CC BY
22
Civil Aviation High Technologies
Vol. 23, No. 05, 2020
UDC: 629.735.33+351.814.2
DOI: 10.26467/2079-0619-2020-23-5-54-66
METHOD OF OPTIMIZATION OF DECISION-MAKING DURING MANAGEMENT OF SAFETY OF FLIGHTS IN THE ACTIVITIES OF
OPERATORS OF AERODROMES
S.A. TOLSTYKH1
Moscow State Technical University of Civil Aviation, Moscow, Russia
The study was conducted with the financial support of the Russian Foundation for Basic Research,
grants № 19-38-90215
In modern conditions of limited budget for enterprises of aerodrome operators, the task of optimizing decision making in flight safety management is becoming extremely urgent. Management decisions, which are a safety management tool, must be not only effective in terms of expected improvements in safety, but also cost-effective and appropriate for the enterprise. Optimization in this article should be understood in terms of the mentioned criteria. The article presents a method for supporting management decision-making as part of a safety management strategy for the activities of aerodrome operators. In the presented methodology, an important place is given to indicators of the level of safety of flights and their use in making managerial decisions. Along with the safety indicator, an indicator of financial damage from recorded events is used, which is calculated in value terms taking into account direct and indirect damage to the aerodrome operator. Regression modeling is used in conjunction with the decision-making technique of "human-machine procedures". Regression analysis is performed using STATISTICA software, and allows you to identify the dependence of indicators on the degree of influence of hazard factors. The resulting model, based on data from last year, makes it possible to forecast the values of indicators for the next. Using the decision-making methodology of "human-machine procedures", an assessment is made of the priority of implementing managerial decisions based on an integrated criterion. The methodology ensures compliance with the requirements of Russian and international air legislation for operators of certified aerodromes. The scope of its application can be expanded to SMS of all aviation service providers, taking into account the relevant specifics of the services provided and the existing hazard factors.
Key words: safety risk, indicator, aerodrome operator, optimization, regression analysis, forecast, management decisions.
INTRODUCTION
In 2020 the civil aviation encountered serious challenges concerning its regular activity. According to the Russian Federation Government Regulation № 4341 the air transportation and airport activity suffered significant impact from the COVID 19 pandemic. It is of utmost importance to consider the balance performing the flight safety management tasks2 in such conditions, and when the situation becomes stabilized as well.
Aerodrome operators in crisis face the acute budget deficit as a consequence to the decreased passenger flow3. The term "airport operator" implies "a person owing an airfield or a helipad under the right of ownership, by lease or any other legal ground and using this airfield or helipad to provide aircraft take off, landing, taxiing and parking"4. Nevertheless, flight safety level should be constantly improved or, at least, maintained, as the present situation in the world should not affect it. Flight safety should not be compromised due to budget issues. Figure 1 (borrowed from SMM ICAO Doc 9859) shows the abstract line of the boundaries of the safety space. The safety space decreases with the insufficient funding, which results in the flight safety deterioration.
1 Regulation No. 434 of the Government of the Russian Federation On Adopting the List of Industries in the Russian Economy Most Severely Affected by Deteriorating Conditions Resulting from Proliferation of the New Coronavirus Infection. (2020). 5 p.
2 Safety Management Manual (SMM) ICAO Doc 9859, 4th Edition. (2018). 218 p.
3 According to international airport association data. MMA. Available at: http://interairports.ru/ (accessed 19.07.2020). (in Russian)
4 Federal Law № 60 -FZ 19 March 1997 "The Air Code of the Russian Federation" (1997). 60 p.
Vol. 23, No. 05, 2020
Civil Aviation High Technologies
Production
Fig. 1. ICAO Safety Flight Management Balance (borrowed from the ICAO SMM. Doc.9859, 4th ed. 2018.)
At the same time, Figure 1 shows that flight safety overfunding may seriously affect the production profitability thus resulting in the enterprise bankruptcy. It is obvious, that the flight safety funding should be balanced (the management dilemma), ICAO SMM does not state the methods for balancing the funding.
Managerial decisions have always been the main tool for flight safety management. Such solutions made by the airdrome authorities are the grounds for the measures aimed at diminishing the danger factors in the aerodrome activity.
The mathematical methods for the managerial decisions effectiveness assessment are shown in [1]. However, the solution to our problem is not possible if only one criterion, the effectiveness, is taken into consideration. The managerial decision priority ranking and feasibility have to be evaluated economically. According to the decision theory [2, 3], this is a two-criteria task.
TERMS OF REFERENCE
At present there are different methods for two-criteria task optimum solutions, shown, for example in [3, 4, 5]. The most frequent way is to limit the number of criteria to one. The result is achieved applying the convolution method, global criterion method, threshold criterion method, distance method. As it was fairly stated in [6], the mentioned methods are not strictly justified, thus the application of them is determined by the conditions of the problem and the preferences of the decision maker.
Alongside with the abovementioned methods, Edgeworth-Pareto principle, Nash principle and man-machine decision making tools are being applied. The latter three are of most interest, as they do not imply a predetermined optimal decision scheme, and allow to keep all the obtained data [6].
The first criterion. In order to determine the decision-making effectiveness, it is appropriate to apply the flight safety record based on the previous events. Any record, typically evaluated by the enterprises can serve as an example for that, for instance the ones described in papers [7-12], or others. The present paper will be using the criterion conditionally called Fspi (flight safety performance indicator). As it has been established within the industry, this indicator actually shows the hazard level during the flight, so the task for the first criterion is to minimize the first indicator.
The second criterion. The method implies the financial indicator as the second criterion. It shows the numeric value of the cost equivalent in provisional monetary units (used to eliminate any national currency as a reference) required to implement the managerial decision. The paper suggests calling the criterion "the financial loss indicator" (Fli). Likewise, the second criterion must also be minimized.
Civil Aviation High Technologies
Vol. 23, No. 05, 2020
In order to evaluate the factors for both criteria separately, it is appropriate to use statistical analyses and modelling. The model allows to forecast the criteria values in case the featuring operational conditions change.
REGRESSION MODEL FORMATION
Regression modelling is one of the statistical analyses and forecast methods. The variety of its application is presented in the summarizing paper [13].
The main idea of the regression analyses is in composing the regression equation. Such an equation demonstrates the dependence of one variable (dependent) from several other dependent variables.
Let us take Fspi and Fli as the feedback (dependent variable) (further -Indicators). Every indicator would require its own regression equation. It is also necessary to determine independent variables. In practice, there can be numerous factors influencing the indicators, and it is impossible to take all of them into consideration. For the sake of simplicity, it is suggested using Generalized Hazard Rate or Generalized Hazard. Hazard assumes "the result of the action or absence of action, circumstance, condition or their combinations, which affect civil aircraft flight safety" as it was defined in the Order No. 1215 of the Government of the Russian Federation5.
Generalized Hazard degree of impact onto the previous events may be evaluated in conditional units(points).
The regression equation general arrangement:
n = PQ +P1&1 + ... + i +... + pn0n (1)
where: n - indicator, dependent variable of the regression model;
Oi - Generalized Hazard (GH), independent variables;
Pi - regression coefficients.
There are six GH areas suggested, the GH areas are structured depending on the activity types. The activity types were determined during the ISAGO6'7 Ground Operation Safety Audit and by aerodrome activity types listed in GOST (National Standard of the Russian Federation)8.
The research suggests the following GH areas:
• AS - Aerodrome support
• SET - Special equipment and transport
• ATC - Air traffic control
• OP - Ornithological provision
• BCH - Baggage cargo handling
• SEC -Security
The listed hazard areas are generalized according to the aerodrome operator activity types. Other types of activity are cases, which can this way or another be related to one of the listed above. Fire-fighting, SAR and emergency services are not listed deliberately, as their activities are intended to diminish the consequences of the hazards.
5 About procedure for development and application of safety management systems of flights of air vehicles, and also collection and data analysis about the factors of danger and risk creating safety hazard of flights of civil air vehicles, storages of these data and exchange of them. Order of the Government of Russian Federation November 18, no. 1215. (2014). 3 p.
6 IATA Safety Audit for Ground Operation (ISAGO). Standards Manual. 7th Edition, Effective February 2018. 350 p.
7 IATA Safety Audit for Ground Operation (ISAGO) Guidelines on auditing a Safety Management System. Based on GOSM 5th Edition, Effective July 2016. 370 p.
8 GOST R 57239-2016 (National Standard of the Russian Federation) Aviation Activity Safety Management System. Database. Aviation infrastructural risks caused by aerodrome activity, Moscow, Standartinform. (2016). 43 p.
Vol. 23, No. 05, 2020
Civil Aviation High Technologies
Consequently, the application of the method is explained using real-world analogous fictional data of one of the Russian airports. The initial data used are the indicators (Fspi and Fli) for the previous year, and each GH degree of impact onto the events of the month, which took place last year. The degree of GH impact is determined by the experts, conversant with the circumstances the reasons of the events subject to expertise.
Every event, which happened the previous year is to be rated using a 10-point scale (with 1 point which stands for minimum or indirect GH impact on the event and 10 points that stand for the event completely resultant from this very GH). The results are to be grouped by months. The fragment of the possible evaluation is shown in Figure 2.
№ Date Event circumstances Previous year indicators Generated Hazards and their impact expert evaluation
F ■ i spi Fli AS SET ATC OP BCH SEC
1 11.01 Airport security vehicle is inoperative, substituted by a follow me car 1500 1 3 3
2 17.01 Standardized lighting markers on the TWY partially covered with snow 600 6 1
3 29.01 Incursion of the active TWY by a luggage trolley (flight delay) 2000 2 8
January total 0,754 4100 7 3 3 0 8 3
4 12.02 SAR portable radio stations transmission inoperable 0 1
5 20.02 Dog on the apron 0 1 1 1 1
6 25.02 A small bird in NLG bay at takeoff 6000 1 6
February total 0,515 6000 1 0 3 7 0 1
Fig. 2. A fragment of the source data (example)
Figure 2 also shows the actual values of the indicators, calculated using the methods accepted by the certain aerodrome operator.
Considering the data obtained, the initial data for the regression model are formulated (fig. 3). In order to perform further regression modelling operations STATISTICA 13.5.0.17 (English version ) software package is used and instruction cited in [14] are applied.
Научный Вестник МГТУ ГА_Том 23, № 05, 2020
Civil Aviation High Technologies Vol. 23, No. 05, 2020
Data: Month-Fspi-Fli-GHF* f8vby12c) I— Швш
Month-Fspi-Fli-GHF m
1 2 3 4 5 6 7 8
Fspi Fli AS SET АТС OP BCH SEC
January 0.754 4100 7 3 3 0 8 3
February 0,515 6000 1 0 3 7 0 1
March 0.368 1200 о 1 4 4 0 0
April 0,294 0 1 0 1 0 2 0
May 0.405 5300 2 1 3 7 2 0
June 0,129 2000 0 1 5 0 0 0
July 0.147 500 0 0 3 0 2 1
August 0.129 800 0 0 4 2 0 0
September 0.423 6110 2 1 9 0 0 0
October 0,257 0 0 0 0 1 0 0
November 0.386 0 0 3 0 0 0 0
December 0,349 5590 1 4 6 0 0 0 v
Fig. 3. Initial data of regression models in the STATISTICA software package
Two basic linear regression models are constructed based on the data obtained. These regression models illustrate the dependence of every indicator on the GH.
The factor correlation analysis is no sense as the forecast will be based on monthly mean per year, without indicator values for a certain month. Therefore, multicollinear factors may occur in a model (linear dependence between the explanatory variables) [15].
The results of the regression model construction for the indicators Fspi and Fli are shown in Figure 4.
Итоги регрессионной модели для показателя Пбп R= .95044945 R? = 0.90335415
Скорректированный Adjusted R~ = 0.78737913
N=12 b* Std. Err. of b* b Std Err of b t(5) p-value
Intercept 0.2819911 0.064340 4.38279 0 007136
AS 1 44574 0461960 0.129607 0.041413 3.12959 0 025970
SET 0.18609 0.174195 0 023684 0.022169 1.06831 0.334220
ATC -0.30767 0.184020 -0,021643 0.012945 -1.67196 0.155395
OP 0.20519 0.163767 0 013403 0.010697 1.25292 0.265637
BCH -1.00523 0.492032 -0.077092 0 037734 -2.04302 0.096497
SEC 0.31380 0.293056 0 062253 0.058138 1.07077 0.333214
Итоги регрессионной модели для показателя Пфу R= 0 93457158 R2 = 0 87342403
Скорректированный Adjusted R~ = 0 72153287
N=12 b* Std. Err. of b* b Std Err of b t(5) p-value
Intercept ) -520.9471 1059.050 -0.49190 0 643623
AS 0.908097 0.528675 1170.890 681.668 1.71768 0.146499
SET 0.174757 0 199352 319.888 364.907 0.87663 0.420804
ATC 0.431384 0 210595 436 465 213.075 2.04841 0.095837
OP 0 428934 0.187418 402.978 176.077 2,28865 0.070769
BCH -0.760775 0.563090 -839 169 621.113 -1.35107 0.234588
SEC 0.166206 0.335378 474.248 956.958 0.49558 0.641202
Fig. 4. The results of the construction of regression models
Here, column "b" shows the regression coefficients, i.e. the constant and the coefficients for every indicator (for every independent variable).
Column "b*" shows the values for standardized coefficient (P).
The GH influence importance values are shown in "p-value" columns.
R2 - the coefficient of linear determination, which is the regression model adequacy value
Std. Err of b (b*) - the standard mean error of b (b*).
Том 23, № 05, 2020_Научный Вестник МГТУ ГА
Vol. 23, No. 05, 2020 Civil Aviation High Technologies
FEASIBILITY ASSESMENT OF THE MODEL
The adequacy and the application spectrum of the model should be evaluated to be able to apply the results obtained.
The determination coefficient for every model is shown in the head of the table (fig. 4) as "R2 .This coefficient may take the values from 0 to 1 and shows the number of factors considered within a model out of those that influence the dependent variable. For this case, the determination coefficients are quite fine (according to [16] if more than 0,8), which proves that the model will be more precise than a simple mean value forecast [17].
Also, STATISTICA software package allows to convey a technical analysis of residual. The theory of regression analyses defines residual as the actual data deviation values from the regression line. Let us analyze the residual for Fspi (likewise for Fli). Figure 5 shows the frequency histogram of residuals, and Figure 6 shows the probability plots - normal for the residuals.
Distribution of Standard residuals - Expected Normal
~7
/ / V
/ \
(.0 -0,5 0.0 0.5 1.0
Fig. 5. The frequency histogram of the distribution of residues
Fig. 6. Normal Probability Plot of Residuals
The given graphs show the normal residual plotting and also the absence of the actual data deviation from the theoretical normal line
Let us enter the monthly mean values of GH for the previous year (Mi mean) from the input data (fig. 3). The model shows the result of the mean indicator values, and the values for their confidence intervals ± 95 % (fig. 7).
Variable Predicting Values for (Месяц-Пбп-Пфу-ОФО) variable: Пбп Variable Predicting Values for (Месяц-Пбп-Пфу-ОФО) variable: Пфу
b-Weight Value b-Weight * Value b-Weight Value b-Weight * Value
AS 0.129607 1.166700 0 151212 AS 1170.890 1.166700 1366.078
SET 0.023684 1.166700 0.027632 SET 319.888 1.166700 373,213
ATC -0.021643 3.416700 -0.073949 ATC 436.465 3.416700 1491 269
OP 0.013403 1,750000 0.023455 OP 402.978 1,750000 705,211
BCH -0.077092 1 166700 -0 039944 BCH -839.169 1 166700 -979.058
SEC 0 062253 0,416700 0.025941 SEC 474,248 0.416700 197.619
Intercept 0 281991 Intercept -520.947
Predicted 0.346337 Predicted 2633,385
-95.0%CL 0 285221 -95.0%CL 1627.403
+95,0%CL 0.407454 +95,0%CL 1 3639.368 \
a b
Fig. 7. Average values and their confidence intervals a) Fspi; b) Fli
Civil Aviation High Technologies
Vol. 23, No. 05, 2020
The analyses of the model shown confirms its applicability within the methods of factor analyses, [18, 19] i.e. the maintenance factors for the aerodrome operators and determining the dependence of the variables. STATISTICA software package allows to construct regressions applicable for the method based on almost any initial data, which exist in practice.
THE APPLICAZTION OF THE MODEL FOR OPTIMIZATION ANY PRIORITY RANKING
OF MANAGERIAL DECISIONS
The flight safety management supposes annual planning. The experts are tasked to evaluate the percental decrease in the impact of every GH after the implementation of every managerial decision. The evaluation criteria are shown in Table 1.
Table 1
Criteria for expert evaluation of the impact of managerial decisions on generalized hazards
Percentile impact onto GH Comment
100 % Maximum impact onto GH. Eliminates the possibility of GHs after the managerial decision implementation
50 % After the managerial decision implementation, the possibility of GH is twice as low
5 % The managerial decision will contribute to the improvement of the only factor out of the total GH
3 % The managerial decision will moderately contribute to the improvement of the only factor out of the total GH
1 % The managerial decision will slightly contribute to the improvement of the only factor out of the total GH
0,5 % The minimum influence onto GH. The managerial decision will slightly contribute to the improvement of the only factor out of the total GH
For example, Table 2 presents the data about the possible managerial decisions showing their implementation costs (in provisional monetary units) and the percentile impact onto the GHs (the managerial decision efficiency coefficients - Kij). These coefficients must be assessed by the experts based on the criteria shown in Table 1.
Table 2
Expert assessments of the impact of management decisions on generalized hazards
The managerial decision Cost Provisional monetary units Efficiency coefficients (KIJ)
AS SET ATC OP BCH SEC
1 2 3 4 5 б l S
l.To renovate the RWY and TWY ground marking 4ОО О,О15 О,О35 О,О15 О,О1 О,ОО5 О,ОО5
2.To develop a data exchange software package 16О О О О,ОО5 О,О1 О,ОО5 О
3.To improve the ATC and ornithological service interaction technology 26О О О О,О2 О,О2 О О
4.To purchase the radio stations 5ОО О О,О35 О,О1 О,ОО5 О,О1 О,ОО5
Том 23, № 05, 2020_Научный Вестник МГТУ ГА
Vol. 23, No. 05, 2020 Civil Aviation High Technologies
Continuance of Table 2
1 2 3 4 5 6 7 8
5.To carry out refresher training for the interacting departments 300 0 0,035 0,005 0,005 0,005 0,005
6.To repair the RWY and TWY pavement 3000 0,015 0,015 0,005 0,05 0,005 0
7.To set the fence from the water body 250 0 0,005 0,02 0,01 0 0
8.To purchase the equipment (parts etc.) 1500 0 0,01 0,05 0,01 0 0
9.The airdrome trolley repair/renovation 400 0,005 0,005 0 0,01 0 0
Considering the data makes it possible to correct the monthly mean (for the previous year) coefficients for each GH contribution to the indicator after the implementation of each managerial decision, using the Formula 2.
Mj = Mi cp- Mi cp* Kj (2)
Where: Mi cp - the monthly mean GH for the previous year; Kij - managerial decisions efficiency coefficients. The new coefficients (My) are shown in Table 3.
Table 3
Contribution coefficients of generalized hazard factors considering the predicted effect of the implementation of each managerial decision
MD Coefficients М^
AS SET ATC OP BCH SEC
1 1,1492 1,1258 3,3654 1,7325 1,1608 0,4146
2 1,1667 1,1667 3,3996 1,7325 1,1608 0,4167
3 1,1667 1,1667 3,3483 1,7150 1,1667 0,4167
4 1,1667 1,1258 3,3825 1,7413 1,1550 0,4146
5 1,1667 1,1258 3,3996 1,7413 1,1608 0,4146
6 1,1492 1,1492 3,3996 1,6625 1,1608 0,4167
7 1,1667 1,1608 3,3483 1,7325 1,1667 0,4167
8 1,1667 1,1550 3,2458 1,7325 1,1667 0,4167
9 1,1608 1,1608 3,4167 1,7325 1,1667 0,4167
Using the models (fig. 4) made on the basis of the initial data for the previous year (fig. 3) enables to forecast the indexes for the following year considering each managerial decision implementation. The changes of the indexes AFspi and AFli may be calculated using the Formula (3):
Anfsj = n/s0 - n/s; Anflj = nfl0 - nflj; (3)
where nfs 0 / nfi 0 - upper fiducial limit (+95%CL) for the indexes not taking the implementation of managerial decisions into account (see fig. 7);
Научный Вестник МГТУ ГА_Том 23, № 05, 2020
Civil Aviation High Technologies Vol. 23, No. 05, 2020
nfs J / Па j - upper fiducial limit (+95%CL) for the indexes considering each managerial decision implementation.
As this is a two-criteria problem, let us introduce two criteria for the managerial decision efficiency:
с; ;С2 = f (4)
' Qj Q,
where Q, is the cost of the j - number managerial decision.
The values of the efficiency criteria are estimated for every managerial decision. Further, it is necessary to calculate the specific values, as it is recommended in [3]. For the criterion C; the specific value CiJ is calculated as (5):
С = Cij ~min Cij (5) max C{ - min C/
where maxC;J are minC;J the maximum and minimal values of the criterion.
The calculations for C2J criterion are likewise, as both criteria (indexes) are supposed to be diminished.
As the theory of man-machine procedures cites, the solutions of a multi-criteria problem of the decision-making theory require to determine weighted coefficients of the criteria importance wl and w2, that should sum to a one. For example, let us suppose that the airdrome operator sets the criteria priority as w1=0,6; w2=0,4. Then we can estimate the complex criteria О for every managerial decision using the formula (6):
Сj = Сij* W; + C2j* w 2 (6)
The results obtained from the calculations (formulae 3-6) are the grounds for the complex criteria evaluation. It is reasonable to rank the managerial decisions in descending order (tab. 4).
Table 4
Ranking managerial decisions by degree of optimization
№ Managerial decision Cost Complex criterion
l 1. To renovate the RWY and TWY ground marking 400 0,945
2 9. The airdrome trolley repair/renovation 400 0,519
3 5. To carry out refresher training for the interacting departments 300 0,474
4 3. To improve the ATC and ornithological service interaction technology 260 0,432
5 7. To set the fence from the water body 250 0,362
6 6. To repair the RWY and TWY pavement 300 0 0,352
7 4. To purchase the radio stations 500 0,282
8 8. To purchase the equipment (parts etc.) 150 0 0,236
9 2. To develop a data exchange software package 160 0,158
Vol. 23, No. 05, 2020
Civil Aviation High Technologies
Managerial decision ranking emphasizes their priority according to the complex criterion chosen. The result obtained may be applied by the person who makes decisions about the feasibility and priority of managerial decisions implementation.
CONCLUSION
The paper suggests the methods for the flight safety management decisions in aerodrome operators' activity. The research introduces the method of the managerial decision efficiency evaluation based on the two-criteria task of "man-machine procedures" using statistical analyses and the method of regression modelling.
The regression analyses allow to forecast the flight safety records based on the previous year data, considering the implementation of the new managerial decisions. The method considers the impact of the managerial decisions onto the flight safety, taking their feasibility into account.
Thus, the suggested method allows to rank the managerial decisions and may be applied as the grounds for the flight safety resources allocation.
STATISTICA software package is suggested as the problem solution tool. The method application does not require mathematical skill as the results may be computed using a personal computer.
The method described may be applied for inspections and flight safety departments of the aerodrome operators. The application area may be extended to Flight Safety Management Systems of all aviation enterprises regarding the services specifics and typical hazards.
The method upgrade to the level of practical application will require further research and trial based on the actual data.
REFERENCES
1. Khrustalev, S.A., Orlov, A.I. and Sharov, V.D. (2013). Mathematical methods for evaluation of the efficiency of management decisions. Zavodskaya Laboratoriya. Diagnostika Materialov, vol. 79, no. 11, pp. 67-72. (in Russian)
2. Orlov, A.I. (2006). Teoriyaprinyatiya resheniy: uchebnoye posobiye [Decision theory: Tutorial]. Moscow: Izdatelstvo «Ekzamen», 576 p. (in Russian)
3. Larichev, O.I. (2008). Teoriya i metody prinyatiya resheniy, a takzhe Khronika sobytiy v Volshebnykh stranakh. Uchebnik dlya VUZov [Theory and decision-making methods, as well as the Chronicle of events in the Magic countries: Textbook for Universities]. Moscow: Logos, 392 p. (in Russian)
4. Larichev, O.I. (2002). Properties of the decision methods in the multicriteria problems of individual choice. Automation and Remote Control, vol. 63, no. 2, pp. 304-315.
5. Orlov, A.I. (2018). Metody prinyatiya upravlencheskikh resheniy. Uchebnik [Management Decision Making Methods: Textbook]. Moscow: KNORUS, 286 p. (in Russian)
6. Romanov, V.N. (2012). Osnovy sistemnogo analiza: V.N. Romanov: uchebno-metodicheskiy kompleks [Fundamentals of systems analysis: V.N. Romanov: educational and methodical complex] St. Petersburg: Izdatelstvo Natsionalnogo minerelno-syrevogo universiteta «Gornyy», 298 p. (in Russian)
7. Guziy, A.G., Kapustin, A.G., Lushkin, A.M. and Fokin, A.V. (2019). Quantitative risk estimation methodology for flight safety in airline segment of commercial aviation. Civil Aviation High Technologies, vol. 22, no. 4, pp. 33-42. DOI: 10.26467/2079-0619-2019-22-4-33-42. (in Russian)
8. Tolstykh, S.A. and Sharov, V.D. (2018). Method of sms basic elements development for the aerodrome operator. Civil Aviation High Technologies, vol. 21, no. 4, pp. 29-38. DOI: 10.26467/2079-0619-2018-21-4-29-38. (in Russian)
Civil Aviation High Technologies
Vol. 23, No. 05, 2020
9. Sharov, V.D. and Vorobyov, V.V. (2017). Fuzzy risk assessment of aviation events. Civil Aviation High Technologies, vol. 20, no. 3, pp. 6-12.
10. Lushkin, A.M. (2015). Methodology of probabilistic evaluating of current accident rate based on the factorial analysis of aviation events. Nauchnyy Vestnik MGTU GA, no. 218 (8), pp. 24-28. (in Russian)
11. Sharov, V.D. (2014). Developing safety indicators on the basis of the ICAO recommendations. Nauchnyy Vestnik MGTU GA, no. 204, pp. 97-104. (in Russian)
12. Popov, Ju.V. (2015). Quantity indicators of safety of flights. Problemy bezopasnosti poletov, no. 2, pp. 10-19. (in Russian)
13. Orlov, A.I. (2018). Diversity of the models for regression analysis (generalizing article). Zavodskaya Laboratoriya. Diagnostika Materialov, vol. 84, no. 5, pp. 63-73. DOI: 10.26896/10286861-2018-84-5-63-73. (in Russian)
14. Borovikov, V.P. (2003). STATISTICA. Iskusstvo analiza dannykh na kompyutere [STATISTICS. The art of analyzing data on a computer]. 2nd ed. ispr. i dop. (+ CD). St. Petersburg: Piter, 688 p. (in Russian)
15. Orlov, A.I. (2018). Errors in the use of correlation and determination coefficients. Zavodskaya Laboratoriya. Diagnostika Materialov, vol. 84, no. 3, pp. 68-72. DOI: 10.26896/1028-68612018-84-3-68-72. (in Russian)
16. Bakhrushin, V.E. (2011). Methods for non-linear statistic bonds estimation. System technologies, vol. 2, no. 73, pp. 9-14. (in Russian)
17. Ershov, E.B. (2008). A choice of the regression maximizing an unbiased estimate of the coefficient of determination. Applied Econometrics, no. 4 (12), pp. 71-83. (in Russian)
18. Kaiser, H.F. (1960). The application of electronic computer to factor analysis. Educational and Psychological Measurement, vol. 20, issue 1, pp. 141-151. DOI: 10.1177/001316446002000116
19. Brown, T.A. (2006). Confirmatory factor analysis for applied research. Guilford Press, 475 p.
INFORMATION ABOUT THE AUTHOR
Sergei A. Tolstykh, Junior Researcher, Postgraduate Student of Moscow State Technical University of Civil Aviation, s.tolstyh@mstuca.aero
МЕТОДИКА ОПТИМИЗАЦИИ ПРИНЯТИЯ РЕШЕНИЙ ПРИ УПРАВЛЕНИИ БЕЗОПАСНОСТЬЮ ПОЛЕТОВ В ДЕЯТЕЛЬНОСТИ
ОПЕРАТОРОВ АЭРОДРОМОВ
С.А. Толстых1
1 Московский государственный технический университет гражданской авиации,
г. Москва, Россия
Работа выполнена при финансовой поддержке РФФИ в рамках научного проекта
№ 19-38-90215
В современных условиях ограниченности бюджета операторов аэродромов задача оптимизации принятия решений при управления безопасностью полетов становится чрезвычайно актуальной. Управленческие решения, являющиеся инструментом управления безопасностью полетов, должны быть не только эффективны с точки зрения ожидаемых улучшений в безопасности, но и экономически выгодны и целесообразны для данного предприятия. Под оптимизацией в данной статье следует понимать минимизацию по этим критериям. В статье представлен метод поддержки принятия управленческих решений в рамках стратегии управления безопасностью полетов в деятельности операторов аэродромов.
Vol. 23, No. 05, 2020
Civil Aviation High Technologies
В представленной методике важное место отводится показателям уровня безопасности полетов и их использования при принятии управленческих решений. Наряду с показателем безопасности полетов используется показатель финансового ущерба от зафиксированных событий, который рассчитывается в стоимостном выражении с учетом прямого и косвенного ущерба для оператора аэродрома. Используется регрессионное моделирование совместно с методикой принятия решений «человеко-машинных процедур». Регрессионный анализ выполняется с применением программного обеспечения STATISTICA и позволяет выявить зависимость показателей от степени влияния факторов опасности. Полученная модель на основе данных за прошлый год дает возможность выполнять прогноз значений показателей на следующий. Используя методику принятия решений «человеко-машинных процедур» выполняется оценка приоритетности внедрения управленческих решений на основе комплексного критерия. Методика обеспечивает выполнение требований российского и международного воздушного законодательства для операторов сертифицированных аэродромов. Область ее применения может быть расширена до СУБП всех поставщиков авиационных услуг при учете соответствующей специфики предоставляемых услуг и имеющихся факторов опасности. Исследование выполнено при финансовой поддержке РФФИ в рамках научного проекта № 19-38-90215.
Ключевые слова: риск безопасности полетов, показатель, оператор аэродрома, оптимизация, регрессионный анализ, прогноз, управленческие решения.
СПИСОК ЛИТЕРАТУРЫ
1. Хрусталев С.А., Орлов А.И., Шаров В.Д. Математические методы оценки эффективности управленческих решений // Заводская лаборатория. Диагностика материалов. 2013. Т. 79, № 11. С. 67-72.
2. Орлов А.И. Теория принятия решений: учеб. пособие. М.: Издательство «Экзамен», 2006. 576 с.
3. Ларичев О.И. Теория и методы принятия решений, а также Хроника событий в Волшебных странах: учебник для вузов. М.: Логос, 2008. 392 с.
4. Ларичев О.И. Свойства методов принятия решений в многокритериальных задачах индивидуального выбора // Автоматика и Телемеханика. 2002. № 2. С. 146-158.
5. Орлов А.И. Методы принятия управленческих решений: учебник. Москва: КНОРУС, 2018. 286 с.
6. Романов В.Н. Основы системного анализа: В.Н. Романов: учебно-методический комплекс. СПб.: Изд-во Национального минерально-сырьевого университета «Горный», 2012. 298 с.
7. Гузий А.Г. Методология количественного оценивания риска для безопасности полётов в самолётном сегменте коммерческой авиации / А.Г. Гузий, А.Г. Капустин, А.М. Лушкин, А.В. Фокин // Научный Вестник МГТУ ГА. 2019. Т. 22, № 4. C. 33-42. DOI: 10.26467/2079-06192019-22-4-33-42
8. Толстых С.А., Шаров В.Д. Метод разработки основных элементов системы управления безопасностью полетов оператора аэродрома // Научный Вестник МГТУ ГА. 2018. Т. 21, № 4. C. 29-38. DOI: 10.26467/2079-0619-2018-21-4-29-38
9. Шаров В.Д., Воробьев В.В. Нечеткая оценка риска авиационного события // Научный Вестник МГТУ ГА. 2017. Т. 20, № 3. C. 6-12.
10. Лушкин А.М. Методология вероятностного оценивания текущего уровня аварийности по результатам факторного анализа авиационных событий // Научный Вестник МГТУ ГА. 2015. № 218 (8). С. 24-28.
11. Шаров В.Д. Разработка показателя безопасности полётов на основе рекомендаций ИКАО // Научный Вестник МГТУ ГА. 2014. № 204. С. 97-104.
12. Попов Ю.В. Количественные показатели безопасности полетов // Проблемы безопасности полетов. 2015. № 2. С. 10-19.
13. Орлов А.И. Многообразие моделей регрессионного анализа (обобщающая статья) // Заводская лаборатория. Диагностика материалов. 2018. Т. 84, № 5. C. 63-73. DOI: 10.26896/1028-6861-2018-84-5-63-73
Civil Aviation High Technologies
Vol. 23, No. 05, 2020
14. Боровиков В.П. STATISTICA. Искусство анализа данных на компьютере. 2-е изд. испр. и доп. (+CD). СПб.: Питер, 2003. 688 с.
15. Орлов А.И. Ошибки при использовании коэффициентов корреляции и детерминации // Заводская лаборатория. Диагностика материалов. 2018. Т. 84, № 3. C. 68-72. DOI: 10.26896/1028-6861-2018-84-3-68-72
16. Бахрушин В.Е. Методы оценивания характеристик нелинейнык статистических связей // Системные технологии. 2011. Т. 2, № 73. С. 9-14.
17. Ершов Э.Б. Выбор регрессии максимизирующий несмещенную оценку коэффициента детерминации // Прикладная эконометрика. 2008. № 4 (12). С. 71-83.
18. Kaiser H.F. The application of electronic computer to factor analysis // Educational and Psychological Measurement. 1960. Vol. 20, iss. 1. Pp. 141-151. DOI: 10.1177/001316446002000116
19. Brown T.A. Confirmatory factor analysis for applied research. Guilford Press, 2006. 475 p.
Толстых Сергей Александрович, младший научный сотрудник, аспирант МГТУ ГА, s.tolstyh@mstuca.aero
СВЕДЕНИЯ ОБ АВТОРЕ
Поступила в редакцию Принята в печать
26.06.2020 24.09.2020
Received
Accepted for publication
26.06.2020 24.09.2020
