[论文] DynaSchedBench: Calibrated Dynamic Scheduling Benchmarks and Obse...

论文概要

研究领域: AI
作者: Shijie Cao, Yuan Yuan, Jing Liu
发布时间: 2026-05-28
arXiv: 2605.27566

中文摘要

神经网络组合优化在动态柔性作业车间调度问题（DFJSP）上的进展正面临方法论的紧张关系：静态基准易导致过拟合，而未校准的生成器则用随机噪声掩盖算法真实能力。为此，本文提出了DynaSchedBench——一个严格控例生成过程的DFJSP诊断框架。该框架不依赖参数采样，而是使用序列事件空间校准器（SESC）计算新颖的调度压力指数（SSI）来按难度分层实例。SESC在计算效率上显著优于进化基线，同时可靠收敛到目标指标。在此校准环境基础上，研究揭示了LLM调度智能体的关键局限：在动态调度的逐步在线决策中，存在一个"可观测性悖论"——给予智能体完整的结构信息反而会降低策略表现，简洁信息反而更优。此外，尽管耗费大量token，工具增强和细化策略仍未能可靠提升性能，大多数LLM智能体始终无法超越强启发式基线。

原文摘要

Progress in neural combinatorial optimization for Dynamic Flexible Job Shop Scheduling Problem (DFJSP) is currently hindered by a methodological tension: static benchmarks encourage benchmark overfitting, while uncalibrated generators obscure algorithmic capability with stochastic noise. To resolve this, we introduce DynaSchedBench, a diagnostic framework for DFJSP that rigorously controls the instance-generation process. Instead of relying on parameter sampling, our approach utilizes Sequential Event-Space Calibrator (SESC) that computes a novel Schedule Stress Index (SSI) to stratify instances by difficulty. We demonstrate that SESC is substantially more computationally efficient than evolutionary baselines while converging reliably to the target metrics. The framework integrates modular components for instance generation, snapshot-based simulation, agents, evaluation, and visualizatio...

自动采集于 2026-05-29

#论文 #arXiv #AI #小凯