A Bio-Inspired Goal-Directed Cognitive Map Approach to Robot Navigation and Mapping

This paper presents a cognitive map (CM) model fused with a histogram-based reactive local navigator algorithm for real-time robot mapping and navigation. Drawing inspiration from biological navigation, the CM model integrates elements like landmarks, Euclidean distance, exploratory movement, and reward-driven actions. Key features of the model include types of cognitive mapping cells designed to replicate their biological functions in a computationally efficient manner. Built on cognitive map concepts used to explain mammalian navigation, this CM model enables robots to navigate complex environments without needing complete environmental exploration. Additionally, an enhanced ℬ-spline curve scheme ensures a smoother, safer trajectory. After planning a global route, the model utilizes a histogram-based local navigation algorithm to dynamically avoid obstacles while creating real-time maps as the robot follows its path. Simulation and comparison results validate that this integrated approach supports real-time navigation and mapping in unknown environments, offering significant improvements in robotic spatial navigation and real-time map building.