A Three-Phase Novel Angular Perturbation Technique for Metaheuristic-Based School Bus Routing Optimization

This article introduces a novel three-phase angular perturbation technique for solv- 1 ing the School Bus Problem (SBP), an NP-hard combinatorial optimization challenge. The three 2 phases—initial solution, global and local search, and solution exploration—focus on optimizing 3 school bus routes by minimizing sharp turns and reducing travel distances. The performance of 4 three metaheuristic algorithms (GAACOSA-AP, BFSPSO-AP, RL-AP) is evaluated, and significant 5 improvements in routing efficiency are observed. we evaluated the performance of our hybrid algo- 6 rithm against the industry-leading optimization software, CPLEX, and other optimization techniques 7 across multiple school scenarios. The results show that the GAACOSA-AP algorithm consistently 8 outperforms its counterparts, achieving substantial reductions in travel distances, particularly for 9 complex routes with numerous stops. The integration of Angular Perturbation was key to improving 10 route geometry by, reducing sharp turns, and enhancing overall path optimization by an average 11 reduction of 27.12% compared to 22.37% without Angular Perturbation.12