Abstract:
Ant nest relocation is smoother and swifter than any other animals. The active ratio within the populations of ants that participate in nest relocation is only 58.0% at b...Show MoreMetadata
Abstract:
Ant nest relocation is smoother and swifter than any other animals. The active ratio within the populations of ants that participate in nest relocation is only 58.0% at best and 31.0% at worst. A considerable number of ants are always inactive and the smaller number of active ants carry and relocate them to a new nest. The swarm behavior of ant nest relocation attracts researchers in the computational intelligence community. Many models and simulations have been proposed, though there exists an open problem whether such low active ratio improves ant nest relocation. A positive answer to the problem would provide a technological inspiration for proposing a promising swarm-based algorithm in focus on the active ratio within the populations of computational agents. In this study, we use a particle swarm optimization (PSO) algorithm and simulate real-world ant nest relocation. Our PSO-based algorithm duplicates the velocity and position of an inactive particle with the velocity and position of an active particle. The number of particles that the algorithm computes is dramatically reduced and the global best position can be identified at an early step of iteration of the algorithm while restricting the loss of diversity in the search-space as the overall result. In simulation, our algorithm performed significantly better and faster than the full active ratio 100%'s performance at the active ratios 15%, 30%, 35%, 45%, 55%, 60%, and 75%-95%. We processed clustering to the simulation results and showed that the low active ratios improved ant nest relocation. Furthermore, three records of the field researches that were carried out by external ant experts in biology empirically supported that we have successfully modeled and simulated real-world ant nest relocation by our PSO-based algorithm.
Published in: 2019 IEEE Congress on Evolutionary Computation (CEC)
Date of Conference: 10-13 June 2019
Date Added to IEEE Xplore: 08 August 2019
ISBN Information: