Study on Maintenance Decision-making Technology Based on Multi-Agent in PHM

Maintenance decision-making technology adopted in PHM system greatly affects the effectiveness of the system. The authors began with describing the architecture for OSA-CBM. Then indicated that there were inherent limitations in model-based method and data-driven method adopted in maintenance decision-making and suggested the Multi-Agents method which is widely applied in many domains may be a proper choice. The components of Maintenance Decision-Making Agent (MDA) were presented in detail. On the analysis of Remaining Life Prediction Agent, Opportunistic Maintenance actions was proposed, which will help to turn the reactive maintenance philosophy to scheduled or preventive maintenance philosophy. It was proved that the capability of cooperation, interaction and parallel processing of autonomous agents had been greatly improved. The Multi-Agents model for maintenance decision-making was more effective in achieving the intelligence of the system.