This paper presents a general framework to cope with full-order H-infinity linear parameter-varying (LPV) filter design subject to inexactly measured parameters. The main novelty is the ability of handling additive and multiplicative uncertainties in the measurements, for both continuous and discrete-time LPV systems, in a unified approach. By conveniently modelling scheduling parameters and uncertainties affecting the measurements, the H-infinity filter design problem can be expressed in terms of robust matrix inequalities that become linear when two scalar parameters are fixed. Therefore, the proposed conditions can be efficiently solved through linear matrix inequality relaxations based on polynomial solutions. Numerical examples are presented to illustrate the improved efficiency of the proposed approach when compared to other methods and, more important, its capability to deal with scenarios where the available strategies in the literature cannot be used
This paper addresses the voltage regulation of dual active bridge converters based on robust PI controllers. The plant model presents a dependence on an uncertain parameter, which defines the operation condition, motivating the design of robust PI controllers for this application. Linear matrix inequalities are then used to synthesize control gains which ensure the closed-loop pole location inside a sector in the left-hand of the complex plane for the entire set of uncertainties. Comparisons with a PI controller designed only for a nominal condition indicate advantages of the robust PI in terms of step responses. It is shown that other criteria can also be associated with the proposed design conditions to ensure the achievement of frequency domain specifications.
This paper addresses the problem of constant sampling discretization of uncertain time-invariant continuous-time linear systems in polytopic domains. To circumvent the difficulty of dealing with the exponential of uncertain matrices, a new discretization method, based on Taylor series expansion, is proposed. The resulting discrete-time uncertain system is described in terms of homogeneous polynomial matrices with parameters lying in the unit simplex and an additive norm-bounded uncertainty which represents the discretization residual error. As a second contribution, linear matrix inequality (LMI) based conditions for the synthesis of a stabilizing state feedback control for discrete-time linear systems with polynomial dependence on the uncertain parameters and an additive norm-bounded uncertainty are proposed. Numerical experiments illustrate the discretization technique advantages of using higher orders in the Taylor series expansion to obtain more precise approximations. The examples also show that, at the price of simple line searches in a scalar parameter, and using Lyapunov functions of higher degrees, less conservative results for robust state feedback control design of discretized uncertain systems can be obtained.