Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A reinforcement learning guided hybrid evolutionary algorithm for the latency location routing problem

Yuji Zou, Jin-Kao Hao, Qinghua Wu

TL;DR

A reinforcement learning guided hybrid evolutionary algorithm following the framework of the memetic algorithm that relies on a diversity-enhanced multi-parent edge assembly crossover to build promising offspring and a reinforcement learning guided variable neighborhood descent to determine the exploration order of multiple neighborhoods.

Abstract

The latency location routing problem integrates the facility location problem and the multi-depot cumulative capacitated vehicle routing problem. This problem involves making simultaneous decisions about depot locations and vehicle routes to serve customers while aiming to minimize the sum of waiting (arriving) times for all customers. To address this computationally challenging problem, we propose a reinforcement learning guided hybrid evolutionary algorithm following the framework of the memetic algorithm. The proposed algorithm relies on a diversity-enhanced multi-parent edge assembly crossover to build promising offspring and a reinforcement learning guided variable neighborhood descent to determine the exploration order of multiple neighborhoods. Additionally, strategic oscillation is used to achieve a balanced exploration of both feasible and infeasible solutions. The competitiveness of the algorithm against state-of-the-art methods is demonstrated by experimental results on the three sets of 76 popular instances, including 51 improved best solutions (new upper bounds) for the 59 instances with unknown optima and equal best results for the remaining instances. We also conduct additional experiments to shed light on the key components of the algorithm.

A reinforcement learning guided hybrid evolutionary algorithm for the latency location routing problem

TL;DR

A reinforcement learning guided hybrid evolutionary algorithm following the framework of the memetic algorithm that relies on a diversity-enhanced multi-parent edge assembly crossover to build promising offspring and a reinforcement learning guided variable neighborhood descent to determine the exploration order of multiple neighborhoods.

Abstract

The latency location routing problem integrates the facility location problem and the multi-depot cumulative capacitated vehicle routing problem. This problem involves making simultaneous decisions about depot locations and vehicle routes to serve customers while aiming to minimize the sum of waiting (arriving) times for all customers. To address this computationally challenging problem, we propose a reinforcement learning guided hybrid evolutionary algorithm following the framework of the memetic algorithm. The proposed algorithm relies on a diversity-enhanced multi-parent edge assembly crossover to build promising offspring and a reinforcement learning guided variable neighborhood descent to determine the exploration order of multiple neighborhoods. Additionally, strategic oscillation is used to achieve a balanced exploration of both feasible and infeasible solutions. The competitiveness of the algorithm against state-of-the-art methods is demonstrated by experimental results on the three sets of 76 popular instances, including 51 improved best solutions (new upper bounds) for the 59 instances with unknown optima and equal best results for the remaining instances. We also conduct additional experiments to shed light on the key components of the algorithm.
Paper Structure (25 sections, 8 equations, 8 figures, 8 tables, 2 algorithms)

This paper contains 25 sections, 8 equations, 8 figures, 8 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Reinforcement learning guided hybrid evolutionary algorithm
  3. Main scheme
  4. Population initialization
  5. Offspring generation based on MPEAX
  6. Repair procedure
  7. Mutation
  8. Reinforcement learning guided VND with strategic oscillation
  9. Rationale and general RL-SOVND framework
  10. Q-learning for deciding the exploring order
  11. Variable neighborhood descent with strategic oscillation
  12. Population updating
  13. Discussions
  14. Computational results
  15. Benchmark instances
  16. ...and 10 more sections

Figures (8)

  • Figure 1: Illustration of the crossover procedure.
  • Figure 2: Cumulative probability distribution for the time to reach a target value.
  • Figure 3: The heatmap for the number of shared edges of solution pairs and the scatter plot for edge sharing ratio between solutions and the best-known solution on instance 121222.
  • Figure 4: The heatmap for the number of shared edges of solution pairs and the scatter plot for edge sharing ratio between solutions and the best-known solution on instance 200-3-1.
  • Figure 5: Comparative results of RLHEA with its two variants RLHEA1 and RLHEA2 on the 54 large instances.
  • ...and 3 more figures