Quadratic Estimation for Discrete Time System with Packet Dropping and Unknown Input

This paper deals with the problem of quadratic estimation for discrete time systems with packet dropping and unknown input. The Kronecker algebra approach is employed to obtain the second order Kronecker powers of original state and measurement. After that, the quadratic system is constructed by means of the initial system and its second Kronecker power. Furthermore, a particular condition is imposed to suppress the interference caused by unknown input. Based on the constraint, we design a minimum variance unbiased quadratic estimator, which is not only suitable to Gaussian noise, but also applicable to non-Gaussian noise. Finally, a simulation example is provided to illustrate the good performance of the quadratic estimator designed.