Integral LQG Model Following Controller

The integral LQG design process presented above is an efficient technique for the design of multiple input / multiple output control systems. Unlike classical root locus methods, requirements of more than one loop can be handled at a time. In addition, the systematic approach to selecting regulated variables and forming criteria outputs affords more insight than LQG techniques using state weighting only. The use of transmission zeros, both inherent in the plant and created by the designer, is crucial since they dictate the asymptotic nature of the closed loop eigenvalues. Although classical root locus techniques treating one loop at a time have been adequate in the past for the design of transport airplane control systems, the need is arising for multiple loop design methods. For example, in search of greater efficiency, airplane structural stiffness has been reduced leading to lower frequency flexible modes. The result is a challenging controls problem since control inputs now risk exciting structural modes. The technique presented here allows engineers with experience applying classical design methods to quickly learn a multiple loop design approach since the single loop compensation ideas they are familiar with are the same tools used to develop criterion outputs, the central feature of this method. This design process has been developed at Boeing over the past decade by building on the wealth of LQG design literature. Although many Boeing people have contributed, credit should be given to Dr. James Blight for his extensive work in this area.