Reducing UWB Indoor Localization Error Using the Fusion of Kalman Filter with Moving Average Filter

Abstract

Indoor localization is important for robot navigation because it allows for robots to accurately determine their location and movement within a space. This is especially important for robots that are used in confined areas, like warehouses or homes, where there is not as much open space to navigate. Indoor localization gives robots the ability to plan their paths strategically and navigate around obstacles in a timely and efficient manner. Therefore, it is crucial for the indoor positioning system (IPS) to be stable and accurate. In this paper, we presented a fusion of non-complex filtering algorithms which combines Kalman filter with Moving Average (MA) filter in order to reduce localization error using Ultra-Wideband (UWB). The performance of the technique was measured against the conventional method of Kalman filtering, and it was found that the average error was reduced even more by the proposed strategy compared to the standard Kalman filtering approach.