Integral estimation of aircraft systems condition based on reference regions

Аuthors

Evdokimenkov V. N.^1*, Kim R. V.^2**, Voronov A. A.^3***, Vekshina A. B.^3****

1. ,
2. ,
3. Moscow Aviation Institute (National Research University), 4, Volokolamskoe shosse, Moscow, А-80, GSP-3, 125993, Russia

*e-mail: evdokimenkovvn@mai.ru; vnevdokimenkov@gmail.com
**e-mail: romanvkim@yandex.ru
***e-mail: woron_al@mail.ru
****e-mail: mopy_ann@mail.ru

Abstract

This paper presents an approach to the problem of aircraft systems technical state estimation method. This method uses information stored by regular on-board flight information monitoring systems. The principle of the method is based on reference pattern formed by way of finite set of points within the space of state variables of controlled on-board system for every generic flight mode upon condition of its normal operation. To estimate current aircraft system condition considering non-controlled factors action including factors of accidental action, we use probability criterions that characterize probability of current controlled parameters set membership to reference pattern. In this work we suggest computational algorithms for such probability criterions. These algorithms intend to apply Mahalanobis distance from reference region points to its geometric center as a variability measure of controlled system parameters values. Estimation of the probabilistic criteria is performed using Mahalanobis distance distribution function. This function is obtained by approximating Mahalanobis distance cumulative distribution function which can be calculated based on the information about previous flights of the exact aircraft. With the help of distribution function we can calculate the probability of occurrence Mahalanobis distance value exceeding the argument of the function in case of normal operating of controlled system. We define this probability as a main probabilistic criteria. We also define an inversed probabilistic criteria as a quantile of the distribution function for given probability value. In this case quantile is a Mahalanobis distance value that will not be exceeded with a given probability value. Developed algorithms allow us to identify the moment of abnormal operation mode occurrence. To solve this task we use inverse probabilistic criteria. This moment occurs when the current Mahalanobis distance calculated using the current flight data exceeds pre-determined quantile value.

Developed methods and algorithms were used to estimate Su-35 aircraft conditioning system state. The analyzed period encompassed 50 flights, beginning from 80^th to 129^th, carried out during two years of exploitation. By using the proposed method an abnormal performance in the temperature channel of air conditioning system was detected in 2 out of 50 flights (105^th and 106^th). Obtained results fully correspond to the state of the system fairly confirmed by technical inspection report. According to the report the breakdown of turbo-compressor due to inadmissible overheating of a bearing was registered during 106-th flight.

Keywords:

aircraft system, aircraft system reference region, probabilistic criteria, Mahalanobis distance, environment control system

References

1. Evdokimenkov V.N., Krasil’shchikov M.N., Ratnikova N.I. Izvestiya RAN. Teoriya i sistemy upravleniya, 2003, no. 6, pp 38-46.
2. Krasil’shchikov M.N., Evdokimenkov V.N., Bazlev D.A. Individual’no-adaptirovannye bortovye sistemy kontrolya tekhnicheskogo sostoyaniya samoleta i podderzhki upravlyayushchikh deistvii letchika. (Individually-adapted onboard systems of control of technical condition of aircraft and support for pilots control actions), Moscow. MAI, 2011, 440 p.
3. Krasil’shchikov M.N., Evdokimenkov V.N., Kim R.V. Vestnik komp’yuternykh i informatsionnykh tekhnologii, 2012, no. 2, pp. 13-19.
4. Krasil’shchikov M.N., Evdokimenkov V.N., Kim R.V. Programma monitoringa tekhnicheskogo sostoyaniya letatel’nogo apparata i ego bortovykh sistem v protsesse ekspluatatsii (Functional software prototype of flying vehicle technical state and its systems monitoring complex in operation). The certificate on the state registration of the computer program RU 2012611202, 30.01.2012.

Download