Sequential design of multivariable control systems

The sequential return difference method is a technique for the design of multivariable control systems in which the individual loops are designed in sequence, using classical tools. In essence, this method implies that at each stage only one loop of a loop gain matrix is closed. The method of this paper, in contrast, takes the view that at each stage only one loop remains to be closed, and uses the interaction factor for its design. The interaction factor is used also to judge the robustness to model uncertainty