RU2559401C1 - Diagnostic of aircraft service systems operating conditions - Google Patents

Abstract

FIELD: aircraft engineering.

SUBSTANCE: claimed method comprises comparison with availability factor K_ava with its threshold magnitude. For every aircraft service system SS, a required set of diagnostics components is composed by of neuronet analytical-simulation model of service systems, controller of SS signals parameters departure, controller of SS operating conditions and controller of SS diagnostics attributes array. Input signals are fed simultaneously to SS and analytical-simulation model inputs. Output signals of said system are fed to inputs of controller of SS signals parameters departure to define the difference in signals. The latter are transmitted to inputs of controller of SS operating conditions to compute partial factors of stability K_spf. Magnitudes of the latter are transmitted to inputs of controller of SS diagnostics attributes array, as well as other indices of reliability, longer life, etc, with the help of recommendations and formulae specified by GOST 27002-89 and GOST R 53111-2008. Computed magnitudes are used to generate the array of diagnostics data to be written in memory of controller of SS diagnostics attributes array.

EFFECT: enhanced diagnostics of aircraft service systems.

1 dwg

Description

The invention relates to the field of aerospace industry and can be used for continuous non-destructive testing and diagnosing the technical condition of the aircraft's service systems (LA) throughout the entire period of its operation.

A number of methods are known for monitoring and diagnosing the technical condition of service systems (“A method for monitoring and evaluating the technical condition of a multi-parameter diagnostic object according to measurement information” / RU 2145735, published on 02.20.2000, - [D1]; “A method for dynamically monitoring deadlocks of information and communication systems and device for its implementation ”/ Patent No. 2502123 dated 12/20/2013 - [D2]).

As an analogue of the invention, the technical solution described in D1 is selected. A disadvantage of the analogue is the absence of any reference model of a multiparameter object (service system). The technical condition is diagnosed by comparing the current state of the multi-parameter diagnostic object and the a priori composed structural-functional state diagram of structural elements constructed from basic structural descriptions. This method does not take into account the dynamics of changing states of both external disturbing factors and the aircraft's service systems.

Closest to the claimed invention in essence, the technical solution is “A method for dynamic monitoring of deadlocks of infocommunication systems and a device for its implementation” [D2]. The specified technical solution has a similar purpose as stated, and therefore selected as a prototype.

- интенсивности отказов критического технического ресурса r_iт, $$h_{r_{jп}}$$

- интенсивности отказов r_jп критического программного ресурса, размера q зоны буферной памяти узла, транслируют на пункт контроля определенные значения $${\lambda }_{r_{iт}},$$

$$h_{r_{jп}}$$

и q, задают на пункте контроля значение t_внп - временного интервала планируемого выполнения процессов и вычисляют коэффициент готовности - К_г по формуле:In the prototype, the technical result is achieved by the fact that dynamic control of the infocommunication system is carried out, while the values are determined: $${\lambda }_{r_{it}}$$

- the failure rate of the critical technical resource r _i $$h_{r_{jP}}$$

- the failure rate r _{jп of the} critical program resource, the size q of the zone of the buffer memory of the node, transmit certain values to the control point $${\lambda }_{r_{it}},$$

$$h_{r_{jP}}$$

and q, set at the control point the value of t _GNP - the time interval of the planned process execution and calculate the availability factor - K _g by the formula:

Where:

- математическое ожидание интенсивности отказов критического технического ресурса r_iт; $${\lambda }_{r_{it}}$$

- the mathematical expectation of the failure rate of the critical technical resource r _i ;

i = 1, 2, 3, ... - the current index of the critical technical resource r _it ;

t _vnp - the time interval of the planned execution of processes;

- интенсивность отказов критического программного ресурса r_jп; $$h{r}_{{}_{jP}}$$

- the failure rate of the critical software resource r _jп ;

j = 1, 2, 3, ... - the current index of the critical software resource r _jп ;

k is the order of the approximating Erlang distribution with the parameter @ _e - the intensity of the Poisson stream to the node for an integer value of the size q of the buffer zone of the system node;

n = 1, 2, 3, ... - the current index of the buffer memory zone in the system;

N is the total number of buffer memory zones in the system,

compare the availability factor determined in the dynamics of the system’s functioning - K _gtr with a threshold level of K _{gtr (0)} and when the condition is met:

K _gtr <K _{gtr (0)}

conclude that there are deadlocks in the service system.

The disadvantage of the prototype is the lack of completeness of diagnosis. However, the use of the prototype and the implementation on its basis of a specific algorithm and the necessary functions of the diagnostic process using the proposed formulas and formulas from GOST allow us to obtain the technical result claimed in the invention.

The basis of any diagnostic method is a diagnostic model of a controlled object, i.e. a formal description of the aircraft’s operational state, which allows to model and calculate the values of its output parameters according to known values of the input parameters (GOST 20911-89 “Technical Diagnostics. Terms and Definitions” - [D3]).

To solve such problems, the so-called regression models are widely used. Such a model is a functional dependence (regression equation) that describes the change in the output parameters of an object when changes in its input parameters are known.

A diagram of the process of diagnosing the technical condition of aircraft service systems built using regression and neural network analytical and simulation models is presented in the drawing. The technical literature provides descriptions of models and applied methodological approaches to the implementation of diagnostic procedures (Guidelines / “Technical Diagnostics. Determining the Technical State of Diagnostic Objects by Indirect Parameters Based on Regression Models.” / RD 50-491-84 / Approved by the Resolution of the State Standard of 24 July 1984, No. 2578 - [D4]; Ivanov AI / “The Subconscious of Artificial Intelligence: Programming Automatons of Neural Network Biometry with the Language of Their Training” / Russia, Penza - 2012, Elek throne book, publishing houses of OJSC PNIII - 125 pp. - [D5]).

When choosing the mathematical apparatus of diagnostic models for solving practical problems of controlling the technical state of an aircraft, it is necessary to take into account not only the accuracy of modeling, but also the convenience of its implementation in user applications. At the same time, they must be provided: the possibility of retraining as experimental data accumulate, invariance with respect to the features of the functioning of the diagnostic object, the number of its input parameters, etc.

In this connection, the apparatus for artificial neural networks (ANN or just neural networks), which allows to solve most of the tasks, and the neural network analytical and simulation model of the aircraft service systems based on these principles for implementing the diagnostic method, are of interest for operational monitoring and diagnosing the technical condition of aircraft service systems proposed in this invention.

A neural network is a set of artificial neurons that are connected in a certain way to each other and to the external environment using connections determined by weighting factors. The properties and capabilities of a neural network depend on its structure and types of functions implemented by a nonlinear converter. Most of the problems of technical diagnostics are successfully solved with the help of multilayer direct distribution neural networks in which artificial neurons are arranged in layers (Amosov OS “Intelligent information systems. Neural networks and fuzzy systems”: Textbook. Manual / OS Amosov. - Komsomolsk- on Amur: GOUVPO KnAGTU, 2004. - 104 p. - [D6]).

In order for the operation of a neural network (i.e., the reaction of its outputs to a change in inputs) to exactly correspond to the work of a real aircraft service system, it is trained, i.e. select the weight values so that the difference between the values of the network outputs and the corresponding outputs of the real aircraft service system is minimal. For training, the values of the input and output parameters of a real aircraft, measured during its operation (the so-called training sample), are used. Network training is carried out automatically using special training algorithms (V. A. Shustov / “Acceleration of neural network training with the selection of training examples” / Samara State Aerospace University, Institute of Image Processing Systems RAS, International Center for Scientific and Technical Information, Computer Optics, issue 24 - 2002 - [D7]).

The apparatus of artificial neural networks allows us to solve the problem of assessing the performance of aircraft service systems as separate functionally separate systems (blocks), while the aircraft is considered as a system of interconnected service systems, each of which is modeled by a separate neural network.

That is, the diagnostic model of the aircraft in this invention is a neural network diagnostic complex, the connections between the neural networks of which correspond to the connections between the service systems of a real aircraft. (Magazine "Locomotive", No. 07-2012 / A.V. Grishchenko, "The apparatus of artificial neural networks for the diagnosis of a modern locomotive" - [D8]). If the output parameters go beyond the permissible limits or the object fails, the result of the complex’s operation will not only be a conclusion about its inoperative state, but also a determination of the place and cause of the emergency or the failure of the element or unit.

In accordance with GOST 27002-89 and GOST R 53111-2008, the functioning of any aircraft service system is evaluated through indicators of its stability (V. Kozlov et al. “On the stability of the functioning of infocommunication systems”, // Basic Research Journal. - 2009 - No. 5, p. 54-55 - [D9]). We clarify the following definitions for their unambiguous interpretation.

Definition 1. Under the stability of the functioning of the service system of an aircraft is understood its ability to perform its assigned functions with specified quality indicators under the influence of destabilizing factors and is characterized by the values of the partial and general stability factors.

Definition 2. The aircraft service system is considered resistant to the influence of a number of destabilizing factors if the overall stability indicator of its functioning is within the specified limits.

Definition 3. The general indicator of the stability of the functioning of the aircraft’s service system is the product of particular stability indicators, provided that the causes and mechanisms of the occurrence of destabilizing factors are independent of each other, and the degradation of any module of any aircraft’s service system occurs independently of each other.

Definition 4. Under a private indicator of the stability of the functioning of the service system of an aircraft is understood the stability coefficient K _uchk , determined from the expression:

Where:

Y _tk - the required value of the quality indicator of the functioning of the spacecraft service system;

Y _k - the value of the indicator of the quality of the functioning of the service system of an aircraft under the influence of a destabilizing factor.

An object of the invention is the complete diagnosis of the technical condition of M service systems of the aircraft, where M = 1, 2, 3 ....

The technical result of the invention is achieved by carrying out the completeness of the diagnostic process based on the application of control procedures using the introduced controllers and the new mathematical models given in the description of the invention, which provide the calculation of stability coefficients for diagnosing the technical condition of aircraft service systems, as well as using those given in GOST 27002-89 and GOST R 53111-2008 formulas for calculating the parameter values of the recommended indicators of reliability, reliability, duty sheep.

вычисляют величины отклонений значений выходных сигналов служебной системы летательного аппарата Y_k, где k=1, 2, 3, …, от значений соответствующих эталонных сигналов Y_TPk, где k=1, 2, 3, …, нейросетевой аналитико-имитационной модели служебной системы летательного аппарата, передают вычисленные величины отклонений на входы контроллера анализа технического состояния служебной системы летательного аппарата, где по формуле $${К}_{учk}=1-\frac{\left|Y_{{{}_{TP}}_k}-Y_k\right|}{Y_{{{}_{TP}}_k}}$$

вычисляют значения частных коэффициентов устойчивости К_учk, передают вычисленные значения частных коэффициентов устойчивости К_учk на входы контроллера формирования массива диагностических признаков служебной системы летательного аппарата, где вычисляют по приведенным в ГОСТ 27002-89 и ГОСТ Р 53111-2008 формулам значения параметров рекомендуемых показателей надежности, безотказности, долговечности, ремонтопригодности, видов резервирования, в том числе и коэффициента готовности К_гтр, формируют из всех вычисленных значений параметров и показателей массив диагностических признаков, который записывают в буферную память контроллера формирования массива диагностических признаков служебной системы летательного аппарата.The essence of the invention lies in the fact that, in contrast to the known technical solution, form for each of M, where M = 1, 2, 3 ..., the aircraft’s service systems, the necessary set of diagnostic components from the neural network analytical and simulation model of the aircraft’s service system, the controller deviations of the signal parameters of the aircraft’s service system, the controller of the analysis of the technical condition of the aircraft’s service system and the controller of the formation of an array of diagnostic signs with uzhebnoy aircraft system, wherein the plurality of real input signals X _k, where k = 1, 2, 3, ..., is supplied simultaneously both to the inputs of performance of the aircraft system and the inputs of the neural network analytical and simulation model aircraft service system, from the outputs which signals Y _k , where k = 1, 2, 3, ..., of the aircraft’s service system and reference signals Y _TPk , where k = 1, 2, 3, ..., of a neural network analytical and simulation model of the aircraft’s service system, are fed to the controller inputs signal deviations s of the service system of the aircraft, where according to the formula $$\left|Y_{TPk}-Y_k\right|$$

calculate the deviations of the values of the output signals of the aircraft service system Y _k , where k = 1, 2, 3, ..., from the values of the corresponding reference signals Y _TPk , where k = 1, 2, 3, ..., the neural network analytical and simulation model of the service system the aircraft, transmit the calculated values of the deviations to the inputs of the controller analysis of the technical condition of the service system of the aircraft, where according to the formula $${\mathrm{TO}}_{\mathrm{at}hk}=\mathrm{one}-\frac{\left|Y_{{{}_{TP}}_k}-Y_k\right|}{Y_{{{}_{TP}}_k}}$$

calculating values of the partial stability coefficients K _uchk transmit calculated values of the partial coefficients _uchk resistance to the inputs of the controller forming the array of diagnostic features of the aircraft system performance, is calculated by where a given GOST 27002-89 and GOST P 53111-2008 formulas values recommended indicators reliability parameters, reliability, durability, maintainability, redundancy species, including the availability factor K _GAD is formed from all calculated values of parameters and azateley array of diagnostic features that are recorded in the buffer memory formation controller of diagnostic performance characteristics of the aircraft system.

The claimed method for diagnosing the technical condition of the service systems of the aircraft is illustrated in the drawing, which adopted the following notation:

1 - X ₁ , the first input signal received at the first input of the aircraft service system;

2 - X ₂ , the second input signal supplied to the second input of the aircraft service system;

3 - X ₃ , the third input signal supplied to the third input of the aircraft service system;

4 - X _k , k-th input signal supplied to the k-th input of the aircraft service system;

5 - X _k , k-th input signal supplied to the k-th input of a neural network analytical and simulation model of the aircraft’s service system;

6 - X ₃ , the third input signal supplied to the third input of a neural network analytical and simulation model of the aircraft's service system;

7 - X ₂ , the second input signal supplied to the second input of a neural network analytical and simulation model of the aircraft service system;

8 - X ₁ , the first input signal supplied to the first input of a neural network analytical and simulation model of the aircraft’s service system;

9 - service system of the aircraft;

10 - neural network analytical and simulation model of the aircraft service system;

11 - M, the number of service systems of the aircraft;

12 - M, the number of neural network analytical and simulation models of aircraft service systems;

13 - Y ₁ , the first output signal from the first output of the aircraft service system;

14 - Y ₂ , the second output signal from the second output of the aircraft service system;

15 - Y ₃ , the third output signal from the third output of the aircraft service system;

16 - Y _k , k-th output signal from the k-th output of the aircraft service system;

17 - Y _TPk , k-th output reference signal from the k-th output of a neural network analytical and simulation model of the aircraft’s service system;

18 - Y _TP3 , the third output reference signal from the third output of the neural network analytical and simulation model of the aircraft service system;

19 - Y _TP2 , the second output reference signal from the second output of the neural network analytical and simulation model of the aircraft service system;

20 - Y _TP1 , the first output reference signal from the first output of the neural network analytical and simulation model of the aircraft service system;

21 - controller deviations of the parameters of the signals of the service system of the aircraft;

22 - M, the number of controllers of deviations of the signal parameters of the aircraft service system;

, величина отклонения k-го выходного сигнала служебной системы летательного аппарата от k-го выходного эталонного сигнала;23 - $$\left|Y_{TPk}-Y_k\right|$$

, the deviation of the k-th output signal of the aircraft service system from the k-th output reference signal;

, величина отклонения третьего выходного сигнала24 - $$\left|Y_{TP3}-Y_k\right|$$

, the deviation value of the third output signal

service system of the aircraft from the third output reference signal;

, величина отклонения второго выходного сигнала25 - $$\left|Y_{TP2}-Y_k\right|$$

, the deviation value of the second output signal

service system of the aircraft from the second output reference signal;

, величина отклонения первого выходного сигнала26 - $$\left|Y_{TP\mathrm{one}}-Y_k\right|$$

, the deviation value of the first output signal

service system of the aircraft from the first output reference signal;

27 - controller analysis of the technical condition of the service system of the aircraft;

28 - M, the number of controllers analyzing the technical condition of the aircraft service system;

29 - K _uchk , k-th _partial coefficient of stability on the k-th signal;

30 - To _uch3 , the third _partial stability coefficient for the third signal;

31 - To _uch2 , the second _partial stability coefficient for the second signal;

32 - To _uch1 , the first _partial stability coefficient for the first signal;

33 - a controller for generating an array of diagnostic features of the aircraft service system;

34 - M, the number of controllers forming an array of diagnostic features of the aircraft service system.

Information confirming the possibility of carrying out the invention is given in the sources D3-D9, as well as in the article "On the diagnostics of the spacecraft service systems", D10, the materials of the source D10 are placed in the Appendix to the application materials.

Note

The applicant placed in the Appendix to the application materials additional information confirming the possibility of carrying out the invention so as not to unnecessarily overload the description of the invention. However, if the examination considers it appropriate, the applicant will not object to the inclusion of the Application in the description.

Claims (1)

A method for diagnosing the technical condition of the aircraft's service systems, which consists in calculating the availability factor (K _gtr ) by the formula:

Where:
$$\lambda r_{it}$$

- the mathematical expectation of the failure rate of the critical technical resource r _i ;
i = 1, 2, 3, ... - the current index of the critical technical resource r _it ;
t _vnp - the time interval of the planned execution of processes;
$$h{r}_{jP}$$

- the failure rate of the critical software resource r _jп ;
j = 1, 2, 3, ... - the current index of the critical software resource r _jп ;
the k - order approximating Erlang distribution with parameter @ _er - intensity of a Poisson flow node for a value of size q of the buffer memory system node zone;
n = 1, 2, 3, ... - the current index of the buffer memory zone in the system;
N is the total number of buffer memory zones in the system,
comparing the calculated readiness coefficient K _gtr with a threshold level of K _{gtr (0)} , characterized in that for each of M, where M = 1, 2, 3 ..., the aircraft’s service systems, the necessary set of diagnostic components is formed from the neural network analytical and simulation service model system of the aircraft, controller of deviations of the signal parameters of the aircraft’s service system, controller of the analysis of the technical condition of the aircraft’s service system and the diagnostic array formation controller x signs service system of the aircraft, wherein the plurality of real input signals X _k, where k = 1, 2, 3, ..., is supplied simultaneously both to the aircraft inputs of performance of the system and the inputs of the neural network analytical and simulation model of the aircraft performance system, from the outputs of which signals Y _k , where k = 1, 2, 3, ..., of the aircraft’s service system and reference signals Y _TPk , where k = 1, 2, 3, ..., of a neural network analytical and simulation model of the aircraft’s service system, are fed to steam deviation controller inputs etrov signaling overhead system of the aircraft, wherein the formula $$\left|Y_{TPk}-Y_k\right|$$

calculate the deviations of the values of the output signals of the aircraft service system Y _k , where k = 1, 2, 3, ..., from the values of the corresponding reference signals Y _TPk , where k = 1, 2, 3, ..., the neural network analytical and simulation model of the service system the aircraft, transmit the calculated values of the deviations to the inputs of the controller analysis of the technical condition of the service system of the aircraft, where according to the formula $${\mathrm{TO}}_{\mathrm{at}hk}=\mathrm{one}-\frac{\left|Y_{{{}_{TP}}_k}-Y_k\right|}{Y_{{{}_{TP}}_k}}$$

calculate the values of the partial stability coefficients K _uchk , transfer the calculated values of the partial stability coefficients K _uchk to the inputs of the controller for forming an array of diagnostic features of the aircraft’s service system, where, according to the formulas given in GOST 27002-89 and GOST R 53111-2008, the parameter values of the recommended reliability indicators, reliability, durability, maintainability, redundancy species, including the availability factor K _GAD is formed from all calculated parameter values and azateley array of diagnostic features that are recorded in the buffer memory formation controller of diagnostic performance characteristics of the aircraft system.

Images