Article citation information:
Ogonowski, K., Kozuba, J. Possibility of
processing parameters obtained from on-board flight data recorders for
diagnostics and predicting reliable operation periods of important aircraft
equipment. Scientific Journal of Silesian
University of Technology. Series Transport. 2017, 95, 133-143. ISSN:
0209-3324. DOI: https://doi.org/10.20858/sjsutst.2017.95.13.
Krzysztof OGONOWSKI[1],
Jarosław KOZUBA[2]
POSSIBILITY OF PROCESSING
PARAMETERS OBTAINED FROM ON-BOARD FLIGHT DATA RECORDERS FOR DIAGNOSTICS AND
PREDICTING RELIABLE OPERATION PERIODS OF IMPORTANT AIRCRAFT EQUIPMENT
Summary. In recent years, much attention has been paid
to the recording of flight data in order to provide objective information
concerning routine and emergency in-flight situations. The data recorded can be
used for assessing the air mission execution by the pilot and flight safety
breaches, for specifying some damage to the equipment, and for preventing
faulty equipment from being used in flights. Flight data recorders (FDRs) are
commonly known as “black boxes”. This article presents issues related to flight
data acquisition and preparing the data for later use, as well as their impact
on flight safety. The systems recording and processing selected parameters not
only enable ongoing diagnostics, but also make it possible to predict the
period of further reliable operations and to analyse the causes that led to
possible damage. In addition to improving economic indicators of the aircraft
operation, flight safety is improved. Obviously, the article will only discuss
certain ways of enhancing safety by applying the processing of data obtained
from FDRs. Nevertheless, I believe it will offer an overall view of how
important it is to collect, process and properly analyse such data for
diagnostics, prediction and flight safety.
Keywords: flight data recorders; diagnostics; prediction; operation period; aircraft equipment.
1. INTRODUCTION
The FDR is an on-board device
designed to record basic parameters characterizing the flight conditions, such
as the position of the controls, the state of the individual instruments and
installations, the radiotelephony communications conducted by the crew, the
internal communication, the cockpit background sounds, the images from the
cameras recording the operation of onboard instruments, the imagery from other
sources (e.g., FLIR cameras or LLLTV), and even the pilot’s psychophysical
condition.
Although the elimination of
accidents and major air incidents remains an ultimate goal, it is believed that
aviation cannot be completely free from hazards and related risks. There is no
guarantee that human activity and man-made systems will be completely free from
operational errors and the consequences thereof. Therefore, safety is a dynamic
feature of aeronautical systems in which safety hazards must be mitigated
constantly.
The fatigue life of the aircraft’s
airframe depends on the manner in which the user operates the aircraft.
Manoeuvre flights generate significantly higher structural loads in comparison
to straight and level en route flights at a constant flight level. The bases
for investigating the manner in which the user operates an aircraft are the
flight data recorded in the on-board recorders. Modern aircraft, however,
record a very large number of various parameters, including those that can be
used for the assessment of the operation of their individual components. In
order to do this, a technical analysis is performed, the main purpose of which
is to detect conditions dangerous to the service life of the engine and other
equipment. Numerous algorithms are developed, in a similar way as for
operational analysis, which can detect excessive fuel consumption, overheating,
vibrations or irregular engine operation.
The propulsion system is one of the
most important and expensive parts of the aircraft. Flight safety depends on
reliable operations. At the same time, aircraft are highly susceptible to
damage due to the extremely high mechanical and thermal loads. Therefore, when
seeking to reduce operating costs, along with maintaining the required
reliability of the power plant, the methods of determining its current
technical condition were sought. This led to the development of aircraft
turbine engine diagnostics and the introduction of the condition-based
maintenance (CBM).
Technical diagnostics represents a
branch of knowledge, which is developing rapidly. It exerts a positive impact
on improving and streamlining the use and operation of technical objects. It
provides the technical and practical basis for the optimum (safe, reliable and
cost-efficient) operation of a technical object in accordance with its
“operating condition”.
As a branch of science, diagnostics
has clearly defined the objectives and tasks undertaken, whose objective is the
evaluation, genesis and prediction of the technical object’s operating
condition using broadly understood indirect research methods. It may,
therefore, be assumed that the purpose of diagnostics is to determine the
activities, the operating decision, methods, commands, and means that will lead
to the achievement of a detailed purpose. Technical diagnostics is based on its
own, as well as original, principles and methods of testing objects and
processes.
The general principle of diagnostics
can be expressed in the words of W.R. Ashby: “the conditions of an object are
dependent on its history”.
This statement constitutes the basis
of diagnostic methods and any other accompanying activities. Diagnostics is an
interdisciplinary science, which by itself results from the scope and
potentialities of its application, including the construction and operation of
machines, automatics, mechanics, electrical engineering, electronics and ergonomics.
Technical objects are subject to
wear-and-tear processes, which are due to the wearing-out of their components
resulting from friction, corrosion, ageing, material fatigue etc. Due to
construction deviations, and technological and material differences, the damage
is not the same for all components of the same object, nor is it the same for
identical elements within a set of objects of the same type. This applies, in
particular, to complex technical components of an aircraft.
Aeronautical structures are
characterized by specific design features, including the low mechanical safety
factor of components, multiple joints and splines, and high durability and
reliability under the conditions of mechanical and climatic exposure.
Consequently, all on-board units are exposed to strong vibrations, which hamper
the identification of unfit components using, for example, conventional
vibroacoustic methods. Moreover, for obvious reasons, aircraft propulsion units
are designed so as to minimize aerodynamic drag and weight, resulting in all
elements being close together, thus making it difficult to further distinguish
diagnostic signals coming from individual kinematic pairs. This is the reason
why new diagnostic methods should be sought.
2. PROCESSING INFORMATION RECORDED IN THE
OPERATIONAL RECORDER FOR THE PURPOSE OF DIAGNOSING THE CONDITION
OF THE AVIATION TURBINE ENGINE
The implementation and widespread
dissemination of aircraft engine parameter recording systems were possible
owing to the benefits that result from their use. First of all, engine
maintenance systems were changed from a system based on the period of reliable
operation guaranteed by the manufacturer (according to the so-called “time
between overhauls”), to a much more economical CBM system based on the engine’s
current operating conditions. This progress was financed primarily from the
budgets of the largest armies, as it allowed for the increased “combat
readiness” of aircraft, owing to the earlier detection of various propulsion
system failures. This, in effect, resulted in a reduced number of aircraft
used. Engine manufacturing companies also showed interest in the implementation
of those recording systems. The objective records of the operation make it
possible to show that the damage results from non-compliance with the operating
requirements and, especially, from exceeding the allowable parameter values.
This means that these systems offer sufficient proof to dismiss many of the
user warranty claims. Meanwhile, the sums in question are huge: the cost of a
jet engine may be as high as several million dollars, and the cost of its
repair may exceed a million. The introduction of automatic recording of engine
operating and flight parameters was also supported by insurance companies, as
this contributed to increased flight safety, fewer accidents and lower
compensation payments. Carriers who use the methods described above in order to
maintain a high level of flight safety pay lower insurance premiums, which
reduces the airline operational costs and ensures the safety of passengers who
use this means of transportation more and more frequently.
FDR design has evolved from
automatic pens marking changes in selected parameters on a paper tape, through
multitrack tape recorders, to modern data recorders equipped with semiconductor
memory. State-of-the-art FDRs not only record the changes of many engine and
airframe operating parameters, but are also equipped with software that
processes signals from sensors, as well as generates real-time warnings or recommendations
to the air crew. More and more frequently, there are systems that allow data
transmission from a flying aircraft to users’ or manufacturers’ ground
facilities.
When diagnostic systems were
introduced, many methods were developed that could be divided into four groups:
- Checking selected engine parameters (parametric
method)
- Vibroacoustic diagnostic methods
- Endoscopy (using fibre optics)
- Oil consumption testing and spectroscopic analysis of
wear-out products (especially metals) accumulated in oil and filters
Operating parameter recording
provides data used in the parametric method and vibroacoustic diagnostics.
When the automatic systems recording
the propulsion and system operating parameters were introduced, both data
recording and the cost of each measurement channel posed a great problem. The
problem gradually subsided with the development of microprocessors and
semiconductor memory technology. In the 1980s, it was possible to record the
changes of a dozen or more parameters sampled several times a second during an
8 h flight on a 1.5 MB solid state semiconductor cartridge. This was possible
owing to various techniques, such as data compression. Today, even several
thousand parameters are recorded, with a large part of them being so-called binary
parameters (“on/off”). Such amounts of data are unsuitable for human
processing, especially with the growing number of aircraft equipped with data
recorders. It is, therefore, natural to automate this process so as not to
overlook the symptoms of potential emergency conditions, as well as in order to
detect those symptoms already during the flight, not after its completion.
While activities related to
measurement data recording have already become largely routine, the formulation
of algorithms to analyse those data requires expert knowledge. Specialists in
diagnostic systems should cooperate with engine designers, analysts and users
who know the “weak points” of particular engines. It is particularly important
to analyse the propulsion unit’s performance under dynamic conditions, such as
start-up, acceleration, deceleration or shutdown, as faults of the engine
components, especially of its control system, will then appear sooner.
Consequently, the first task is to
develop algorithms that recognize various phases of engine operation based on
the values of recorded parameters. In order to identify the dynamic states of
the turbine engine, Table 1 lists the characteristics of the engine and flight
parameter recording. They concern an engine with an electric starter and a
start-up fuel system activated with an electrovalve. The duration of the
observed records arising from the nature of the observed phenomena was
determined on the basis of the experience obtained in the research conducted
during the ground and in-flight tests of the aircraft. It is included in the
third column of Table 1. For the aircraft under examination, the duration of
the particular phases is as follows: start-up on the ground approximately 60 s,
airborne start-up approximately 2 min, acceleration approximately 10 s and
deceleration 15 s, rotor coast-down after switching off the combustion chamber
approximately 2 min.
Table 1. The characteristics of the
particular phases of the turbine engine operation recording
|
Engine operation |
Phase identification parameters |
Observed recording duration |
|
Start-up on the ground |
1.
Appearance of
the supply voltage in the electric starter 2.
Indicated
airspeed less than, e.g., 200 kph |
80 s from the moment supply
voltage was delivered to the starter |
|
In-flight engine start-up |
1.
Appearance of
the supply voltage in the start-up fuel valve 2.
Indicated
airspeed less than, e.g., 200 kph |
180 s from the moment supply
voltage was delivered to the start-up fuel valve |
|
Acceleration |
1.
Engine control
lever (ECL) position change 2.
Pressure rise in
the main injector manifold |
30 s after the commencement of the
ECL movement |
|
Deceleration |
1.
ECL position
change in the direction of decreasing the rotor speed in less than 0.5 s |
30 s after the commencement of the
ECL movement |
|
Shutdown |
1.
Closing the fuel
shut-off valve |
150 s from the moment of fuel flow
shut-off |
For example, the differences between
the airborne engine start-up and the ground start-up are as follows (Fig. 1):
- Idle speed is higher than on the ground and usually
depends on the flight altitude and speed
- The rotor is not driven by the energy received from
the starter, but by “windmilling”, i.e., the energy of the air flowing through
the engine channel (this energy depends on the flight speed and the air
density, i.e., the flight altitude)
Therefore, the algorithm for
analysing this phase of the engine operation includes:
- Checking whether the flight conditions are within the
so-called “engine performance envelope” specified for a given engine built on
the aircraft by the values of maximum altitude and airspeed
- Checking whether the “windmilling” speed has reached
the minimum value necessary to commence the start-up
- Calculating idle speed under conditions determined by
speed and altitude
Start-up duration Start-up injectors Time [s]
Fig. 1. Fragment of the rotor speed
change (n), exhaust temperature after turbine (Tc4), altitude (H) and indicated
airspeed (VIAS) during the airborne engine start-up [1]
The results presented were applied
in the engine diagnostic system of the I-22 Iryda aircraft. In addition to the
above, numerous other parameter processings are used in the diagnostic system:
observation of long-term trends, calculation of gas-dynamic indices, efficiency
of components etc.
3. INCREASING SAFETY THROUGH THE USE OF
ON-BOARD FLIGHT DATA RECORDERS AND PROGRAMS FOR THE ONGOING MONITORING
OF FLIGHT SAFETY
In its half-century history, the
analysis of flight parameters has undergone many changes. Today, flight
parameter analysis is used not only in air accident investigation, but also in
accident prevention. The detected breach that may affect flight safety or the
technical condition of the aircraft is, in the first phase, a signal to
immediately undertake tasks related to an additional or special inspection, and
then to thoroughly analyse the process of crew training and in-flight
methodology. Data from on-board flight recorders are used extensively in a
variety of optimizations: crew working times, route planning, aircraft rotation
etc. They also serve as the basis for programs that monitor the operation of
engines and other equipment, which make it possible to control fuel and oil
consumption, and even radiation levels at high altitudes. The results of
specialized analyses are sent to programs that calculate crew working time, as
well as the cycles and flying time of aircraft, engines and equipment.
For safety oversight purposes,
documentation is controlled, analysed and archived. Such documentation
includes:
- All documents received before the flight, records made
during aviation observations, weather forecasts, loading sheets etc.
- Recordings from an FDR (when the aircraft is equipped
with such equipment)
- Recordings from other flight data recording equipment
Recordings from the recorders are
archived on electronic media and must be read at a frequency that will allow
each flight to be analysed before “parameter overwriting” occurs.
Additionally, systematic reviews of
“flight records” recorded on aviation monitoring systems are performed within
the scope of air operation safety oversight.
The fatigue life of the aircraft’s
airframe depends on the manner in which the user operates the aircraft. Manoeuvre
flights generate significantly higher structural loads in comparison to
straight and level en route flights at a constant flight level. The bases for
investigating the manner in which the user operates an aircraft are the flight
data recorded in the on-board recorders. In the case of aircraft maintained
according to their technical condition, the recording, storage and processing
of content from onboard recorders constitute the basis of the Aircraft
Structural Integrity Program (ASIP) for ensuring the structural integrity. A
good example of this may be the F-16 aircraft purchased by Poland, for which
recordings from the on-board recorders will be collected within the framework
of the ASIP. Fatigue wear will be calculated for each aircraft individually on
the basis of its actual operating history.
Individual aircraft tracking (IAT)
is a standard activity within modern structural integrity programmes. For IAT
to be possible, the following conditions are to be met:
- All aircraft must be equipped with digital FDRs
- Data reading and archiving systems must be organized
at airbases
- The entire history of an aircraft must be known
4. AIRCRAFT CONDITION
MONITORING SYSTEM
Modern aircraft are equipped with
the means for performing multiple analyses and self-monitoring their systems.
These solutions integrate almost the entire avionics with other aircraft
components in order to both control them and determine their condition. At
present, an aircraft condition monitoring system (ACMS) offers great
capabilities for testing, diagnosing and helping with problem-solving. The
major benefit of using these systems is the reduction of operating costs as a
result of the reduced amount of maintenance required and, as originally
intended by their developers, the improvement of air safety as a result of the
reduced number of failures or non-optimal operating conditions.
What is an ACMS? It is a solution
that allows for an analysis of multiple aircraft systems on the basis of the
available parameters and reports, which may be obtained either in-flight, on
board the aircraft or at ground-based ground support equipment service
stations. The general operating principle of an ACMS and its interaction with
other avionic and communication systems are illustrated by Fig. 2.
As can be seen from the above
illustration (Fig. 2), the communication between the aircraft and the ground
station may be two-track: data transfer using a variety of media, and radio
communication. In the first case, information is both received and delivered,
enabling us to load the software into a plane, as well as modify the
assumptions according to different users’ needs. This is most commonly done by
means of a variety of data loader devices, such as floppy disc drives,
magneto-optical drives, CDs or memory card readers. The same device is usually
used for receiving information after the flight as a set of recorded parameters
or maintenance reports generated on the basis thereof. An increasingly popular
solution is the automatic bulk dump of the above data when the aircraft stops
at the airport stand, e.g., by means of wireless devices mounted in the air
bridges or through a LAN port located in an easily accessible place on the
aircraft side. Information obtained from the aircraft is then analysed by the
ground station and used in further applications, such as monitoring engine
condition or performance. An alternative way of transmitting information to the
ground is the radio, although, to date, more information is transmitted in this
way from the ground to the aircraft than vice versa. This is due to the fact
that, during the operation, the aircraft structures must not be interfered
with. Obviously, there are examples of ground-aircraft radio transmissions,
which do not involve an ACMS, rather involve, e.g., sending numerous auxiliary
data, such as a flight plan or weather reports.
Fig. 2. General schematic diagram of
an ACMS system [3]
5. ACMS: CAPABILITIES
AND LIMITATIONS
The functionality and scalability of
the ACMS system in a modern commercial airliner makes that particular system
very useful to any aviation operator. The most valuable functions
(problem-solving, flight parameter analysis, selection of data for monitoring
aircraft engine trends/condition) will be applicable in air operations based on
an extensive connection network. The capability of uploading analysis results
via radio, even during the flight, increases the benefits from the use of an
ACMS. In situations where immediate aircraft service action is required, the
ability to receive ACMS messages in real time via the Aircraft Communications
Addressing and Reporting System will be invaluable for the airline.
The complex and integrated system,
however, imposes high requirements. The large number of parameters involved
also requires an equally large number of sensors, transmitters, relays and
similar technical devices, which must be of the highest quality and maintained
in good operating conditions. This puts the highest demands both on the
manufacturer of the aircraft and on its operator. A number of innovative
solutions was incorporated into this area at the design stage of the prototype.
However, it was impossible to apply so many of the above-mentioned transmitters
powered by traditional electrical circuits, which is why manufacturers decided
to introduce virtual circuit elements, such as virtual fuses and relays.
6. CONCLUSIONS
Constantly improved FDRs play an
increasingly important role in ensuring the safety of air operations and
improving the operational performance of aircraft. They are used in both civil
and military aviation. Flight data recording systems can be used in almost
every aircraft from commercial airliners and helicopters to general aviation
aircraft. The complexity of problems related to contemporary air traffic
requires the use of modern techniques and methods in the management process, as
well as the conduct of broadly understood research. Dynamic technological
changes cause both aircraft handling and air traffic management to increase
skills and experience, which can no longer be gained solely by improving skills
on real objects under normal operating conditions. Furthermore, increasing
attention is being paid to the reduction of all costs associated with
conducting such activities.
All this leads to the search for
methods and tools that make it possible to replace the study of the actual
object (aircraft) and actual conditions with mathematical models, which offer a
high degree of accuracy in mapping the reality. In these models, it is
necessary to use the flight data recorded by on-board recorders.
The level of safety depends on many factors. Human and
technical factors can be mentioned here. Meanwhile, the research currently been
carried out, on a global scale, on factors related to the means of transport,
will significantly help to reduce their participation in contributing to
accidents and their consequences. An interesting non-invasive diagnostic method
is presented in [4-17].
Every modern ACMS is a powerful
maintenance tool, which enables fuller and quicker diagnostics of aircraft in
the air and on the ground. After a short period of operating the Embraer 170
aircraft, equipped with such systems, it was possible to observe that the time
between traditional inspections carried out by maintenance personnel increased
several times, while the scope of inspections was limited to necessary
maintenance, such as fluid replenishment, tyre replacement or mechanical
repairs. This has had a great impact on cost reduction. With proper design and
use, monitoring systems will improve safety indicators and the effectiveness of
air operations. Their development will continue towards the fullest possible
use of computers and avionics, automation, and possibly even robotics in the
future, in aircraft diagnostics and maintenance.
References
1.
Balicki W. 2011. The Processing of Information Written in the
Exploitation Recorder for the Purposes of Diagnosing the State of an Air
Turbine Engine (Works of the Institute of Aviation No. 213). Warsaw:
Institute of Aviation. ISSN: 0509-6669.
2.
Gębura A. 2010. Method of the Frequency Modulation of
Stretching Deck Generators in Diagnosing Driving Teams. Warsaw: Publishing
Company of the Technical Institute of the Air Force. ISBN: 978-83-61021-20-9.
3.
Witkowski
R., P. Lipiec. Aircraft Condition Monitoring System ACMS. Polskie Linie Lotnicze LOT S.A.
4.
Czech P. 2012. “Diagnosis of
industrial gearboxes condition by vibration and time-frequency,
scale-frequency, frequency-frequency analysis”. Metalurgija 51(4): 521-524. ISSN: 0543-5846.
5.
Czech P. 2011. “An intelligent approach to wear of
piston-cylinder assembly diagnosis based on entropy of wavelet packet and
probabilistic neural networks”. In: J. Mikulski (ed.). 11th International Conference on Transport
Systems Telematics. Katowice Ustron, Poland. 19-22 October 2011. Modern
Transport Telematics. Book series: Communications
in Computer and Information Science, Vol. 239: 102-109.
6.
Czech P. 2011. “Diagnosing of disturbances in the
ignition system by vibroacoustic signals and radial basis function -
preliminary research”. In: J. Mikulski (ed.). 11th International Conference on Transport Systems Telematics. Katowice
Ustron, Poland. 19-22 October 2011. Modern Transport Telematics. Book
series: Communications in Computer and
Information Science, Vol. 239: 110-117.
7.
Czech P. 2012. “Determination of the course of
pressure in an internal combustion engine cylinder with the use of vibration
effects and radial basis function - preliminary research”. In: J. Mikulski
(ed.). 12th International Conference on
Transport Systems Telematics. Katowice Ustron, Poland. 10-13 October 2012.
Telematics in the Transport Environment. Book series: Communications in Computer and Information Science, Vol. 329:
175-182.
8.
Czech P. 2012. “Identification of leakages in the
inlet system of an internal combustion engine with the use of Wigner-Ville
transform and RBF neural networks”. In: J. Mikulski (ed.). 12th International Conference on Transport Systems Telematics.
Katowice Ustron, Poland. 10-13 October 2012. Telematics in the Transport
Environment. Book series: Communications
in Computer and Information Science, Vol. 329: 414-422.
9.
Czech P. 2013. “Diagnosing a car engine fuel
injectors’ damage”. In: J. Mikulski (ed.). 13th
International Conference on Transport Systems Telematics. Katowice Ustron,
Poland. 23-26 October 2013. Activities of Transport Telematics. Book
series: Communications in Computer and
Information Science, Vol. 395: 243-250.
10.
Czech P. 2013. “Intelligent approach to valve
clearance diagnostic in cars”. In: J. Mikulski (ed.). 13th International Conference on Transport
Systems Telematics. Katowice Ustron, Poland. 23-26 October 2013. Activities of
Transport Telematics. Book series: Communications
in Computer and Information Science, Vol. 395: 384-391.
11.
Czech P. 2013. “Diagnose car engine exhaust system
damage using bispectral analysis and radial basic function”. In D. Zheng, J.
Shi, L. Zhang (ed.). Proceedings of the
International Conference on Computer, Networks and Communication Engineering
(ICCNCE). Beijing, China. 23-24 May 2013. Book series: Advances in Intelligent Systems Research, Vol. 30: 312-315.
12.
Czech P. 2013. “The Use of DWT Analysys and PNN Neural
Networks in Diagnostics of Gasket under Engine Head Damage”. In: B. Baksys,
A. Bargelis, S. Bockus, A. Fedaravicius, V. Leonavicius, P. Ziliukas, R. Dundulis,
T. Pilkaite (eds.). Proceedings of the
18th International Conference on Mechanika. Kaunas University of Technology,
Kaunas, Lithuania. 4-5 April 2013. Kaunas University of Technology. Book
series: Mechanika Kaunas University of
Technology: 58-61.
13.
Czech P., H. Madej. 2011. “Application of cepstrum and
spectrum histograms of vibration engine body for setting up the clearance model
of the piston-cylinder assembly for RBF neural classifier”. Eksploatacja i Niezawodnosc - Maintenance
and Reliability 4(52): 15-20. ISSN: 1507-2711.
14.
Czech P., J. Mikulski. 2014. “Application of Bayes Classifier
and Entropy of Vibration Signals to Diagnose Damage of Head Gasket in Internal
Combustion Engine of a Car”. In: J. Mikulski (ed.). 14th International Conference on Transport
Systems Telematics. Katowice Ustron, Poland. 22-25 October 2014. Telematics -
Support for Transport. Book series: Communications
in Computer and Information Science, Vol. 471: 225-232.
15.
Figlus T., M. Stańczyk. 2016. “A method for detecting damage to rolling bearings in
toothed gears of processing lines”. Metalurgija
55(1): 75-78. ISSN: 0543-5846.
16.
Figlus T., M. Stańczyk. 2014. “Diagnosis of the wear of gears in the gearbox using the
wavelet packet transform”. Metalurgija
53(4): 673-676. ISSN: 0543-5846.
17.
Madej H., P. Czech. 2010. “Discrete wavelet transform
and probabilistic neural network in IC engine fault diagnosis”. Eksploatacja i Niezawodnosc - Maintenance
and Reliability 4(48): 47-54. ISSN: 1507-2711.
Received 04.03.2017; accepted in revised form 04.05.2017
Scientific Journal of Silesian
University of Technology. Series Transport is licensed under a Creative
Commons Attribution 4.0 International License