LLM Supported Complex System Anomaly Detection and Intelligent Defect Classification Model

Abstract

With the increasingly frequent application of complex systems, anomaly detection and defect classification have become important tasks to ensure the stable operation of systems. This paper is based on a complex system anomaly detection and intelligent defect classification model with Large Language Model (LLM) as the core. Through the data processing and learning capabilities of LLM, multi-source information is fused to establish an efficient and accurate detection and defect classification method. The model extracts abnormal patterns from the system data through LLM and intelligently classifies them against known defects, thereby achieving early identification and classification of potential faults in complex systems. The experimental results show that this model demonstrates excellent detection accuracy and classification ability in fields such as power, manufacturing, and transportation. According to the actual application scenarios, the model can adaptively process, thereby ensuring that complex systems have better robustness and security. This research provides new technical ideas and practical paths for future intelligent operation and maintenance and precise fault diagnosis.