Robust control systems design using plant step response data

Summary form only given. The paper describes an approach to the design of robust controllers using data derived from plant step responses. These could be obtained from simulation studies or from plant step tests. The approach is based on a suitable combination of model fitting, control system CAD and robustness theory. The methods fit naturally into traditional SISO and MIMO control system CAD methods. Both graphical frequency-domain and time-domain methods of analysis are consistent with the approach. A detailed model of the plant is not required. The system can contain delays and other infinite-dimensional elements. The use of time-domain characterizations of the error enables both robust stability margins and closed-loop performance deterioration to be estimated. The techniques are simple and easy to apply. The designer has control over controller complexity