This research introduces a novel Graph-RAG based approach that leverages LLM agents to automate optimization modeling for complex sensor array signal processing problems.

• Overcomes limitations of existing prompt-based techniques by incorporating domain-specific knowledge
• Uses a graph retrieval-augmented generation (RAG) architecture to enhance model performance
• Demonstrates superior results in the specialized SASP domain where traditional methods struggle
• Provides a foundation for automated engineering optimization in technically complex fields

This advancement matters for engineering teams working with sensor arrays by reducing modeling complexity while improving accuracy, potentially accelerating development cycles for security applications and other signal processing systems.

Leveraging LLM Agents for Automated Optimization Modeling for SASP Problems: A Graph-RAG based Approach