Performance Analysis of Selected Swarm-Based Robot Search Algorithms for Target Tracking

This study focuses on comparing the performance of several swarming algorithms, which are mainly basic and correlated with random walk, Brownian motion, and levy flight. These algorithms were applied to a swarm of E-pucks for tasks such as target tracking and surveying. The analysis evaluates algorithm accuracy, swarm and target density impacts, and consensus times. Correlated random walk achieved the highest accuracy, while Brownian motion and correlated random walk showed that higher swarm densities correlate with greater accuracy. The same insight may be acquired for target densities. From the simulation, it was seen that a greater number of targets allowed the swarm to acquire more accurate results. Lastly, for consensus times, it was a basic random walk that provided the most consistent and fastest consensus times, whereas correlated random walk provided high and greatly varying consensus times – indicating inconsistencies.