Adaptive Unscented Kalman Filter for Robot Navigation Problem (Adaptive Unscented Kalman Filter Using Incorporating Intuitionistic Fuzzy Logic for Concurrent Localization and Mapping)

Abstract

The navigation of a mobile robot is a very important issue, especially for an autonomous mobile robot. A robot autonomously can track the area by interpreting the arena, building an adequate map, and localizing itself to this map. This paper proposes a Hybrid filter for Concurrent Localization and Mapping (CLAM) in the navigation to rectify the faults, basically Unscented Fast Simultaneous Localization and Mapping (SLAM) (UFS). We also interrogate the effectiveness of the IF system to investigate nonlinear attributes. A probabilistic method has planned the solution to the CLAM issue, which is an essential requirement for the navigation of robots. The Hybrid filter CLAM contains an Intuitionistic Fuzzy Logic (IFL) and Unscented Kalman Filter (UKF). The IFL is first ordered by using a correctness function explained on score functions for the non-membership function (NMF) and membership function (MF) of the IFL. Then this ordering is utilized to develop a method for a sufficient decision on the CLAM issue. The proposed method has a few privileges in management robot navigation with nonlinear movements owing to the inference feature of the IFL, which also needs a fewer quantity of comparisons than the UFS and shows much better efficiency from the robustness, perspective assessment exactitude, and reliability than the UFS, also, for learning the measurement and control noise covariance matrices for increasing correctness and consistency are utilized IFL. The Hybrid filter CLAM proficiency compared with the UFS has a smaller quantity of computations and good efficiency for bigger areas as demonstrate in the results of simulation and experimental.