V.G. Avgustinovich - Dr.Sc. (Eng.), Professor, Department of Aircraft Engines, Perm National Research Polytechnic University. E-mail: firstname.lastname@example.org
T.A. Kuznetsova - Ph.D. (Eng.), Associated Professor, Department of Design and Technologies in Electrical Engineering, Perm National Research Polytechnic University; The Head of Distance Education Technologies Centre. E-mail: email@example.com
In modern digital automatic control system of aircraft engines (ACS GTE) the flight reliability improvement is achieved through the creation of algorithmic information redundancy based on built-in on-board a mathematical engine model.
The technical and theoretical difficulties of practical implementation of algorithmic reservation by the model are associated with the high dimensionality of the engine state space, that are significantly higher than the dimension of the vector of parameters measured on board. There is a problem of identification of sensor fault with subsequent replacement of the value by modeling information and recognition of engine «fault» (reconfiguration), which is the general theoretical, regardless of the engine model is used. Therefore, improvement of models’ level does not automatically lead to ACS reliability increase. And for successful identification of faulty information channel and fault of the engine component to its substitution in the ACS by use of on-board mathematical model, the important feature is the adaptability to the mentioned state of the object changes.
The linear adaptive onboard engine model (LABEM) is proposed. LABEM is designed for work in conjunction with the aircraft turbofan engine control system in a real environment and satisfy the requirements for compactness, speed and accuracy of the display parameters of the engine in statics and dynamics in a wide the range of operating modes, flight and engine conditions.
The static model is based on the throttle characteristics of the individual engine. The throttle characteristics was obtained in the acceptance tests or «race» in the operation after the service. The lower level dynamic linear mathematical model of a gas-turbine engine is obtained by state space method.
The most important for LABEM reliability is the validity of input information: atmospheric pressure and temperature at the inlet of the engine, fuel flow. The additional logic blocks for checking the input information was proposed. The developed algorithms validate the input parameters’ measurement with LABEM use and possible uncontrolled fault of one of the s at the LABEM input. The algorithms are based on dual channel sensors’ measurements of atmospheric pressure and temperature, the rotor speed of the low and high pressure turbine, the outlet compressor pressure and the outlet gas turbine temperature. By calculating of analytical and interchannel deviation the sensor faults are identificated through there comparing with the selected threshold. For detecting the engine component fault a component fault signature (CFS) is calculated. CFS is also compared to a threshold.
To improve the reliability of the fuel circuit input information the Kalman-filtering algorithms with integrated fault detection and isolation logic for the measuring channels are used. The algorithms are based on the calculation of the fault signature as weighted sum of the squares of residuals (WSSR), which is compared with the selected threshold value.
The practical motor-stand testing of the developed fault-tolerant algorithms as part of LABEM was showed that the average relative error of the dynamics is 0.168 %. In statics at maximum flow rate GT = 3800kg/hr the error is reduced to 0.01%. The results are satisfied the modern ACS GTE accuracy requirements. All this confirms the efficiency and the practical value of the developed algorithms.
Panov V. Auto-tuning of real-time dynamic gas turbine models //
Proceedings of ASME Turbo Expo 2014: Turbine Technical Conference and
Exposition (June 16 – 20, 2014, Düsseldorf, Germany). GT2014-25606. 2014. 10 p.
Kong Ch., Kang M., Koh S., Park G. A study on practical condition monitoring system for
2-spool Turbofan Engine using artificial intelligent algorithms // American
Institute of Aeronautics and Astronautics. ISABE-2013-1328. 2013. 9 p.
Malloy D.J., Webb A.T., Kidman D. F-22/F119
Propulsion System Ground and Flight Test Analysis Using Modeling and Simulation
// Proceedings of
ASME Turbo-Expo 2002 (June 3-6, 2002, Amsterdam, Netherlands). GT2002-30001.
2002. 8 p.
Avgustinovich V.G i dr. Identifikacija sistem upravlenija
aviacionnykh GTD / pod obshhejj redakciejj V.T.
Dedesha. M.: Mashinostroenie. 1984. 196 s.
Avgustinovich V.G., JAkushev A.P. Statisticheskijj analiz processov v sistemakh upravlenija
aviacionnykh dvigatelejj s bortovymi matematicheskimi modeljami //
Nauchno-tekhnicheskijj vestnik Povolzhja. 2014. № 35. S. 236-238.
Kobayashi T., Simon D.L. Aircraft Engine On-Line
Diagnostics Through Dual-Channel Sensor Measurements: Development of Baseline
System // Proceedings of
ASME Turbo-Expo 2008 (June 9-13, 2008, Berlin, Germany). GT2008-50345. 2008. 13 p.
Kuznecova T.A., Gubarev
E.A., Likhacheva JU.V.Algoritmy optimalnojj filtracii v zadachakh identifikacii parametrov sistem
avtomaticheskogo upravlenija aviacionnykh dvigatelejj //
Informacionno-izmeritelnye i upravljajushhie sistemy. 2014. T. 12. № 9. S. 12-20.
Kobayashi T., Simon D.L. Application of a bank of Kalman
filters for aircraft engine fault diagnostics // Proceedings of ASME Turbo Expo 2003. GT2003–38550.
Borguet S., L´eonard O. A sensor-fault-tolerant
diagnosis tool based on a quadratic programming approach// Proc. of ASME Turbo
Expo 2007. GT2007–27324.
Kobayashi T., Simon D.L. Integration of on-line and
off-line diagnostic algorithms for aircraft engine health management // Proceedings of ASME Turbo Expo 2007. GT2007-27518. 11 p.