Dual-Drive Agent for Accurate Gas Turbine Degradation Identification

Researchers from Harbin Engineering University have developed a novel dual-drive agent combining data-driven and thermodynamic models to accurately identify gas turbine degradation states. The proposed method significantly reduces the number of required training samples while enhancing identification accuracy, achieving a classification accuracy of over 96.55% for individual components. This breakthrough has the potential to improve the online health monitoring of complex gas turbine systems.

Key Takeaways:

• The dual-drive agent combines a data-driven model with a thermodynamic model to identify gas turbine degradation states, achieving a classification accuracy exceeding 96.55%.
• The proposed method reduces the number of required training samples while enhancing identification accuracy, outperforming conventional model-based and data-driven approaches.
• The dual-drive agent exhibits superior performance in both accuracy and stability, making it well-suited for online health monitoring of complex gas turbine systems.
• The research concludes that the proposed method is an effective solution for identifying gas turbine degradation states, with a maximum identification error limited to 0.0122.
• The study used the Rank Whale Decision Optimization Algorithm (RWDOA) within the reduced space to identify the degradation state.
• The research highlights the importance of combining data-driven and thermodynamic models for achieving accurate gas turbine degradation identification.
• The study was conducted by a team of researchers from Harbin Engineering University, led by Xuemin Li, and included Jingjing Zhang and Jian Li as co-authors.
• The research was peer-reviewed and published in the Energy journal, Volume 335, in 2025.

Statistics:

• Classification accuracy of over 96.55% for individual components.
• Maximum identification error limited to 0.0122.
• The proposed method reduces the number of required training samples.
• The dual-drive agent exhibits superior performance in both accuracy and stability.
• The research was published in the Energy journal, Volume 335, in 2025.

Sources:

• "An Agent Composed of Data Model and Thermodynamic Model for Multi-component Degradation Identification of Gas Turbine Online." Energy, 2025; 335.
• NewsRx. Studies Conducted at Harbin Engineering University on Data Modeling Recently Reported (An Agent Composed of Data Model and Thermodynamic Model for Multi-component Degradation Identification of Gas Turbine Online). Information Technology Newsweekly. November 4, 2025; p 810.
• Xuemin Li, Harbin Engineering University, College of Power & Energy Engineering, 145 Nantong St, Harbin 150001, People's Republic of China.