control framework bycombining inverse compensation with adaptive control techniques, andinvestigate the parameter identification methods for the hysteresismodel. We prove the asymptotic tracking property of the proposedadaptive inverse control algorithm, discuss the issue of parametersconvergence and illustrate the performance of the proposed controlmethod through simulations.Adaptive nonlinear control is a more challenging task and hasreceived increasing attention in recent years. The challengeaddressed here is how to fuse hysteresis models with availableadaptive control techniques to have the basic requirement ofstability of the system. In this dissertation, an adaptive variablestructure control approach, serving as an illustration, is fusedwith the homogenized energy model without constructing a hysteresisinverse. The global stability of the system and tracking a desiretrajectory to a certain precision are achieved under certainconditions. Simulations are performed on an unstable nonlinearsystem. The purpose of exploring new avenues to fuse the model ofhysteresis nonlinearities with the available adaptive controllerdesigns without constructing a hysteresis inverse is achieved andillustrated with the promising simulation results. This provides astep toward the development of a general nonlinear adaptive controlframework for hysteretic systems.