Information Fusion Estimation for Nonlinear Systems with Multiplicative Noise

This paper investigates the distributed information fusion estimation problem for non-linear systems with multiplicative noise in multi-sensor systems. The paper uses virtual states with virtual process noise and designs one-step predictors and state update equations for the virtual conditions. An extended Kalman estimator with a forgetting factor is applied to optimize the valuation of each sensor for the non-linear system. matrix-weighted is used in the fusion process and compared to verify the optimization of the state valuation by adding the forgetting factor to the filtering process.