Model identification, controller design and validation of nano positioning system using time domain analysis

A linear standing wave ultrasonic motor driven Nano Positioning System with resolution 100nm requires model identification and robust controller design from its time domain data for application in high speed micromanufacturing system. From the time domain data acquired during experiments and iterative model estimations three best are chosen for controller design. Instrumental Variable method is used for model identification giving fitting accuracy in the range of 94%. Using this estimation internal model control (IMC) and Robust response time controllers have been designed. Using different methods and model transfer functions 12 sets of controllers have been designed and tested on the set up. The most optimum controller show rise time (ms), settling time (ms) and overshoot (%) in range of [1.2, 1.6, 0] respectively. which is significantly above the factory setting controller performance i.e improvement by 80% in rise time and 25% in settling time. Detail analysis of various system models and controller combination have been conducted. MATLAB® simulation testing is used for estimation and primary validation of models and controllers whereas final validation is achieved by implementing the designed controllers on the Nano Positioning System using PIMikroMove® software.