A Behavioral Decision-Making Model of Learning and Memory for Mobile Robot Triggered by Curiosity

Abstract

Learning and memorizing behavioral decision in the process of environmental cognition to guide future decision is an important aspect of research and application in mobile robotics. Traditional rule-based behavioral decision approaches have difficulty in adapting to complex and changing environments. The offline decision-making approaches lead to poor adaptability to dynamic environments, while behavioral decision-making based on reinforcement learning relies on data acquisition, and the learned knowledge cannot guide mobile robots to quickly adapt to new environments. To address this issue, this article proposes a brain-inspired behavioral decision model that can perform incremental learning by simulating the logical structure of memory classification in the brain, as well as the memory conversion mechanisms of hippocampus, prefrontal cortex, and anterior cingulate cortex. The model interacts with the environment through semisupervised learning and learns the current decision online, simulating the memory function of humans to enable mobile robots to adapt to changing environments. In addition, an internal reward mechanism driven by curiosity is designed, simulating the reinforcement mechanism of curiosity in human memory, encoding the memory of unfamiliar behavioral decisions for mobile robots, and consolidating the memory of frequently made behavioral decisions, improving the learning and memory capacity of mobile robots in environmental cognition. The feasibility of the proposed model is verified by physical experiments in different environments.