Towards robust sliding mode control of electrostatic microactuators

Micro Electro Mechanical Systems refer to formation of electro-mechanical systems in the scales of micrometers. In terms of popularity of the technology and interest toward this, Borovic et al have mentioned in their recent paper[28]: “MEMS are the next step in the silicon revolution that began 40 years ago. Currently, MEMS are in an exponential growth stage such as that enjoyed by thesemiconductor industry.”. While MEMS technology enables significant reduction in physical size of various types of sensors, actuators and systems by several orders of magnitude, does not necessarily imply less system complexity. In fact, modelling or identifying characteristics of such devices can be difficult as a combination of advanced modelling techniques and Finite Element analysis would be necessary to obtain an appropriate mathematical model approximation. In the context of MEMS actuator technology, the electrostatic microactuators have received much attention for real applications such as DLP technology and research work on more advanced control algorithms. Traditionally, such devices have been controlled in an open loop fashion but problems with the so-call ed pull-in phenomenon (a form of instability in the system due to the electrostatic forces that arise between the plate of the microactuator) limits performance. Feedback control has enabled improved performance, although still most works exist in research labs rather than already implemented and in mass production. The work presented in this dissertation relates to voltage control of parallel-plate electro-static microactuator devices, with the particular aim of extending the travelling range of the microactuator plates beyond the “pull-in” condition thus taking advantage of increase range of operation (which is useful when finer motion resolution is required in applications such as DLP etc.). The particular aim is studying the capabilities of state feedback related Sliding Mode Control to impro ving system performance and robustness to uncertainty compared to the linear control equivalents. Simulation studies are performed with the help of Matlab and Simulink