Nonlinear control of electrostatic MEMS with applications to optical systems

由于该领域的多学科性质，MEMS的发展需要在许多科学和工程领域进行共同努力。该计划将专注于开发高性能和高功能MEMS应用的控制策略，特别是静电驱动设备的控制及其在自适应光学（AO）和扫描微镜中的应用。 由于系统固有的非线性、不确定性、时变特性和干扰，MEMS的控制面临着严峻的挑战。解决这样的控制问题是至关重要的未来的进步和MEMS技术的新应用的出现。在我们以前的工作中，非线性控制技术已被引入到MEMS控制系统的设计。通过具体的应用表明，寻址系统的性能可以大大提高。 在这个项目中，我们将继续研究单个微器件的非线性控制，但强调输出反馈控制方案，闭环系统分析和避免与某些特定点的不可控性相关的奇异性的算法。本论文的研究工作将集中在互连MEMS阵列的控制上，这是典型的自适应光学系统的模型。空间分布系统的分布式控制的技术将被采用和系统架构，控制器合成，最优性，计算效率，控制任务调度和同步等问题将得到解决。 开发的系统将使用软件模拟进行测试。为此，将解决更现实的建模和软件包ANSYS，CoventorWare和COMSOL（以前FEMLAB）将用于模拟MEMS器件。硬件在环仿真和实时实验实施也将用于控制系统验证和确认。最终目标是在实际应用的背景下，例如自由空间激光通信系统，将控制系统与AO系统集成。

Due to the multidisciplinary nature of the field, the development of MEMS requires that joint efforts be deployed in many scientific and engineering domains. The proposed program will focus on developing control strategies for high performance and high functionality MEMS applications, in particular the control of electrostatically actuated devices and their applications to adaptive optics (AO) and scanning micro-mirrors.     Control of MEMS involves serious challenges due to inherent system nonlinearity, uncertainties, time-varying properties, and disturbances. Addressing such control problems is vital for future advancement and the advent of new applications of MEMS technology. In our previous work, nonlinear control techniques have been introduced to the design of MEMS control systems. It is demonstrated through concrete applications that the performance of addressed systems can be drastically improved.     In the proposed program, we will continue to work on the nonlinear control of individual micro-devices, but emphasize the output feedback control schemes, closed-loop system analysis, and algorithms for avoiding the singularity related to the uncontrollability in some particular points. The research work will then focus on the control of interconnected MEMS array, which is the model of typical AO systems. The technique of distributed control of spatially distributed systems will be employed and issues like system architecture, controller synthesis, optimality, computational efficiency, control task scheduling, and synchronization will be addressed.     The developed systems will be tested using software simulations. For this purpose, more realistic modeling will be addressed and software packages ANSYS, CoventorWare, and COMSOL (previously FEMLAB) will be used to simulate MEMS devices. Hardware-in-the-loop simulations and real-time experimental implementations will also be used for control system verification and validation. The ultimate goal is to integrate the control system with AO systems in the context of practical applications, e.g. free space laser communication systems.