Genetic algorithms for the resource-constrained project scheduling problem in aircraft heavy maintenance

TL;DR

This paper develops genetic algorithms to efficiently solve the NP-hard resource-constrained project scheduling problem in aircraft heavy maintenance, reducing downtime by optimizing maintenance schedules.

Contribution

It introduces a genetic algorithm approach with heuristic initial population and resource allocation methods tailored for aircraft heavy maintenance scheduling.

Findings

Algorithms outperform existing solutions in computational efficiency.

Proposed methods effectively minimize maintenance plan makespan.

Resource allocation strategies improve scheduling quality.

Abstract

Due to complex sets of interrelated activities in aircraft heavy maintenance (AHM), many airlines have to deal with substantial aircraft maintenance downtime. The scheduling problem in AHM is regarded as an NP-hard problem. Using exact algorithms can be time-consuming or even infeasible. This article proposes genetic algorithms for solving the resource-constrained project scheduling problem (RCPSP) in AHM. The objective of the study was to minimise the makespan of the maintenance plan. The proposed algorithms applied five heuristic dispatching rules to generate an initial population based on activity list formation. Resource allocation methods for RCPSPearliest start time (EST) and workgroup and earliest start time (WEST)-were used to evaluate the fitness value. The elitist and roulette wheel methods were applied in the selection process. The sequences of the activity lists were then iteratively improved by crossover and mutation operations. The results show that the proposed algorithms perform efficiently compared to the existing solutions in terms of computational time and resource allocation.