Hybrid particle swarm optimization for vehicle routing problem with time windows

  • Authors:

  • S. Masrom;Siti. Z. Z. Abidin;A. M. Nasir;A. S. Abd. Rahman

  • Affiliations:

  • Faculty of Computer and Mathematical Science, Universiti Teknologi MARA, Malaysia;Faculty of Computer and Mathematical Science, Universiti Teknologi MARA, Malaysia;Faculty of Computer and Mathematical Science, Universiti Teknologi MARA, Malaysia;Computer and Information Science Department, Universiti Teknologi PETRONAS, Malaysia

  • Venue:
  • MAMECTIS/NOLASC/CONTROL/WAMUS'11 Proceedings of the 13th WSEAS international conference on mathematical methods, computational techniques and intelligent systems, and 10th WSEAS international conference on non-linear analysis, non-linear systems and chaos, and 7th WSEAS international conference on dynamical systems and control, and 11th WSEAS international conference on Wavelet analysis and multirate systems: recent researches in computational techniques, non-linear systems and control
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

Vehicle routing problem with Time Window (VRPTW) has received much attention by researchers in solving many scheduling applications for transportation and logistics. The objective of VRPTW is to use a fleet of vehicles with specific capacity to serve a number of customers with various demands and time window constraints. As a non-polynomial (NP) hard problem, the VRPTW is complex and time consuming, especially when it involves a large number of customers and constraints. This paper presents a hybrid approach between Particle Swarm Optimization (PSO) and Genetic Algorithm (GA) for solving VRPTW. The reason for hybridization is to overcome the problem of premature convergence that exists in standard PSO. Premature convergence often yields partially optimized solutions because of particles stagnation. The proposed hybrid PSO implements a mechanism that automatically trigger swarm condition which will liberate particles from sub-optimal solutions hence enabling progress toward the maximum best solution. A computational experiment has been carried out by running the hybrid PSO with the VRPTW benchmark data set. The results indicate that the algorithm can produce some improvement when compared to the original PSO.