Systematization of approaches to parameters settings of genetic algorithms for optimizing public transport schedules

Abstract

The research investigates the problem of the practical implementation efficiency of genetic algorithms (GAs) for public transport schedule synchronization, specifically focusing on the impact of heuristic parameter selection.

The purpose of the research is to systematically summarize contemporary scientific approaches to justifying the parametric tuning of genetic algorithms for synchronizing public transport timetables, and to identify the specifics of parameter selection in research aimed at improving the efficiency and reliability of solutions under variable transport network operating conditions. The study employs methods of systemic and comparative analysis of current scientific works dedicated to the use of GAs in transport schedule optimization problems.

The findings of current scientific works on the influence of GAs parametric tuning on its functional capacity and result relevance in solving the timetable synchronization problem have been systematized. The key GAs tuning parameters that form the architecture of the computational process were identified: population size, crossover probability, and mutation probability. It was established that in the majority of analysed scientific works these parameters are determined empirically without adequate theoretical justification, indicating the need to develop methodological approaches to their selection. A critical analysis of existing works revealed that researchers pay insufficient attention to the heuristic selection of GAs parameters when solving the public transport timetable synchronization problem, which limits the reliability of algorithmic solutions under variable transport network operating conditions. A promising direction for further research was identified, consisting in the development of adaptive GAs parameter tuning strategies to enhance solution efficiency and reliability in line with real public transport operating conditions.

The practical significance of the study lies in its systematized analytical generalization of existing approaches, which will enable researchers and transport industry specialists to use the findings as a toolkit for justifying GAs parametric tuning in order to improve the effectiveness of its application in solving public transport timetable synchronization problems.