Loop Transfer Recovery Techniques for Discrete-Time Non-Minimum Phase Plants with Feedback Delays. (Q4244840)

Consider a discrete-time system \[ x(t+1)=Ax(t)+Bu(t)+w(t), \] \[ y(t)=Cx(t)+v(t). \] Introduce a quadratic performance index (with delay) defined as \[ V=\lim_{T\to \infty} E\left\{ \frac{1}{T}\sum_{t=0}^{T-1} [y'(t)Q_cy(t)+u'(t+L)R_cu(t+L)\right\}. \] The optimal control is given by \[ u(t+L)=-F\hat x(t+L| t), \] where \[ F=(B'P_cB+R_c)^{-1}B'P_cA, \] \(P_c\) is a non-negative definite solution to the matrix Riccati equation \[ P_c=A'P_cA-A'P_cB(B'P_cB+R_c)^{-1}B'P_cA+C'Q_cC. \] For \(L>1\), the prediction by connecting the \((L-1)\)-step state predictor \[ \hat x(t+L| t)=A^{L-1}\hat x(t+1| t)+ A^{L-2}Bu(t+1)+\dots +Bu(t+L-1) \] is obtained, where \[ \hat x(t+1| t)=A\hat x(t| t-1)+ Bu(t)+K[y(t)-C\hat x(t| t-1)]. \] In the article, various systems with delay are considered.