V Kharchenko, V Chepizhenko, Y Zaitsev - Impact of poly-linguistic load on air traffic control and monitoring quality - страница 1

Страницы:
1  2 

ISSN 1813-1166. Proceedings of the NAU. 2012. №2

UDC 629.735.05:519.226

1Volodymyr Kharchenko, Prof. 2Valery Chepizhenko, Sn. Researcher 3Yuriy Zaitsev, Head of the Department

IMPACT OF POLY-LINGUISTIC LOAD ON AIR TRAFFIC CONTROL AND MONITORING QUALITY

National Aviation University 1E-mail: kharch@nau.edu.ua 2E-mail: chiv@nau.edu.ua 3E-mail: vtnik@nau.edu.ua

Abstract. We have defined the structure and basic characteristics of the poly-linguistic audio-acoustic channel within the framework of controller - pilot communication, and set limits of poly-linguistic load impact on air traffic control.

Keywords: controller, dynamic flight status, linguistic interaction, operator model, poly-linguistic audio-acoustic channel, time delay.

Introduction

One of the main features of the modern air traffic system is consolidation of all its participants to ensure maximum security. This feature is essential for creation, existence and operation of the air navigation environment and is basic for the synthesis of its component behaviors. The quickest solution to the air traffic safety problem follows from the regulatory and other documents of the world's leading organizations [1-3], notably: International Civil Aviation Organization (ICAO), European Organization for the Safety of Air Navigation

(EUROCONTROL), National Agency of Space and Aeronautics (NASA).

The usual air safety problem is gaining today a fresh impetus due to an increase in air traffic, expansion of aircraft operation and the range of issues aviation is facing in the given conditions.

The projections below [4-6] show a rise in annual air traffic 5 to 7 percent. This refers to increasing topological complexity of conflicts and consequently, stricter requirements for air control quality.

The analysis of abnormal situations, accidents and incidents caused by linguistic misunderstanding and semantic distortions in verbal interaction between air traffic operators shows a profound impact of the linguistic load on communication (information channel) between crew members and controllers that leads to deterioration in air traffic performance and affects flight safety.

© Kharchenko Volodymyr, Chepizhenko Valery, Zaitsev y

Review of latest publications

The study into the controller-pilot audio-acoustic channel generally highlights its mono structure [7]. Therefore, the results of the studies do not suggest any conclusion about the degree of a linguistic component impact on the quality of Air Traffic Control (ATC) and monitoring.

However, the recent studies [8-10] indicate that the interaction between pilot-controller data components has different levels of their dominance that makes the audio-acoustic channel somewhat different from the mono one. In Ukraine we use a poly-linguistic audio-acoustic channel in controller-pilot communication.

The poly-linguistic structure is based on several factors: two commonly used languages in Ukraine (Russian and Ukrainian), an adjacent air traffic control complex with Russian as working language (Russian Federation and most CIS countries), and the International ATC System (English for professional purposes) [8; 11].

The poly-linguistic structure of the controller-pilot audio-acoustic channel will hereinafter be cited as the audio-acoustic channel with several language components sharing common logic and purpose - to ensure effective air traffic control.

Identification of Unsettled Issues

The analysis of recent studies shows a lack of an effective scientific and methodological instrument:

- to determine a structure and characteristics of the controller - pilot poly-linguistic audio-acoustic channel;

iy, 2012

- to establish an influence pattern for the poly-linguistic load on air traffic control characteristics;

- to develop efficient methods for increasing capacity and reliability of controller-pilot communication and avoid the effect

The purpose of the article is to specify the structure and characteristics of the controller-pilot poly-linguistic audio-acoustic channel, and to outline proposals against negative effect of the poly-linguistic load on the air traffic control.

Description of the Basic Material (Idea)

Establishment of structure and characteristics of poly-linguistic audio-acoustic channel.

The poly-linguistic audio-acoustic channel was established when studying controller competency in abnormal situations [10].

To determine and study characteristics of the poly-linguistic audio-acoustic channel we adopted the following technology: the examinee was being displayed graphic stimuli to choose two of the three; if the answer is false, each subsequent exposure, beginning with the initial one, is increasing or reducing if the answer is correct. The time of stimuli differentiation is being recorded to set the level of functional mobility level in nervous processes [8-10].

The examinee is given the task consisting of three consecutive programs:

program A is Ukrainian language; program B is Russian language; program C is English. The speed of data processing is measured by three types of verbal information. Then they compare the same parameters received for each program and determine how suitable the examinee is for an air traffic controller on international airlines and in abnormal situations.

The comparative histogram shows that percentage rate of the examinees capable of fast data processing and rendering (grades 8 to 12) in increased workload exceeds that of the control group; air controllers with slow data processing and rendering (grades 1 to 3) were not identified [12-14].

Thus, to operate in abnormal conditions the examinee must achieve 8 to 12 grades in data processing and rendering. Applicants (persons) with 4 to 7 grades are eligible conditionally, while those attained lower than 4 grades are turned away (fig. 1).

Based on the statistical methods of data processing and nervous processes all the examinees were classified according to the 12-grade scale taking into account the average arithmetic values

and the average square deviations О of these indicators

n -1

where E is the sign of product deviation power, option 'Xi of their average x to weight Pi of the deviations from 1st to k class; n is volume.

If the examinee receives high grades (8-12) in A, B programs, he is good for working as operator of fast data processing and rendering operator in abnormal conditions (air traffic controller in Ukraine and the CIS countries).

Getting high grades in the C program the examinee fits for data processing in English (air traffic controllers on international flights), while low grades (1 to 3) in the C program do not allow operating international flights [8; 10; 13; 15].

20

15

10

4—

О

CD

E Z

о 4

Air controllers Control group

18,7

12,5 7,1

1

2

3

4

5

8

9

10

11

12

6 7 Grades

Fig. 1. Comparative histogram for differentiation between examinees working in abnormal conditions and the control group under the 12-grade scale of data processing and rendering speed

2

i=1

5

Low grades in the A and B programs mean the examinee slowly processes data when experiencing emotional stress, and, at large, cannot be employed as air traffic controller even for local destinations.

Every controller in Ukraine actually uses the poly-linguistic channel in English for international flights or Russian for domestic flights. Ukrainian is used to validate translation if the Ukrainian language is the mother tongue for a controller.

The structure of the poly-linguistic channel:

a) controller - native language Ukrainian or Russian depending on a birth place and a workplace;

b) commonly spoken language, Ukrainian or Russian (with the above mentioned reasons);

c) professional English language.

Time frame for reconstruction of a Dynamic

Flight Status (DFS) in ATC.

1. Reconstruction of the ATC dynamic flight status in Russian:

a) controller's native language is Russian;

b) commonly spoken language is Ukrainian;

c) professional English language:

^DFS

DFS1

E (r

s=1

N

E k

DFS     DFS DFS

s

DFS

s=1

1,s DFS1

2,s DFS1

(1)

where rD1,FsS is time required for reconstruction of the dynamic flight status in Russian;

2,s

is time required for partial validity check to reconstruct the dynamic flight status (influence of Ukrainian and articulation differences);

rD3F,sS is time for acute translation of the dynamic flight status in English involved s aircraft;

s = 1, N is number of aircraft (if ATC needs duplicating by international standards at crew's request or other reasons);

k

DFS

is coefficient of time spent depending on

pilot's language competence;

r1,s

DFS1

is time required for report perception and

understanding (request, confirmation) by s crew;

t2,s  is time for data transmission to s crew;

DFS1

a) controller's native language is Ukrainian;

b) Russian is a commonly spoken language;

c) professional English language:

tDFS2 = E (

rl,s     + r 2,s    + r 3,s

DFS2     DFS2 DFS2

where

'dfs2

is time required for reconstruction of

the dynamic flight status in Russian (ATC operation under the regulations of the Russian Federation and certain CIS countries);

is time required for parallel reconstruction

r 2,s

'dfs2

of the dynamic flight status in Ukrainian (validated translation is not communicated);

rD3F,sS2   is time required for making accurate

translation of the dynamic flight status in English, including s aircraft;

s = 1, N is number of aircraft (aircraft transition to international ATC zone).

Coefficient of time spent depending on pilot's language competence kDs FS see formula (1);

a) controller's native language is Ukrainian;

b) Russian is a commonly spoken language;

c) professional English language (aviation English) [8-11; 13].

The controller-pilot audio-acoustic channel is of a bilingual structure. If the controller is capable of fast data processing and rendering in English, he may conduct air traffic control according to international rules:

N

E 0

s=1

,1,s

+ r,

2,s

DFS3 DFS2

,)■

where rD1,FsS3 is time required for reconstruction of

the dynamic flight status in professional English (in ATC under international rules).

The coefficient for time spent is similar to the above mentioned.

2. Compulsory reconstruction of the dynamic flight status in English:

a) controller's first language is Ukrainian or Russian (depending on a birth place and a workplace);

b) commonly spoken language, Ukrainian or Russian (with the above mentioned reasons);

c) professional English language:

^DFS4

N

E (

s=1

DFS4      rDFS4 'DFS4

where r1,s   is time required for reconstruction of

DFS4

the dynamic flight status in Russian (not communicated);

r 2,s   is time required for reconstruction of the

DFS4

dynamic flight status in Ukrainian (not communicated);

s

s=1

rD3F,sS4  is time required for making accurate

translation of the dynamic flight status in English, including s aircraft.

The coefficient for time spent is similar to the above mentioned [10; 13-15].

In modeling, professional English is considered a constant due to its most universal and adequate interpretation of dynamic flight status. The variables in this complex are language competences of a pilot and a controller, as well as their psychological and emotional characteristics. The Russian language is defined as a specialized language with limited regional restrictions. Russian is used in ATC in the territory of the Russian Federation and the CIS countries, although the basic rules of air traffic control in the Russian Federation differ from international ones.

Ukrainian is defined as variable linguistic value, practically unsuitable for ATC.

Lengthy delays in simulating the dynamic flight status are vital for air traffic control. The distinctive feature of the controller - pilot audio-acoustic channel is an incomplete reply and a quite conventional and situational language flow. This type of speech requires highly accurate display of situational time intervals.

Analysis of the poly-linguistic load

on air traffic control efficiency (quality)

To study the effect of operators' delay on air traffic services we'll consider a simple control loop

(fig- 2).

Under the primitive control loop we mean the smallest closed loop of air traffic control system, including a controller, communications and surveillance tools, a pilot and a plane. The division of the primitive loop into smaller elements is not advisable as it breaches the conditions for solutions of functional tasks assigned to the air traffic control system [8; 15].

The phase coordinate control system for each aircraft in the control traffic area consists of two primitive control loops: inner and outer (fig. 2).

The inner control loop is formed by transfer functions which give a formal description of:

- an i-aircraft dynamic in a particular flight mode WAPi;

- psycho-physiological characteristics of pilot performance WPi within aircraft control loop;

- dynamics of on-board information system WISi

in receiving, processing and delivering data about aircraft phase coordinates to the pilot [8; 14; 16].

In abnormal situations caused by equipment fail­ures or pilot's distrust in the data from the on-board information system the inner loop opens (fig. 1 depicts the process as a switch). In this case the air traffic con­trol system is managed by the outer control loop in­cluding a controller, a pilot and a plane. Fig. 2 features controller's psycho-physiological performance as transfer function WFD .

To carry out an analytical study into the influence of the poly-linguistic load on qualitative characteristics of the controller - pilot performance in the primitive control loop it is convenient to use an operator's mathematical model as part of the control system.

Today there are plenty of models describing human behavior in the control loop (see table).

Mathematical model of man - operator control loop

Fig. 2. Structural chart of simple ATC loops sp- setpoint: W    is surveillance transfer function

sur

Model

Equation

Model parameter value

Refe­rence

1

W (s) = K

K = 5...100

[1]

2

Ke~xs W (s) = 7;s +1

K = 5...9 t = 0,1...0,3 T1 = 0,1...1

[1; 2]

3

W    = Ke-s(T2s + 1) V '   (Ts + VfJs +1)

K = 5...100 t = 0,1...0,3 T1 = 0,1...1

T2 > 0,1 T3 £ 25

[1-3]

Note: s=a± jb - complex number of Laplace transform

The parameters of the above models reflect man - operator physiological features:

K is the operator gain system is set depending on an input frequency range. The higher frequency component, the lower operator gain is. So, the man -operator enters derivative signals keeping control margin at 40 - 800;

T1 is time of effector's (motor) delayed action;

T2 is indicates person's ability to anticipate

changes in input signal (differentiate);

T3 is indicates man - operator's integrating

capacity.

Model 1 is not used in practice because of severe air traffic control restrictions. The analysis of models 2 and 3 shows that transfer functions contain pure delay links characterized by time delay t.

This parameter specifies delay in the man -operator response to input data [18].

As shown above, the delay strongly depends on the poly-linguistic load. Consequently, the delay is a general characteristic integrally reflecting the impact of the poly-linguistic load on the air traffic control loops.

In general, the delay time in the controller - pilot closed loop is a complicated function depending on dynamics of set and actual parameters of aircraft movement and on external disturbances.

The external disturbances occurred due to variations in the controller-pilot audio-acoustic data channel under the poly-linguistic factors described above are shown in fig. 1 by f and signals

respectively [17; 19; 20].

As known from the theory of automatic control, delay elements in the control loop do not influence accuracy of input signal processing, but significantly affect the phase margin I the system.

The assessment of the impact of transient delays on qualitative characteristics of the primitive control loop in view of the poly-linguistic load on the operator is given below. The primitive open loop transfer function (fig. 1) is as follows:

W(s) =Wfd(s)•      Wp(s) •           (s)     (s) =

= Wfd (s) Wp (s)     (s)     (s) e-tD+Tp )s. (2)

The aircraft transfer function in the autopilot pitch altitude control is as follows [4]:

Wap (s) =

2

KH wa

tas(s + w1)(s 2 + 2Xw2 s + co2) where ta, w1, w2, wa, X, KH are model parameters.

To describe the controller - pilot psycho-physiological characteristics let's use model 3 in table.

In addition, we assume that Wsur = 1.

Application of the algebraic criteria in the study of primitive loop stability characteristics represented in model 1 is challenging due to the high-order characteristic equation obtained by the expansion of the delay transfer function component in power series, and the high-order polynomial of the transfer function's denominator (2).

There are methods reducing components (members) with sufficient accuracy of delay elements' approximation, for example, number Pada. However, it is more convenient to study the system with time delay using the Nyquist frequency method [16-18].

The approach of Boeing 737 regional passenger plane served as a model experiment.

The modeling was conducted in the MATLAB integrated computing environment using the capabilities of the Control System Toolbox package [5].

The modeling has shown negativity of all performance equation roots (fig. 3) of the limit open-loop system ( ti = 0 ).

0.935

0.88

0.І       0.66   048 025

0.975

0.994

9  x 8      7      6      5       4   о 3       2       1

-2

Fig. 3. Location of characteristic equation roots in open loop system (2)

This has led to the conclusion about its stability [17].

The study in the systems with delay has demonstrated a significant reduction in the system stability with unsteady control parameters.

This concerns both the ATC system characterized by  air  traffic  intensity  in  the  control zone,

3

2

1

meteorological characteristics of the environment, flight patterns and other parameters of ATC radio data transmission devices.

The factors are shown in fig. 1 as parallel connected models to the controller model, each model combines a pilot model and an aircraft model multiplicatively [16; 18; 20].

Fig. 4 demonstrates the family Nyquist curves for system (2).

-2        -1 0 1 2 3 4

Re

Fig. 4. Nyquist curves for different values of delays t   + tFD :

1 - 0.6 s;

2 - 2 s;

3 - 4 s

Its analysis indicates loss of systems' stability with a four-second delay.

Conclusions

1. Characterization of the poly-linguistic channel in the interaction with the components comes up with recommendations for reducing load on controller's short-term and long-term memory and increasing efficiency and reliability of the controller - pilot communication.

2. The analysis of the existing man-operator mathematical models shows that the delay is a general characteristic integrally reflecting the impact of the poly-linguistic load on the air traffic control loops.

3. The transient delays caused by the poly-linguistic loads on the primitive control loop seriously undermine stability margin that may switch to the off-design critical unstable modes.

4. To eliminate these negative effects one should adjust phase characteristics of the primitive control loop components.

References

1. World Wide CNS/ATM Systems Implementer // ICAO, RIO Conference 1998.

2. Demystifying CNS/ATM // Report CANSO CNS/ATM Working Group, Final Version. June

1999.

3. Циркуляр ІСАО 9750-AN/963, 2007. [ІСАО   Circular   9750-AN/963,   2007.] (in

Russian).

4. Hoekstra, J.; Ruigrok, R.; Gent, R. 2000. Free Flight in a Crowded Airspace? - Napoli: 3rd USA/Europe Air Traffic Management R&D Seminar, 13-16 June.

5. Cloerec, A.; Zeghal, K.; Hoffman, E. 1999. Traffic Complexity Analysis To Evaluate The Potential For Limited Delegation Of Separation Assurance to the Cockpit. - IEEE Trans: 5.A.5-1 -5.A.5-8.

6. EUROCONTROL.  Long-Term Forecast of

Flights (2004-2025). 2004. 32 p.

7. Крыжановский, Г. А. Введение в прикладную теорию УВД. - Москва: Машиностроение, 1984. - 364 с.

Страницы:
1  2 


Похожие статьи

V Kharchenko, V Chepizhenko, Y Zaitsev - Impact of poly-linguistic load on air traffic control and monitoring quality