A Multi-Agent Framework for Zero-Dimensional Reduced-Order Model Planning

2026-07-12

AI Summary

This work addresses the limitations of traditional zero-dimensional reduced-order models (0D ROMs), whose design relies heavily on manual expertise, suffers from restricted topological exploration, and struggles with long-horizon, tightly coupled global optimization. To overcome these challenges, the authors propose Z-COPA, a multi-agent framework that introduces a novel graph-based representation for 0D ROMs, reformulating flow network synthesis as a graph structure optimization problem. By integrating a Symbolic Action Graph Engine (SAGE) with a MILP-guided navigation optimizer (MGN), Z-COPA enables synergistic forward and inverse co-design. The approach uniquely combines large language model agents, Chain-of-Thought reasoning, ReAct, retrieval-augmented generation (RAG), mixed-integer linear programming (MILP), and graph neural representations to transcend the long-term planning limitations of single-agent systems. Evaluated across six benchmarks—including aircraft engine bleed systems, IEEE power grid reconfiguration, and water distribution networks—the method significantly enhances automation and solution quality, achieving state-of-the-art performance.

Abstract

Zero-dimensional reduced-order models (0D ROMs) are central to multi-dimensional design workflows for high-end complex equipment. However, the planning process currently relies on manual expertise, limiting topological exploration and prolonging iterations. Even traditional optimization methods such as Genetic Algorithms (GA) are typically confined to local parameter tuning. Although Large Language Model (LLM) agents have shown promise in exploring large sample spaces, and frameworks such as Chain of Thought (CoT) and Reason and Act (ReAct) improve reasoning reliability, while Retrieval-Augmented Generation (RAG) overcomes domain knowledge barriers, a single agent still falls short for the long-horizon and highly coupled nature of complex 0D ROM planning. This paper proposes the Zero-dimensional reduced-order model CO-Planning framework (Z-COPA), a multi-agent architecture featuring a Symbolic Action Graph Engine (SAGE) and a MILP-Guided Navigation (MGN) optimizer. Its core innovation is a dedicated graph representation method that accurately encodes the 0D flow network topology, converting the empirical planning process into a rigorous graph structure optimization problem. We validate the forward and inverse design capabilities and generalization performance of Z-COPA on two real aircraft engine secondary-air systems, two IEEE power-distribution reconfiguration benchmarks, and two water-distribution network benchmarks. The results show superior task completion quality, obtaining the best performance in both forward and reverse design of air systems. Z-COPA disrupts the traditional 0D model planning paradigm, providing a new technical approach for exploring broader topological space and achieving highly automated, globally optimal air system architectures.

Problem

• Zero-dimensional reduced-order models
• topology exploration
• automated planning
• complex system design
• graph optimization

Innovation

• multi-agent framework
• zero-dimensional reduced-order model
• graph representation
• MILP-guided optimization
• topology exploration