A Comparative Study of Nature Inspired Algorithms for Prioritized Foraging

Prioritized foraging in swarm intelligence refers to the ability of a swarm to collect items while accounting for differences in their importance. Prioritized foraging is examined in settings which contain multiple item types, where each item type is associated with a distinct priority level. The objective is to study and adapt nature-inspired foraging models to efficiently solve these multi-item-type foraging problems, addressing the challenge of collecting resources that are not equally important and should be selected according to their assigned priorities. The nature-inspired models explored include a swarm-based baseline model, a desert ant model, and a honey bee model. These models are evaluated with respect to the foraging ability, gain, energy consumption, and collisions between agents. A variety of environment scenarios is generated, varying in both the number of item types and the spatial distribution of items. Sensitivity analysis is used to quantify how key parameters influence performance, and optimal parameter configurations are identified for each model across the different item-type scenarios. The results show that each model brings distinct strengths and trade-offs to multi-item-type prioritized foraging. The baseline model is simple and reliably consistent, the desert ant model offers a balance between efficiency and limited communication, and the honey bee model performs well in complex environments with many item types. Across all models, the adaptive prioritization strategy improves performance by directing agents toward high-priority items while preserving overall foraging efficiency. This work extends current knowledge of prioritized foraging to richer environments involving multiple item types and priority levels. It further demonstrates how nature-inspired models can be adapted and improved for complex resource-collection tasks. The findings are relevant to real-world applications where efficient prioritization is essential, including logistics, environmental monitoring, and disaster response.

Grantor dc:publisher

Stellenbosch : Stellenbosch University

Year dc:date.issued

2026

Author dc:creator

• Lee, Jin Ree

Advisor dc:contributor.advisor

• Engelbrecht, Andries

Language dc:language.iso

en

Repository record dc:identifier.uri

https://scholar.sun.ac.za/handle/10019.1/136288

OAI identifier oai:identifier

oai:scholar.sun.ac.za:10019.1/136288