General optimal control system design by structured uncertainty description and eigenstructure assignment method

In this paper, the structured uncertainty description of non-minimum phase systems in the linear quadratic Gaussian and loop transfer recovery (LQG/LTR) process is firstly proposed. With above process, we can confirm that there is no difference between minimum-phase and nonminimum-phase systems in Kalman filter design. Furthermore, the well-developed eigenstructure assignment (ESA) method applied to a LTR process is firstly proposed. The proposed method can efficiently improve the high-gain problem in a conventional LQG/LTR one. In addition, our method provides better performance in time-domain and frequency-domain responses. With the ESA technique in a LTR process, the control system can simultaneously provide prescribed stability and take into account of both time-domain and frequency-domain specifications. Finally, the numerical results illustrate the advantage of our proposed method.