A robust multivariable control system design and tuning software (IMCTUNE)

Software has been developed to support research and education in model-based control. The controller design and tuning concepts are based on a linear internal model control structure (IMC). Tuning for robustness of the control system is accomplished by adjusting controller parameters, the filter time constants, to assure that the maximum magnitude of the closed loop frequency responses of all possible processes in the set of uncertain processes meets the user specified values. Such criteria relate approximately to time domain performance. The process uncertainty set is described in terms of uncertain process parameters that lie between upper and lower bounds. The research has mostly concentrated on how to reliably and efficiently utilize the computer power to correctly tune the control system.; The current research unifies and improves both single-input single-output (SISO) and Multivariable (MIMO) version into single software and substantially improves the user interface. The new MIMO version is totally different from the earlier versions. The new user interfaces were built for the user to easily input the MIMO system and interact with the software. The software can automatically suggest a possible multivariable process model which result in a controller. The user may make changes to the model/controller, as long as it pass s the Nyquist test included in the software. A new method of tuning of the MIMO system was developed. Several examples show that the new tuning method is more reliable and more computationally efficient than the existing method. Time simulations of the MIMO system (up to a 5 x 5 system) were also implemented using Simulink.