Closed-loop subspace Predictive Control for Linear Parameter Varying systems (i) - the nominal case

An MPC approach for LPV systems in input-output form

“…Now, given the current value and the future trajectory of the scheduling variable and the input of the system, the future output of G I in (5) can be computed as follows: (6). Therefore, the key prediction equation for G I with h 0 (k) ≡ 0 is given by…”

“…The use of observers to access the state information may deteriorate significantly closed-loop performance in terms of input disturbance rejection when plant input constraints become activated, as in that case, the nonlinearities start to dominate the behavior of the closed-loop system, see [5]. To cope with these issues, a subspace-based predictive control for LPV systems has been proposed in [6]. The critical issue in this approach is that no stability guarantee has been provided.…”

Robust data driven H-infinity control for wind turbine

“…For the evaluation of performance, residual signal which is the difference between measured output and predicted one is used. Thus, residual, , is derived from (6,7,8) with inclusion of fault as follows:…”

An improved robust model predictive control for linear parameter‐varying input‐output models

“…for j = 0, 1, 2, … , N −1, whereθ(k+j) ∈ R n x ×n y is computed as in Equation 16 except its rows from (1+n y n a ) to (n u +n y n a ) are given byθ [1+n y n a ,n u +n y n a ] (k + j) = θ [1+n y n a ,n u +n y n a ] (k + j)…”

“…Moreover, the use of observers to estimate the states may also deteriorate closed‐loop performance significantly in terms of input disturbance rejection when input constraints become activated . To handle this, a subspace‐based predictive control for LPV systems has been proposed in the work of Dong et al without stability guarantee. However, the complexity of this scheme increases exponentially with the order and number of scheduling variables.…”