Coolant Bleed Flow Modeling and Prediction for Aircraft Engines Based on TSO-RF Algorithm

Enlai Zhang (School of Mechanical and Automotive Engineering, Xiamen University of Technology, No.600 Ligong Road, Jimei District, Xiamen, Fujian Province, 361024)
Jiaqiang Jiang (School of Mechanical and Automotive Engineering, Xiamen University of Technology, No.600 Ligong Road, Jimei District, Xiamen, Fujian Province, 361024)
Zijuan Ren (School of Mechanical and Automotive Engineering, Xiamen University of Technology, No.600 Ligong Road, Jimei District, Xiamen, Fujian Province, 361024)

Journal of Engineering and Technological Sciences
Vol. 58 No. 4 (2026): Vol. 58 No. 4(2026): August

Coolant bleed flow forecasting is critical for the engineering health of aircraft engines. Using the datasets from Commercial Modular Aero-Propulsion System Simulation (C-MAPSS), this paper aims to establish high-precision coolant bleed flow models by the random forest (RF) algorithm combined with tuna swarm optimization (TSO). 17 sensor variables that are moderately or highly correlated with the low-pressure turbine (LPT) and high-pressure turbine (HPT) are selected as inputs with LPT and HPT as outputs. After being trained and validated on the FD002 and FD004 datasets, the TSO-RF model significantly reduces mean squared error, mean absolute error and root mean squared error, and improves the determination coefficient R² compared to other RF models. It verifies the superiority of the TSO-RF model in predicting engine coolant bleed flow, providing reliable technical support for subsequent evaluation and management in aircraft engine bleed air system.