Deadbeat Simultaneous Parameter-State Estimation for Continuous-time Systems: a Kernel-based Approach

The simultaneous estimation of parameters and states in continuous-time linear MISO systems is considered in this paper. The estimates of system’s parameters and states are provided simultaneously by a single estimator, where Volterra operators are suitably applied to the I/O measurements. This makes the estimator independent on the initial conditions thanks to the design of suitably shaped kernel functions equipped with non-asymptotic properties. As a result, in the noise-free scenario, instantaneous convergence can be obtained. No high-gain injection nor periodic resetting is necessary. Numerical examples are reported showing the effectiveness of the proposed estimator