Real-Time Localization Application of MEKF-VDPL Algorithm on Autonomous Wheelchair in a Dynamic Environment

The problem of determining the location of a mobile robot while moving is called the localization problem in autonomy. During navigation, it is vital to obtain high accuracy information about where the mobile robot is on the global coordinate system so that it can reach a predetermined goal smoothly and safely without hitting obstacles. For this reason, there has been an increased focus on localization in recent years with the aim of designing algorithms that can make both faster and more accurate position estimation. In this study, the MEKF-VDPL (Modified Extended Kalman Filter using Virtual Dynamic Point Landmark) localization algorithm, which was designed in MATLAB and has been introduced to the literature recently, has been made available on a wheelchair for the first time in real time, and real-time localization tests of this algorithm have been performed using a test scenario closer to the real-world dynamic environment. The localization test of the AMCL (Adaptive Monte Carlo Localization) algorithm has been also performed with the same test scenario. When the results obtained are compared, it is understood that the MEKF-VDPL localization algorithm has both faster and more accurate performance results.