A human error mechanism for pilot based on fault tree analysis and Bayesian network

Abstract

At present, aircraft accidents caused by human factors occur frequently. As the flight crew is the last line of defense for aviation safety, how to identify and quantitatively evaluate human errors in flight and reduce the error rate has become an important issue for the civil aviation, and it is also the most effective way to control flight operation risks. Thus, the human error mechanism for pilot was investigated by the combine of the fault tree and the Bayesian network theories based on historical research data of typical unsafe events caused by flight human errors. Firstly, the fault tree was used to identify and qualitatively analyze the risk of systems, and then transformed into a Bayesian network model to obtain the relative probability of intermediate events and top events. Finally, the system hierarchy of unsafe events caused by flight human error was quantitatively evaluated. The results showed that there were 96 failure modes in the system, and the flight human error was caused by the coupling of multiple risk factors. The probability of non-technical skill loss is the highest, followed by that of the lack of technical skills and violations. The basic events in the organization, environment and equipment factors have a great impact on the flight human error, which is a weak link in the system. The results provide some theoretical basis for developing preventive measures of flight human error and improving the level of flight safety.

Keywords

Flight human error fault tree Bayesian network risk management flight safety

1 Introduction

Flight safety is the foundation of the civil aviation. With the continuous improvement of the cockpit automation level and the reliability of mechanical equipment, the human factor reliability of flight crews is becoming more and more critical for the entire civil aviation safety production system [1]. According to International Civil Aviation Organization (ICAO) statistics, flight human error in the civil aviation field has accounted for about 76% of the causal factors of accidents, among which the accidents caused by flight crew has reached more than 60%, which has become the main bottleneck in improving flight safety [2]. The civil aviation industry is a complex social technology system, and its safety has always been highly concerned by all social sectors [3]. The occurrence of a flight accident is a multi-stage dynamic system, and it is difficult to strictly distinguish the independence between various risk factors. Furthermore, the causal mechanism is complicated. Therefore, the research on the internal mechanism of flight human error is also the basic work to improve the level of flight safety.

However, the existing researches on the factors affecting flight human error mostly focus on the statistical analysis of pilots’ behavioral errors [4, 5], whilst pay little attention to organizational factors that lead to these behavioral errors like the unsound management standards and regulations, the insufficient supervision and review and the unreasonable flight operation plans as well as the inherent psychological factors including safety awareness, skills and experience, psychological status etc. In addition, the research on flight human error mechanism also pays less attention to the polymorphism of events and the uncertainty of the logical relationship of the causal factors of unsafe events, and lacks a dynamic description of such complex systems [6]. Overall, the existing researches on flight human error mostly focus on the analysis of behavioral errors at the subject level of risk events, lacking discussions on organizational-level management factors and employees’ psychological-level factors and paying insufficient attention to the interaction between risk factors. In particular, there is a lack of in-depth research on causal correlation and mechanism of risk factors. Therefore, this paper screens the risk factors of human error from human-machine-environment-organization and other multiple dimensions, analyzes the relationship and intensity of different influencing factors and uses both qualitative and quantitative analysis to diagnose system fault more clearly and objectively and identify the path of risk communication in the system, which provides managers with certain decision-making support for preventive measures for flight human error.

2 State of the art

In view of influencing factors of flight human error, the existing literature mostly analyzes the correlation between intrusion event data and influencing factors through mathematical statistical methods to determine the influence degree of different factors on the errors. Venera [7] established a non-linear mode by improving the Boolean algebra operation in the fault tree, and conducted a risk assessment for the dangerous environment, insufficient risk management, inability to control the aircraft and other factors in flight accidents. Saada et al. [8] used the dynamic Bayesian network model to diagnose individual faults of pilot under variable conditions during approach phase; Chen et al. [9] used Human Factors Intervention Matrix (HFIX) and hierarchical analysis. The method ranks the importance of various unsafe pilot behavior index factors and proposes intervention measures. Liu et al. [10] combined the SHELL model with the Reason model to design a model framework for the risk of human factors in civil aviation accidents and identify key factors that cause accidents. Since the Federal Aviation Administration (FAA) aircraft certification experts report that the human factors should be considered in the design of aircraft hardware, Michelle et al. [11] put forward the design concept of the human factors of aircraft control components through big data analysis, which aims to reduce the probability of flight human error.

There are many research results on the risk analysis of human factors. Generally, the steps of index selection, weight determination and comprehensive evaluation are used. The research methods mostly use BP neural network [12], system theory model [13], analytic hierarchy process [14], entropy method [15], event analysis model [16], accident chain model [17], etc. For example, Wang et al. [12] established a civil aviation flight safety risk assessment index system based on the fish-bone diagram of civil aviation accidents by using the analytic hierarchy process. They determined the weights among various factors, and established civil aviation flight safety risk assessment model using BP neural networks. The assessment model uses a four-level evaluation system to evaluate and analyze the output results. Zhao et al. [15] used set pair analysis and entropy method to dynamically evaluate the flight operation risk of airlines. Rosa María Arnaldo Valdés et al. [18] used Bayesian inference and hierarchical structure to build statistical estimation and prediction models with different complexity and goals, so as to analyze safety-related data and predict future risks. Jiang et al. [19] referred to the risk index evaluation method, using the judging parameters and frequency of occurrence of over-limit events in the QAR data to represent the severity of the risk consequences and the probability of risk occurrence, and using the product of the two as a measure of the pilot’s safety capability. Based on the QAR raw data of 36 captains of an airline for 3 years, with the model applied, the safety capability score of each pilot was obtained, and the safety capability level of each pilot was evaluated.

Existing research results provide a certain reference for improving the safety of civil aviation. However, the current researches on flight human error pay more attention to pilot operation-level errors and violations. Although such risks are the direct causes of intrusion events, less attention has been paid to management-level factors that lead to errors and violations, as well as employees’ psychological awareness-level factors. Focusing on mathematical statistics and comprehensive evaluation analysis, the current researches on flight human error lack the research on mechanism and sufficient analysis of the correlation between the factors, which fails to reveal the deep mechanism of flight human error.

Therefore, in view of the shortcomings of the existing researches and based on previous studies, this paper takes the respective applicability of fault trees and Bayesian network models into consideration and uses fault trees to build logical relationship between risk factors that lead to unsafe events caused by flight human error and analyze them qualitatively. Then the historical data is used to determine the prior probability of each basic event, and the probability of the top events of the basic event is determined by the Bayesian network. Finally, the backward derivation ability of the Bayesian network is used to calculate the posterior probability of each basic event. By using a combination of qualitative analysis and quantitative analysis, the system fault diagnosis is more clearly and objectively provided, so as to provide a basis for decision-making for civil aviation safety managers.

The remainder of this paper is organized as follows: The third section describes the screening method of the influencing factors of flight human error, the fault tree analysis method, the construction of the Bayesian network, and the mapping of the fault tree to the Bayesian network. The fourth section is the analysis of results. The last part summarizes this article and gives relevant conclusions.

3 Methodology

3.1 Fault tree analysis (FTA)

The fault tree analysis uses a tree diagram to analyse the logical relation among the causal factors of top events, intermediate and basic events. It is widely used in the calculation and analysis of safety and reliability of complex systems such as nuclear power, railways and water conservancy. This method can not only identify and analyze the cause of the accident layer by layer, but also calculate the minimum cut set, minimum path set, structural importance of fault tree to show the reliability of the system and the impact of basic events, so as to find the key factors, formulate effective intervention measures, and provide reference for system security. The basic symbols of the fault tree are shown in Table 1.

Table 1
Examples of basic symbols of fault tree

Event Symbol Logic gate Symbol

Top event or intermediate event and gate

Basic event or gate

Event	Symbol	Logic gate	Symbol
Top event or intermediate event		and gate
Basic event		or gate

3.2 Bayesian network (BN)

Bayesian network is a technology based on probabilistic graph model. Nodes and directed edges represent random variables and their conditional dependencies, so that potential relationships between data are discovered and the mutual relationship between prior and posterior probability between events are realized. Among them, the nodes of the Bayesian network represent variables, and the directed edges represent conditional dependencies between the variables [20], and each node has a conditional probability distribution table. Let P (X₁, X₂, ⋯ X_n) be the joint probability of the Bayesian network model, then $P (X_{1}, X_{2}, \dots X_{n}) = \prod_{i = 1}^{n} P (X_{i} | parent (X_{i}))$ (1)

Where X_i represents the i node and parent (X_i) represents the i parent node. Probability of top event T£º $\begin{matrix} \sum_{parent (T)} P (T = 1 | parent (T)) P (parent (T)) \\ = \sum_{λ (T)} P (parent (T), T = 1) \end{matrix}$ (2)

Where represents the set of child nodes.

3.3 Applicability analysis of the two theories to flight human error

The fault tree theory is mostly used for the safety and reliability analysis of static systems, in which it is assumed that events have only two states: working and failing, so it is difficult to analyze quantitatively polymorphic events effectively [21]. However, the occurrence of a flight accident is a multi-stage dynamic system, and it is difficult to strictly distinguish the independence of each risk factor. And the causal mechanism is complicated. Therefore, quantitative analysis using a fault tree will inevitably cause certain errors. The Bayesian network can take into account the polymorphism of the event and the uncertainty of the logical relationship between the causal factors of the unsafe event, avoiding the two-state assumption of the event “true or false” in the fault tree theory and has a strong ability to describe the accident status of complex system. Therefore, in order to enhance the accuracy of the flight human error analysis, with the advantages of the two methods combined, the Bayesian network is introduced for quantitative analysis based on the qualitative analysis of the fault tree.

3.4 Mapping of fault trees to Bayesian networks

The top event, intermediate event, and basic event in the constructed fault tree are converted into corresponding nodes of the Bayesian network according to certain rules. The probability of the corresponding event is converted into the initial probability of the corresponding node, and the logical relationship corresponds to the conditions between the nodes Probability. According to references [22, 23], the mapping rules of fault trees and Bayesian networks are obtained, as shown in Fig. 1.

Fig. 1

Mapping from fault tree to Bayesian network.

4 Results analysis and discussion

This paper uses the official accident / incident investigation and analysis reports of civil aviation around the world, China’s civil aviation safety information system, and worldwide civil aviation unsafe incident investigation and tracking reports as sources of information. Based on the analysis of 304 typical flight human error accidents / incidents occurring from 1980 to 2018 at home and abroad, combining a variety of flight human error classification methods such as the SHELL model, HFACS, and information processing theory, this paper analyzes and combs the previous extraction methods for the influence factors of flight human error. On this basis, in December 2018, we conducted expert interviews via email, telephone, video conference, etc., and further optimized the influencing factors. The interviewees were all scholars of airport safety management or middle and senior management personnel of the enterprise. Five of them were from the Civil Aviation Administration Accident Investigation Center, five were from the Zhengzhou Xinzheng Airport Management Department, and five were from the China Southern Airlines Henan Branch. According to the collation and analysis of the interview results, 20 influencing factors (Fig. 2 and Table 2) were obtained after simplification. The index was analyzed from the four aspects of “human-machine-environment-organization” as the main risk factors of flight human error.

Fig. 2

Fault tree model of flight human error.

Table 2

Events numbers of flight human error

Number	Event	Number	Event
T0	Flight human error event	X5	Terrible weather
M1	Trigger factor	X6	Complex airspace
M2	Unsafe behavior	X7	Poor human-machine interface design
M3	Organization factor	X8	Equipment failure
M4	Environment factor	X9	Invulnerability
M5	Equipment factor	X10	Macho
M6	Error	X11	Poor level of operation skills
M7	Violation	X12	Lack of experience
M8	Lack of technical skills	X13	Lack of crew resource management ability
M9	Lack of non-technical skills	X14	Loss of situation awareness
M10	Physical / psychological factors	X15	Poor stress management ability
M11	Disciplinary issues	X16	Decision making error
X1	Human resource management issues	X17	Poor physiological state
X2	Insufficient standards and regulations	X18	Poor metal state
X3	Insufficient supervisory and review	X19	Poor safety awareness
X4	Unreasonable operation plan	X20	Poor regulation awareness

4.1 Results and analysis of fault tree model

Taking the flight human error unsafe event as a top event, direct reason is the coupling of triggering factors such as organization factors, environment factors, and equipment factors with human unsafe behavior. These factors are analyzed in depth from top to bottom to find the basic reason which is taken as the basic event, and then the fault tree of the unsafe event due to flight human error is established, as shown in Fig. 2.

The event numbers in the fault tree are shown in the Table 2.

The calculation and analysis of the fault tree are mainly based on three aspects: minimum cut set, minimum path set, and structural importance. The minimum cut set represents the combination of the probability of the top event and the risk factors that cause the top event [24]. The larger the number of minimum cut set is, the greater the risk of the system is. Contrary to the concept of the minimum cut set, the minimum path set refers to the set of minimum basic events that cannot cause a fault tree top event to occur. The more minimum path sets, the more measures and schemes that represent the system’s risk control there are, and the more secure the system is. The structural importance only analyze influence degree of the basic event on the occurrence of the top event from the structural system aspect, ignoring the probability of the basic event or assuming that the probability is equal.

The above fault tree is represented by Boolean algebra: $\begin{matrix} T 0 = M 1 M 2 = (M 3 + M 4 + M 5) (M 6 + M 7) \\ = (X 1 + X 2 + X 3 + \dots + X 7 + X 8) (X 9 + X 10 \\ + M 8 + M 9 + M 10 + M 11) \\ = (X 1 + X 2 + X 3 + \dots + X 7 + X 8) (X 9 + X 10 \\ + X 11 + \dots X 19 + X 20) \\ = X 1 X 9 + X 1 X 10 + X 1 X 11 + \dots + X 8 X 19 \\ + X 8 X 20 \end{matrix}$

According to the above formula, the minimum cut set of the fault tree is {X1, X10}, {X1, X11}, ... ... {X8, X19}, {X8, X20}, which is 96 in total, representing a total of 96 failure modes of the system and indicating that the probability of flight human error unsafe events is great, and that there is a certain degree of prevention difficulty. For example, the minimum cut set indicates that if the organization’s supervision and review are insufficient, when the flight crew’s safety awareness is poor, it is easy to cause flight human error unsafe events.

According to the dual success number algorithm of the fault tree, the minimum path sets of the fault tree are {X1X2X3X4X5X6X7X8}, {X9X10X11X12X13X14X15X16X17X18X19X20}, indicating that the way to prevent this event is to develop corresponding measures from the aspects of triggers factors and unsafe behavior of pilot.

The structural importance of the fault tree is calculated as follows: $\begin{matrix} I (X 1) = I (X 2) = I (X 3) = I (X 4) = I (X 5) \\ = I (X 6) = I (X 7) = I (X 8) > I (X 9) \\ = I (X 10) = I (X 11) = I (X 12) = I (X 13) \\ = I (X 14) = I (X 15) = I (X 16) = I (X 17) \\ = I (X 18) = I (X 19) = I (X 20) \end{matrix}$

It shows that human resource management issues (X1), inadequate standard regulations (X2), insufficient supervision (X3), unreasonable operation plans (X4) and other basic events have a great impact on the flight human error.

4.2 Analysis results based on Bayesian network theory

In the calculation of the project, the frequency of basic events is often used to replace its probability value [25, 26]. Due to the complexity of the system of the flight human error, the probability of basic events is difficult to accurately obtain. In this paper, the statistical analysis of the historical events caused by flight human error is used to obtain the frequency of each basic event, which is approximately prior probability, as shown in Table 3.

Table 3
Prior probability of basic events

Basic events Probability Basic events Probability Basic events Probability

X1 0.00287 X8 0.00299 X15 0.00687

X2 0.00305 X9 0.00264 X16 0.00741

X3 0.00329 X10 0.00209 X17 0.00385

X4 0.00332 X11 0.00905 X18 0.00341

X5 0.00248 X12 0.00751 X19 0.00470

X6 0.00235 X13 0.00979 X20 0.00718

X7 0.00304 X14 0.00959

Basic events	Probability	Basic events	Probability	Basic events	Probability
X1	0.00287	X8	0.00299	X15	0.00687
X2	0.00305	X9	0.00264	X16	0.00741
X3	0.00329	X10	0.00209	X17	0.00385
X4	0.00332	X11	0.00905	X18	0.00341
X5	0.00248	X12	0.00751	X19	0.00470
X6	0.00235	X13	0.00979	X20	0.00718
X7	0.00304	X14	0.00959

According to the mapping rules in Fig. 1, a Bayesian network model of the unsafe event of the flight human error is established, as shown in Fig. 3.

Fig. 3

Bayesian network model of flight human error.

With the probability of the basic event of flight human error unsafe event in Table 3, using the forward reasoning ability of the Bayesian network model, the relative probability of the intermediate event and the top event is calculated by the GENIE2.0 software, as shown in Table 4.

Table 4

Prior probability of intermediate events and top events

Basic events	Probability	Basic events	Probability	Basic events	Probability
M1	0.00145	M5	0.00151	M9	0.00835
M2	0.00137	M6	0.00118	M10	0.00181
M3	0.00312	M7	0.00429	M11	0.00296
M4	0.00121	M8	0.00413	T0	0.00071

Using the backward reasoning abilities of the Bayesian network model, the most probable cause of the result can be discovered by calculating the posterior probability of basic events. Therefore, by setting the child node T0 in the GENIE2.0 software as an evidence node, which means that the probability of unsafe events due to flight human error is 100%, the mutual inference of the prior and posterior probabilities between the nodes is performed for system fault diagnosis. Among them, the posterior probability of each basic event is shown in Table 5.

Table 5

Posterior probability of basic events

Basic events	Probability	Basic events	Probability	basic events	Probability
X1	0.06561	X8	0.05657	X15	0.03866
X2	0.07073	X9	0.03760	X16	0.04708
X3	0.07521	X10	0.03398	X17	0.02276
X4	0.06889	X11	0.05789	X18	0.02364
X5	0.04591	X12	0.04806	X19	0.03016
X6	0.03893	X13	0.06916	X20	0.05405
X7	0.05978	X14	0.05533

It can be seen that the key factors affecting the flight human error unsafe incidents are human resource management (X1), inadequate standards and regulations (X2), insufficient supervision and review (X3), unreasonable operation plans (X4), and poor human-machine interface design (X7), failure of facilities and equipment (X8), low level of operation technology (X11), lack of crew resource management capability (X13), loss of situation awareness (X14), and weak awareness of regulations (X20). It is shown that these basic events are the weak links of the system. Upon failure, it is very easy to cause unsafe incidents. It is also an important basis for the system to establish a safety barrier.

5 Conclusion

Aiming at the inherent mechanism of flight human error, the fault tree and Bayesian network model of the flight human error unsafe events was constructed and analyzed based on the system-safety theory and the applicability of the FTA and BN models. The failure mode, the importance of various basic events, the direct reasons and the improvement methods were determined. The work made a thorough analysis of the system hierarchy of unsafe events caused by flight human error. The conclusions are as follows:

(1) Based on the FTA theory, a systematic analysis of the flight human error can better clarify the logical relationship among the influencing factors and identify the combination of various failure modes in the civil aviation operation process caused by flight human error. It shows that flight human error is caused by the coupling of multiple risk factors and difficult to prevent. Through the calculation of the minimum path set of the fault tree, it can be clearly pointed out that the safety managers need to strengthen management in terms of trigger factors and unsafe behavior of pilot in order to reduce unsafe events caused by flight human error.

(2) Based on the forward and reverse directions of the BN theory, the respective probability of various influencing factors in the flight human error can be relatively accurately obtained. Among intermediate events, the lack of non-technical skills has the highest probability, followed by lack of technical skills and violations. It indicates that the lack of non-technical skills, the lack of technical skills and the neglect of regulations of the flight crew are the direct reasons of the incident. The unsafe behaviors of the flight crew are mainly caused by the poor level of operation skills, the lack of knowledge and experience, the lack of crew resource management capability, the loss of situation awareness and weak regulation awareness. These basic events are the key factors in the human error of the flight crew. Therefore, flight crew need to be trained in these aspects to ensure aviation safety.

(3) Based on the analysis of FTA-BN theory for unsafe events caused by flight human error, the result shows that the basic events in organization, environment and equipment factors have a great impact on causing flight human error, which are the weak links of the system. It is necessary to formulate countermeasures from a systematic perspective. It shows the inherent consistency of the fault tree and Bayesian network analysis methods in the fault diagnosis of key risk factors that lead to flight human error. Firstly, it shows that organization factors are regarded as the biggest potential threat to system safety, which is the key to the “front control” of system risk management. Secondly, the harsh flight environment will directly lead to the increase in the complexity of the flight crew information processing as well as the difficulties of control and operation management, which will severely affect the flight crew’s situation awareness and decision-making capability. Finally, with the improvement of the reliability of aircraft equipment, flight human error has become an important factor in flight safety nowadays.

(4) Applying the FTA and BN theories on the analysis of the internal mechanism of flight human error can deeply dig into the system hierarchy and clarify the controlling factors and improvement methods. It not only provides a simple and reliable theoretical analysis model for improving the level of flight safety, but also provides certain application value for safety managers’ scientific decisions.

The work digs out the key risk factors that lead to flight human error based on the fault tree analysis and Bayesian network. The difficulties as the independence and complex cause mechanism among various the risk factors of the flight human error are processed scientifically and effectively. However, the path analysis of unsafe events caused by flight human error and how to enhance the situational awareness of flight crews in flight for effectively intercepting the development path of unsafe events are need to further research.

Footnotes

Acknowledgments

This paper was supported by Humanity and Social Science Youth foundation of Ministry of Education of China (17YJC630124), science and technology planning project of Henan Province (192102310253, 182102310725); Key technology projects for the prevention and control of serious and especially serious accidents in safety production (henan-0012-2017AQ).

References

Kharoufah

, Murray

and Baxter

, A review of human factors causations in commercial air transport accidents and incidents: From to 2000–2016, Progress in Aerospace Sciences99 (2018), 1–13.

Kramer

L.J.

, Etherington

T.J.

and Bailey

R.E.

, Quantifying Pilot Contribution to Flight Safety during Hydraulic Systems Failure, Advances in Human Aspects of Transportation597 (2017), 15–26.

Tamasi

and Demichela

, Risk assessment techniques for civil aviation security, Reliability Engineering & System Safety96(8) (2011), 892–899.

Zhou

and Hu

, Research on Inducement to Accident/Incident of Civil Aviation in Southwest of China based on Grey Incidence Analysis, Procedia Engineering45 (2012), 942–949.

Lalis

, Socha

and Kremen

, Generating synthetic aviation safety data to resample or establish new datasets, Safety Science106 (2018), 154–161.

Hooper O’Hare

B.J.

and David

P.A.

, Exploring Human Error in Military Aviation Flight Safety Events Using Post-Incident Classification Systems, Aviation Space and Environmental Medicine84(8) (2013), 803–813.

Venera

B.C.

, Boolean algebra Application in Analysis of Flight Accidents, Incas Bulletin7(4) (2015), 41–49.

Saada

, Meng

and Huang

, A novel approach for pilot error detection using Dynamic Bayesian Networks, Cognitive Neurodynamics8(3) (2014), 227–238.

Chen

J.C.

and Yu

V.F.

, Relationship between human error intervention strategies and unsafe acts: The role of strategy implementability, Journal of Air Transport Management69 (2018), 112–122.

10.

Liu

and Wang

L.L.

, Civil Aviation Accident Human Factors Analysis Model, Advanced Materials Research955–959 (2014), 1825–1828.

11.

Michelle

, Swider

and Young

J.J.

, et al., Human factors considerations in the design and evaluation of flight deck displays and controls: version 2.0, Washington DC: Federal Aviation Administration, 2016.

12.

Wang

Y.Y.

, Zhang

Z.H.

and Qian

, Research on Civil Aviation Flight Security Risk Assessment Based on BP Neural Network, Journal of Nanchang Hangkong University: Social Sciences33(2) (2019), 89–95.

13.

Wang

, Sun

and Li

, Analysis of military aircraft flight training safety: based on STAMP model, China Safety Science Journal28(09) (2018), 72–77.

14.

Wang

, Du

and Ding

, Flight quality evaluation of aircraft take-off stage based on AHP-TOPSIS method, Flight Dynamics37(01) (2019), 83–88.

15.

Zhao

and Wan

, Evaluation of Airline Risk Based on Five Element Connection Number-Entropy Weight Method, Science Technology and Engineering18(05) (2018), 347–352.

16.

Sun

and Zhao

, Study on ECAR Model for Human Error Analysis of Aviation Accident/Incident, China Safety Science Journal (02) (2012), 19–24.

17.

, Liu

and Jin

, Study on a Structured Pilot Error Model Based on Information Processing and Accident Chain Principle, China Safety Science Journal21(6) (2011), 25–31.

18.

Valdés

, María

, Comendador

, et al., Prediction of aircraft safety incidents using Bayesian inference and hierarchical structures, Safety Science104 (2018), 216–230.

19.

Jiang

, Shi

and Yang

, Establishment of a QAR Data-Based Model for Assessing Safety Capabilities in Airline Pilots, Space Medicine & Medical Engineering32(3) (2019), 208–212.

20.

Esmaeil

, Ali

and Mohammadfam

, Dynamic safety assessment of natural gas stations using Bayesian network, Journal of Hazardous Materials321(5) (2017), 830–840.

21.

Kabir

, An overview of fault tree analysis and its application in model based dependability analysis, Expert Systems with Applications77 (2017), 114–135.

22.

Kang

, Zhong

and Haijun

, Research on Probabilistic Safety Analysis Approach of Flight Control System Based on Bayesian Network, Procedia Engineering99 (2015), 180–184.

23.

Zhao

and Zhao

, Risk analysis of the runway excursion and overrun event based on the fault tree and Bayesian network, Journal of Safety and Environment14(3) (2014), 141–145.

24.

Ruijters

and Stoelinga

, Fault tree analysis: A survey of the state-of-the-art in modeling, analysis and tools, Computer Science Review15-16(03) (2015), 29–62.

25.

Garkushenko

V.I.

and Lazareva

P.A.

, On the Problem of Improving the Flight Safety in Manual Control of the Aircraft, Russian Aeronautics60(4) (2017), 528–533.

26.

Nabdi

, Zaghar

and Herrou

, Bayesian Modeling of a Wind Turbine System in the Preliminary Design Phase, Journal of Engineering Science & Technology Review10(4) (2017), 185–190.

A human error mechanism for pilot based on fault tree analysis and Bayesian network

Abstract

Keywords

1 Introduction

2 State of the art

3 Methodology

3.1 Fault tree analysis (FTA)

Table 1 Examples of basic symbols of fault tree Event Symbol Logic gate Symbol Top event or intermediate event and gate Basic event or gate

3.4 Mapping of fault trees to Bayesian networks

4.2 Analysis results based on Bayesian network theory

Footnotes

Acknowledgments

References

Table 1
Examples of basic symbols of fault tree

Event Symbol Logic gate Symbol

Top event or intermediate event and gate

Basic event or gate