Distributed Fusion Estimator for Multi-Rate Multi-Sensor Systems with Unknown Inputs, Time Delays and Fading Measurements

A distributed fusion estimator (FE) is proposed for multi-rate multi-sensor systems (MMSs) with unknown inputs (VIs), one-step time delays and fading measurements, in which the state update (SU) rate is a positive multiple of the measurement sampling (MS) rates of different sensors. The fading measurements are described by using independent random variables with known intervals and obeying the certain probability distributions. Firstly, the time-delayed MMS is transformed into a single-rate system at the MS points without delays. The local filter (LF) at the SU point is derived based on the filter with Kalman filtering form that does not depend on UI information. Then, the estimation error covariance matrix is derived on the basis of LFs, and the distributed optimal FE is proposed. Finally, simulation results of numerical examples validate the effectiveness of FE.