Monitoring Aero-Engine Fuel Flow Based on Neural Net

Abstract:

Engine fuel flow is one of the most important parameters of aero-engine performance, which can accurately reflect the actual condition of aero-engine. It'is possible to monitor the aero-engine fuel flow (FF) by reading the data of QAR (Quick Access Recorder). However, it's difficult to monitor the actual condition of aero-engine through the traditional method because of the vast amount of QAR data and the complexity of the engine itself. Feed forward process neural network is adopted to monitor the QAR-based aero-engine fuel flow . The results of the simulation are acceptable and show that the Neural Net model is an effective method to monitor aero-engine conditions.

• Periodical: Advanced Materials Research (Volumes 490-495)
• Pages: 979-984
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.490-495.979
• Online since: March 2012
• Authors: Zhao Hui Ji, Shen Jiang
• Keywords: Fuel Flow, Monitoring, Quick Access Recorder (QAR)

References

[1] Geng Hong, Jie Jun. QAR data based on the cruise segment of aircraft fuel flow regression model. Aero-engine Vol. 43 (2008) , pp.46-50, in Chinese.

[2] Liu Jing. Based on the analysis of flight data estimation model aircraft fuel. Master thesis of Nanjing University of Aeronautics and Astronautics. 2010, pp.5-10, in Chinese

[3] He Xingui, Liang Jiuzhen. Some neural network theory of the process issues. Engineering Science of China Vol. 2 (2000) , pp.40-44, in Chinese.

[4] Zhang Dinghui, Shao Huihe. Fault diagnosis based on neural network inference methods.Learning Journal of Shanghai Jiaotong University Vol. 33(1999) , pp.619-621, in Chinese.

[5] Zhou Kaili, Kang Yaohong. Based on neural network model and its MALAB emulator programming [M].Tsinghua University Press, (2005) , pp.211-212, in Chinese