Simultaneous Continuous- and Discrete-Time Unknown Input Filtering

This study presents a simultaneous continuous- and discrete-time unknown input filtering for both linear time-invariant continuous- and discrete-time stochastic systems with arbitrary unknown inputs. This new approach proposes a system reformation technique to transform the original linear time-invariant system into one that can reflect all online estimable unknown inputs on the output equation. Using a hybrid unknown input filtering method, the simultaneous estimation of unknown inputs and states for both the continuous- and discrete-time systems can be achieved. The relationship with the existing optimal continuous-time input and state filter is also addressed.