Design of an improved quantum-inspired evolutionary algorithm for a transportation problem in logistics systems

  • Authors:

  • Lixing Wang;S. K. Kowk;W. H. Ip

  • Affiliations:

  • Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hung Hom, Hong Kong;Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hung Hom, Hong Kong;Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hung Hom, Hong Kong

  • Venue:
  • Journal of Intelligent Manufacturing
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

Logistics faces great challenges in vehicle schedule problem. Intelligence Technologies need to be developed for solving the transportation problem. This paper proposes an improved Quantum-Inspired Evolutionary Algorithm (IQEA), which is a hybrid algorithm of Quantum-Inspired Evolutionary Algorithm (QEA) and greed heuristics. It extends the standard QEA by combining its principles with some heuristics methods. The proposed algorithm has also been applied to optimize a problem which may happen in real life. The problem can be categorized as a vehicle routing problem with time windows (VRPTW), which means the problem has many common characteristics that VRPTW has, but more constraints need to be considered. The basic idea of the proposed IQEA is to embed a greed heuristic method into the standard QEA for the optimal recombination of consignment subsequences. The consignment sequence is the order to arrange the vehicles for the transportation of the consignments. The consignment subsequences are generated by cutting the whole consignment sequence according to the values of quantum bits. The computational result of the simulation problem shows that IQEA is feasible in achieving a relatively optimal solution. The implementation of an optimized schedule can save much more cost than the initial schedule. It provides a promising, innovative approach for solving VRPTW and improves QEA for solving complexity problems with a number of constraints.