Impact of Different Noise Distributions in the Application of Kalman Filter in Sensor Fusion

This paper explains the concept of the Kalman filter as an estimator of the system state based on measurements in the presence of noise. Sensor fusion allows for the measurement of factors that are not immediately measurable as well as the creation of more accurate data. The theoretical underpinning for the definition of a filter and Kalman-based sensor fusion, and filter and sensor fusion algorithms are explained and implemented in simulation. The application of the filter for sensor fusion is presented, and in the end, simulation results are obtained for different types of noise distribution - normal as the usual one for physical quantities and uniform as dominant in digitized values.