Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Keep Rehearsing and Refining: Lifelong Learning Vehicle Routing under Continually Drifting Tasks

Jiyuan Pei, Yi Mei, Jialin Liu, Mengjie Zhang, Xin Yao

TL;DR

The paper tackles the problem of neural VRP solvers failing under continual task drift when per-task training is limited. It introduces Dual Replay with Experience Enhancement (DREE), a lifelong RL framework that combines problem instance replay, behavior replay, and an experience enhancement mechanism to rehearse and refine buffered knowledge. Empirically, DREE outperforms existing lifelong learning solvers on drifting CVRP and TSP tasks and generalizes well to unseen benchmarks, while remaining competitive in periodically stationary settings and across different base solvers. The work demonstrates that jointly reusing problem instances and solver behaviors, together with active improvement of buffered experiences, yields superior plasticity-stability trade-offs and practical applicability for real-world continual learning in routing problems.

Abstract

Existing neural solvers for vehicle routing problems (VRPs) are typically trained either in a one-off manner on a fixed set of pre-defined tasks or in a lifelong manner on several tasks arriving sequentially, assuming sufficient training on each task. Both settings overlook a common real-world property: problem patterns may drift continually over time, yielding massive tasks sequentially arising while offering only limited training resources per task. In this paper, we study a novel lifelong learning paradigm for neural VRP solvers under continually drifting tasks over learning time steps, where sufficient training for any given task at any time is not available. We propose Dual Replay with Experience Enhancement (DREE), a general framework to improve learning efficiency and mitigate catastrophic forgetting under such drift. Extensive experiments show that, under such continual drift, DREE effectively learns new tasks, preserves prior knowledge, improves generalization to unseen tasks, and can be applied to diverse existing neural solvers.

Keep Rehearsing and Refining: Lifelong Learning Vehicle Routing under Continually Drifting Tasks

TL;DR

The paper tackles the problem of neural VRP solvers failing under continual task drift when per-task training is limited. It introduces Dual Replay with Experience Enhancement (DREE), a lifelong RL framework that combines problem instance replay, behavior replay, and an experience enhancement mechanism to rehearse and refine buffered knowledge. Empirically, DREE outperforms existing lifelong learning solvers on drifting CVRP and TSP tasks and generalizes well to unseen benchmarks, while remaining competitive in periodically stationary settings and across different base solvers. The work demonstrates that jointly reusing problem instances and solver behaviors, together with active improvement of buffered experiences, yields superior plasticity-stability trade-offs and practical applicability for real-world continual learning in routing problems.

Abstract

Existing neural solvers for vehicle routing problems (VRPs) are typically trained either in a one-off manner on a fixed set of pre-defined tasks or in a lifelong manner on several tasks arriving sequentially, assuming sufficient training on each task. Both settings overlook a common real-world property: problem patterns may drift continually over time, yielding massive tasks sequentially arising while offering only limited training resources per task. In this paper, we study a novel lifelong learning paradigm for neural VRP solvers under continually drifting tasks over learning time steps, where sufficient training for any given task at any time is not available. We propose Dual Replay with Experience Enhancement (DREE), a general framework to improve learning efficiency and mitigate catastrophic forgetting under such drift. Extensive experiments show that, under such continual drift, DREE effectively learns new tasks, preserves prior knowledge, improves generalization to unseen tasks, and can be applied to diverse existing neural solvers.
Paper Structure (44 sections, 3 equations, 3 figures, 9 tables, 1 algorithm)

This paper contains 44 sections, 3 equations, 3 figures, 9 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Neural VRP Solvers
  4. Lifelong Learning Neural VRP Solvers
  5. Lifelong Learning under Continually Drifting VRP Tasks
  6. Continually Drifting VRP Task
  7. Problem Formulation
  8. DREE
  9. Experience Buffer
  10. Problem Instance Replay
  11. Behavior Replay
  12. Experience Enhancement
  13. Experiments
  14. Experimental Setups
  15. Dataset.
  16. ...and 29 more sections

Figures (3)

  • Figure 1: VRP instances arise in lifelong learning scenarios, with node distribution of the task changes from uniform to clustered.
  • Figure 2: DREE in the lifelong learning scenario where the task drifts between every two consecutive time steps.
  • Figure 3: Learning curve of task order 1, measured by test performance. Grey background indicates the epochs that the corresponding principal task is involved in generating intermediate tasks.