Analysis of Semantic Probabilistic Inference Control Method in Multiagent Foraging Task

Abstract

Adaptation in robotics systems is often implemented as some form of learning. While much research is dedicated to studying policy and value approximation in reinforcement learning, some methods are based on rule inference and logical descriptions. One of these methods is based on a semantic probabilistic inference algorithm that has its roots in the theory of functional systems. In this article, the method is applied to a distributed multiagent foraging problem that has an important property of providing an environment that allows to study a decentralized system of individually learning agents. We compare the performance of this method to other methods: Q-learning and a random choice algorithm as a baseline. We also propose a modification of the algorithm that includes an exploration behavior. Experiments are carried out in a computer simulation system. The results show the performance of the algorithms with different parameters, as well as the effect of exploration on the performance.