Dynamic remodeling model based on chemotaxis of slime molds.

Abstract

Social infrastructure networks, essential for daily life and economic activities, encompass utilities such as water, electricity, roads, and telecommunications. Dynamic remodeling of these systems is crucial for responding to continuous changes, unexpected events, and increased demand. This study proposes a new dynamic remodeling model inspired by biological mechanisms, focusing on a model based on the chemotaxis of slime molds. Slime molds adapt spontaneously to environmental changes by remodeling through the growth and degeneration of tubes. This capability can be applied to optimizing and dynamic remodeling social infrastructure networks. This study elucidated the chemotactic response characteristics of slime molds using biological experiments. The mold's response was observed by considering changes in the concentration of chemicals as environmental changes, confirming that slime molds adapt to environmental changes by shortening their periodic cycles. Subsequently, based on this dynamic response, we propose a new dynamic model (oscillated Physarum solver, O-PS) that extends the existing Physarum solver (PS). Numerical simulations demonstrated that the O-PS possesses rapid and efficient path-remodeling capabilities. In particular, within a simplified maze network, the O-PS was confirmed to have the same shortest-path searching ability as the PS, while being capable of faster remodeling. This study offers a new approach for optimizing and dynamically remodeling social infrastructure networks by mimicking biological mechanisms, enabling the rapid identification of solutions considering multiple objectives under complex constraints. Furthermore, the variation in convergence speed with oscillation frequency in the O-PS suggests flexibility in responding to environmental changes. Further research is required to develop more effective remodeling strategies.