AKH-NFIS: Adaptive Krill Herd Network Fuzzy Inference System for Mobile Robot Navigation

Mobile robot navigation has been a current issue in the most recent two decades. Mobile robots are necessary to explore in obscure and dynamic situations. To solve the aforementioned issues an extended Kalman filter (EKF) and adaptive Krill Herd network fuzzy inference system (AKH-NFIS) techniques are proposed for the self-sufficient portable robot route. This is in charge of avoidance of obstacles in an obscure static and dynamic environment. Initially, the start and goal position will be set and the obstacles identified in front of the robot will be checked using the sensor. This sensor captures the environmental information around the mobile robot. Subsequently, to deal with the filtering problem of sensor data, EKF will be used. By EKF more accurate position estimation will be obtained by using dynamic information of data. Subsequently, the obstacle distances from the robot and the obstacle avoidance angle are calculated and fed as input to the training dataset. This training data set trains AKH-NFIS controller obtained by designing a Krill herd optimization algorithm adaptive network fuzzy logic-based navigation controller. The left wheel velocity and right wheel velocity are the output from the proposed system. The robustness of the proposed navigation controller will be assessed by exploring the mobile robot in various conditions. The experimental result demonstrates that our proposed strategy outperforms by correlation with existing strategies.