Robust adaptive control using integral quadratic constraints

This paper introduces a new approach to robust adaptive H ∞ control. The main idea of the paper is to extend an existing approach to robust H ∞ control to allow for constant or slowly varying unknown parameters in an adaptive way. This is achieved by treating the presence of the unknown parameters as states resulting in a bilinear nonlinearity. A copy of this nonlinearity is repeated in the controller and the presence of the repeated nonlinearity is exploited using a special integral quadratic constraint.