GLOP: Learning Global Partition and Local Construction for Solving Large-scale Routing Problems in Real-time
Haoran Ye, Jiarui Wang, Helan Liang, Zhiguang Cao, Yong Li, Fanzhang Li
TL;DR
GLOP tackles real-time, large-scale routing by hierarchically partitioning instances into sub-TSPs and solving SHPPs with a local autoregressive policy. It jointly learns a non-autoregressive global partition heatmap $m{ ext{H}}_{oldsymbol{ ho}}$ and autoregressive revisers, enabling scalable yet accurate construction across $TSP$, ATSP, CVRP, and PCTSP using a single trained set of local policies. The two-stage curriculum and heatmap-driven decomposition enable a one-size-fits-all solver that achieves state-of-the-art real-time performance, including a $TSP_{100K}$ gap of $5.1\%$ with $174\times$ speedup over LKH-3 and superior CVRP/PCTSP real-time results. Overall, GLOP offers robust cross-distribution performance and practical impact for industrial-scale routing under tight latency constraints, while remaining extensible to further hybridizations of AR and NAR paradigms.
Abstract
The recent end-to-end neural solvers have shown promise for small-scale routing problems but suffered from limited real-time scaling-up performance. This paper proposes GLOP (Global and Local Optimization Policies), a unified hierarchical framework that efficiently scales toward large-scale routing problems. GLOP partitions large routing problems into Travelling Salesman Problems (TSPs) and TSPs into Shortest Hamiltonian Path Problems. For the first time, we hybridize non-autoregressive neural heuristics for coarse-grained problem partitions and autoregressive neural heuristics for fine-grained route constructions, leveraging the scalability of the former and the meticulousness of the latter. Experimental results show that GLOP achieves competitive and state-of-the-art real-time performance on large-scale routing problems, including TSP, ATSP, CVRP, and PCTSP.