Fuzzy State Noise Driven Kalman Filter for Sensor Fusion

Abstract: This article proposes a fuzzy state noise-driven Kalman filter for sensor fusion to estimate the instantaneous position and attitude of an unmanned air vehicle for navigation purpose. The formulation of the state noise covariance matrix has been carried out using the fuzzy regression method applied to the state residuals. This algorithm has been embedded in the real-time hardware and tested for performance on ground and not in real flight. A comparative study between the proposed and conventional algorithm illustrates its efficacy.