Adaptive stabilization of bilinear systems

For a class of deterministic and stochastic bilinear systems, “modified self-tuning regulators” are given by using recursive least-squares and modified recursive least-squares, respectively. The closed-loop systems are shown globally stable and asymptotically optimal in some sense without any restriction on location of poles or zeros of nominal linear part of the open-loop system considered. Simulations show that the methods used here work very well