Computational model of locust swarm formation, implemented in NetLogo (v4.1.2). The model can be used to explore the notion of 'critical density', and how that affects swarm formation.
The model loosely incorporates elements of the SIR (susceptible-infectious-removed) model. Grasshoppers are treated as 'susceptible', capable of turning into locusts. A grasshopper can be 'infected' (i.e., turn into a locust) by making contact with a sufficient number of other grasshoppers & locusts within a prescribed window of time (the criterion is motivated by the idea that overcrowding induces swarm behavior). A locust is treated as 'removed', in the sense that once it turns into a locust, it remains a locust (i.e., it could not be 'infected' again).
The model also incorporates the notion of flocking. A grasshopper is assumed to wander randomly, whereas after turning into a locust, it follows flocking/swarming behavior.
Model assumptions:
- only solitary (grasshopper) and gregarious (locust) members
- population is fixed
- members cannot escape world boundaries
- solitary can become gregarious
- gregarious remain gregarious
- grasshopper wanders randomly
- locust adopts swarming behavior
Model parameters:
population: number of members in the grasshopper-locust population (all are initialized as grasshoppers)locust-mobility: distance a locust moves per cyclevision: distance around a locust within which neighbors are considered as swarm matesminimum-separation: distance below which a locust will separate from its nearest neighbor, and above which it aligns to the average swarm headingmax-align-turn: maximum turn angle allowed by a locust aligning to the average swarm headingmax-cohere-turn: maximum turn angle allowed by a locust cohering to the average swarm headingmax-separate-turn: maximum turn angle allowed by a locust separating from its nearest neighbor
Michael E. Rowan — mrowan137 — michael@mrowan137.dev.
- The model is a modification of Uri Wilensky's flocking model, which mimics e.g. the flocking of birds or fish.