Optimization to Robustify Controllers Using the Disk Margin Tool

Multiple design methods are conceived exclusively for a process described by an ideal mathematical model. On the other hand, if the process presents non-negligible uncertainties in practice, then it becomes a requirement to improve the regulator to ensure the desired performance metrics for the entire uncertainty family. As such, the purpose of this paper is to enforce robustness on a precalculated regulator through the disk margin framework using a proposed optimization problem. Additionally, several performance constraints will also be enforced. Linear time-invariant (LTI) single-input single-output (SISO) uncertain systems are considered. The solutions are provided with readily-available tools. Improvements are shown in a comparative case study performed using a common linear-quadratic regulator servo structure for a single axis CNC machine model.