Data Fusion Applied to the LBBA Algorithm to Improve the Localization of Mobile Robots

Bioinspired optimization algorithms derive their effectiveness from the number of particles used to find the global minimum or maximum of a problem. However, as the number of potential solutions increases, the computational effort also increases due to the exponential expansion of the calculations required to evaluate the objective functions of each particle throughout the algorithms. Another fundamental aspect to be considered is the algorithm's ability to avoid local optima efficiently and not compromise the results, leading to failure to achieve an ideal or acceptable outcome. This article discusses a solution to this challenge in the context of localizing mobile robots in known environments, proposing sensory data fusion applied to the Leader-Based Bat Algorithm (LBBA). This new approach exploits the algorithm's ability to work with leader bats, whose function is to influence specific groups (colonies), employing information from a sensor (compass) to assist in the distribution of particles (bats) on the map during the search process. The idea is to allow precise localization of Unmanned Ground Vehicles (UGV) on known maps when a motion capture system is unavailable, allowing controlling such robots based on good position feedback.