| High precision tools that allow for direct mechanical manipulation and sensing are enabling new manufacturing, scientific, and medical explorations into submicron length scales. The manipulation and sensing capabilities of these tools are determined not only by the hardware motion system, but also the algorithms that plan and execute the intricate tasks. The unprecedented levels of precision and response time demanded by applications on submicron length scales provide new challenges in the algorithm development.; In this research we consider precision motion control techniques for achieving very high performance of micro- and nano-tracking systems. In particular, we focus on the iterative learning control (ILC) technique, which uses repetition of the task to update and improve the control with each trial. By leveraging the high repeatability of tracking systems, ILC is able to use information from previous responses to automatically tune a fast, even noncausal, precision motion control algorithm. A thorough survey of ILC analysis and design techniques for multi-input, multi-output linear time-invariant systems is presented. Many micro- and nanoscale tracking systems fall under this class. Time- and frequency-domain ILC analysis results are used to identify key design parameters and their roles in convergence, performance, and robustness.; This dissertation develops two distinct contributions to the ILC technique. First, we show how ILC can be used in the identification of a state-dependent nonlinearity on the input channel. This class of nonlinearities is relevant to micro- and nano-scale tracking systems because the input channel typically represents force, and thus these nonlinearities describe the mechanical interactions of the tool and workspace, as well as parasitic disturbances in the actuators. For the purpose of illustration, the technique is applied to a permanent magnet linear motor to measure and map force ripple and frictional forces.; Second, we develop a set of time-frequency ILC design and analysis techniques. These techniques are used to custom-design the ILC for precision tracking of aggressive trajectories suitable for rapid microscale and nanoscale manufacturing, assembly, or scanning. They are also useful for applications where disturbance signals may change rapidly such as in micro- and nano-materials testing. The time-frequency ILC is applied to a 3-DOF piezo driven flexure stage yielding 4x higher bandwidth over the original feedback controller and achieving 1500 Hz tracking bandwidth during critical partitions of a given trajectory. |
