The emergence of collective obstacle avoidance based on a visual perception mechanism

Coordinated and ordered collective behaviors that emerge from decentralized and local self-organizing interactions among individuals widely exists in many biological groups. The classical models often focus on the imitation of biological phenomena with the given position and velocity, but ignore the modeling of individual perception during the emergence of collective behaviors. Based on the mechanism of visual perception, we present a model named visual perception-decision-propulsion (VPDP) and clearly define several evaluation indicators to explore the emergence of collective obstacle avoidance in flocks. Within this model, there is no centralized control and no information exchange among individuals. Instead, individuals interacting with others merely rely on the perceptual information represented by whether the visual field is occupied by neighbors. We demonstrate that our model effectively achieves collective obstacle avoidance, preserving a safe distance between individuals and obstacles while maintaining relatively good coherence. Moreover, the sensitivity analysis of the parameters indicates the robustness of our VPDP model. Hence, this paper provides a novel model for collective obstacle avoidance based on visual perception as well as a detailed analysis. (c) 2021 Elsevier Inc. All rights reserved.