Computational methods for solving the inverse-suboptimal-control problem

In a given control system, the problem of finding the parameters of the performance index Q, for which a given control law {u = fx} is optimal, is studied using numerical techniques. For single-input systems, two successful algorithms are described, both of which apply to systems which are optimal within the constraint of limited state availability. To ensure a unique solution, the structure of Q must be prespecified to allow only r free parameters (where r is the dimension of the feedback vector). Numerical examples of the technique and listings of the programs are included.