Railway Track Maintenance Scheduling using Artificial Bee Colony and Harmony Search

Abstract

The objective of this paper is to propose a heuristic algorithm to optimize the railway track maintenance scheduling, a NP-hard problem, by reflecting conditions of the actual field more quickly and easily. We develop the mechanism based on Binary Artificial Bee Colony (BABC) and Binary Harmony Search (BHS), and verify their performance through simulation experiments. Our proposed BABC and BHS mechanisms were applied to problems composed of 30, 60, 100, and 200 operations for railway track maintenance scheduling to carry out experiments and analysis. On comparing it with the results solved by CPLEX, it is found that the mechanism could present an optimal solution within limited time by user.Keywords: Railway Track Maintenance Scheduling, Artificial Bee Colony, Harmony Search Background and Purpose of the StudyThe most efficient and effective maintenance schedule is established by considering the abilities and availabilities of various resources such as manpower and equipment in the track maintenance field [15]. An efficient and effective maintenance plan can be explained in various ways. The criterion would be varied depending on the purpose or circumstances of the organization carrying out the maintenance. However, performing more jobs with given resources for an appropriate period is the most general and common purpose for optimization of a maintenance schedule.At present, maintenance jobs for railway tracks are carried out through various methods depending on the conditions of the field. Therefore, field conditions reflected in a schedule also show a variety of differences. This paper proposes a methodology applying a heuristic algorithm for flexible handling of such a situation and a quicker establishment of a schedule.Even though a heuristic algorithm could not guarantee optimal solutions, it can present optimal or best solutions within a short time or period given by schedulers. Furthermore, it also has an advantage that it can solve various problems more easily.Therefore, various conditions at the railway track maintenance fields can also be reflected more easily to optimize it.Among them, Artificial Bee Colony (ABC) and Harmony Search (HS) have been actively studied, their performances were proved, and related studies are constantly being carried out even in resource assignment and scheduling fields [2,6,7].The purpose of this paper is to mathematically model the railway track maintenance scheduling problem and propose a mechanism that can be applied to obtain the best solution to the track maintenance schedule with sample data of KORAIL's maintenance history of high-speed railway tracks. Status of Related StudiesPresent methods which are used to solve a resource assignment or scheduling problem in various fields could also be exploited as a method to solve optimization problems for track maintenance scheduling. As there is an availability of several resources, which are able to carry out a single job process in the problem for track maintenance scheduling covered in this paper, it develops a cor...