Control system design for nanometer scale positioning systems

Nanometer positioning presents unique challenges because devices capable of this accuracy exhibit hysteretic behavior. Hysteresis is characterized by rate independent memory and multi-valued output, limiting even the application of nonlinear control. Previous researchers have pursued a control strategy dependent on precision models of hysteresis. The physical principles of these hysteretic devices are not well understood and the models exhibiting best fidelity to experimental evidence are phenomenological, not analytic, thus they are computationally intensive for reasonable accuracy and their behavior is unique to each device.; Our thesis is that one may treat hysteretic behavior as a disturbance and compensate for it as one would for other disturbance. Three hysteresis compensation strategies are demonstrated which exhibit performance superior to prior reported results and none of which require a hysteresis model. Novel passive and active disturbance rejection strategies, as well as a hybrid combination inheriting favorable characteristics of both strategies, are successfully developed and implemented.