An Improved BAT Algorithm for Solving Job Scheduling Problems in Hotels and Restaurants

TL;DR

This paper introduces an improved Bat Algorithm (MBA) that enhances convergence speed and effectiveness for large-scale optimization, demonstrated through benchmark tests and real-world hotel and restaurant job scheduling problems.

Contribution

The paper proposes a modified Bat Algorithm with better global search and convergence capabilities, tailored for complex job scheduling applications.

Findings

MBA outperforms original BA in benchmark tests

MBA achieves better convergence in real-world scheduling

Enhanced algorithm is suitable for large-scale problems

Abstract

One popular example of metaheuristic algorithms from the swarm intelligence family is the Bat algorithm (BA). The algorithm was first presented in 2010 by Yang and quickly demonstrated its efficiency in comparison with other common algorithms. The BA is based on echolocation in bats. The BA uses automatic zooming to strike a balance between exploration and exploitation by imitating the deviations of the bat's pulse emission rate and loudness as it searches for prey. The BA maintains solution diversity using the frequency-tuning technique. In this way, the BA can quickly and efficiently switch from exploration to exploitation. Therefore, it becomes an efficient optimizer for any application when a quick solution is needed. In this paper, an improvement on the original BA has been made to speed up convergence and make the method more practical for large applications. To conduct a comprehensive comparative analysis between the original BA, the modified BA proposed in this paper, and other state-of-the-art bio-inspired metaheuristics, the performance of both approaches is evaluated on a standard set of 23 (unimodal, multimodal, and fixed-dimension multimodal) benchmark functions. Afterward, the modified BA was applied to solve a real-world job scheduling problem in hotels and restaurants. Based on the achieved performance metrics, the proposed MBA establishes better global search ability and convergence than the original BA and other approaches.