Adaptive Landmark-Based Navigation System Using Learning Techniques

Abstract

Abstract. The goal-directed navigational ability of animals is an essential prerequisite for them to survive. They can learn to navigate to a distal goal in a complex environment. During this long-distance navigation, they exploit environmental features, like landmarks, to guide them towards their goal. Inspired by this, we develop an adaptive landmarkbased navigation system based on sequential reinforcement learning. In addition, correlation-based learning is also integrated into the system to improve learning performance. The proposed system has been applied to simulated simple wheeled and more complex hexapod robots. As a result, it allows the robots to successfully learn to navigate to distal goals in complex environments.

Decision Making Process of Hexapods in a Model of Complex Terrains

“…Not only does our approach allow for the description of the complexity of a terrain, it also provides description of the size of the obstacle with respect to LW [2,23]. Thus, the proposed method may complement existing methods designed for trajectory planning [5][6][7][8].…”

“…A number of methods based on artificial intelligence has been developed to control the trajectory of a walking hexapod robot [5][6][7], [8] however, these have not been designed for leg coordination [2]. There are also methods for controlling legs on a flat surface without any obstacles [9][10][11][12][13].…”

Synaptic plasticity in a recurrent neural network for versatile and adaptive behaviors of a walking robot

“…The results presented here show that the employed embodied adaptive neural closed-loop system (Supplementary Figure 1 ) is a powerful approach for achieving versatility and adaptivity in machines. As the neural mechanisms are modular, it is flexible and offers the future possibility of integrating other modules, like a goal-directed navigation learning module (Zeidan et al, 2014 ) and a neural path integration module (Goldschmidt et al, 2015 ). This will enable the robotic system to be capable of navigating in complex environments toward given goals and autonomously return to its home position.…”