State Estimation Based on Multirate Kalman Filter for Power Systems Driven by Switching Inverter

Abstract

This paper presents a design methodology of a Kalman filter for a multirate control system with a fast input system. There are many applications whose sensory system is slower than the input system in a power system. The sensory system is required to perform fast sensing while ensuring a low signal-to-noise ratio, although it requires additional time compared with that of a low-resolution converter. The Kalman filter is an effective tool in such a situation and it estimates parameters while eliminating noise. Because the optimality of this filter is ensured when the system model and noise variance on the system are well identified, rigorous discretization should be considered on the multirate control system. With this in mind, the paper presents the design method of a multirate Kalman filter as is the case for a single-rate controller. Further, the noise variance and oversampling ratio of the multirate controller is confirmed using the Monte Carlo method. Numerical simulation of a voltage inverter theoretically validates the design methodology.