A new multi-agent system to simulate the foraging behaviors of Physarum

Abstract

Physarum Polycephalum is a unicellular and multi-headed slime mold, which can form high efficient networks connecting spatially separated food sources in the process of foraging. Such adaptive networks exhibit a unique characteristic in which network length and fault tolerance are appropriately balanced. Based on the biological observations, the foraging process of Physarum demonstrates two self-organized behaviors, i.e., search and contraction. In this paper, these two behaviors are captured in a multi-agent system. Two types of agents and three transition rules are designed to imitate the search and the contraction behaviors of Physarum based on the necessary and the sufficient conditions of a self-organized computational system. Some simulations of foraging process are used to investigate the characteristics of our system. Experimental results show that our system can autonomously search for food sources and then converge to a stable solution, which replicates the foraging process of Physarum. Specially, a case study of maze problem is used to estimate the path-finding ability of the foraging behaviors of Physarum. What's more, the model inspired by the foraging behaviors of Physarum is proposed to optimize meta-heuristic algorithms for solving optimization problems. Through comparing the optimized algorithms and the corresponding traditional algorithms, we have found that the optimization strategies have a higher computational performance than their corresponding traditional algorithms, which further justifies that the foraging behaviors of Physarum have a higher computational ability.Keywords Physarum Polycephalum Á Multi-agent system Á Self-organized computational system Á Maze

Morphogenesis in robot swarms

“…(ii) Different types of Turing patterns on fish skin. gradients (21)(22)(23)(24)(25)(26)(27)(28)(29) or have used more complicated interactions between signals, such as RD systems (30,31), gene regulatory networks (GRNs) (32)(33)(34)(35), and swarm chemistry (36) [for a more extensive review, see (37)].…”

Simulating Transport Networks With a <italic>Physarum</italic> Foraging Model

“…The foraging behavior and the computing ability of Physarum have inspired many novel models, which can be used to reveal the evolving mechanisms and computational intelligence of Physarum. The existing Physarum foraging models can be divided into two main categories [26], [36]: the top-down description based on the population-based modeling [23], [25], [37] and the bottom-up simulation from the perspective of individual behaviors [24], [26], [38], [39].…”

“…The experimental results illustrate that simple local behaviors, according to the characteristics of chemotaxis, can guide the agent population and generate more complex and dynamic transport networks [38]. Later, Liu et al [36] and Wu et al [47] have improved the multi-agent model by optimizing the structure of each agent and designing two types of agents to simulate the search and the contraction behaviors of Physarum respectively.…”

A Slime Mold-Ant Colony Fusion Algorithm for Solving Traveling Salesman Problem

“…Subsequently, Bonifaci et al [32] verified the feasibility and convergence of SMA. What's more, Chinese researchers such as Southwest University, have proposed a strategy by using SMA to optimize the original pheromone of ACO [33][34][35][36]. However, SMA started later, so the algorithm is not systematic and mature enough and the stability needs to be improved.…”