Gucma M., Gucma S. An optimisation method for constructing a pilot navigational system.

AN OPTIMISATION METHOD

FOR CONSTRUCTING A PILOT NAVIGTIONAL

SYSTEM

Gucma M., Gucma S.

Abstract: The article presents a method of optimising the navigational information system and

the user’s interface of a pilot navigational system. The method worked out was applied for the construction of prototypes of a stationary and portable navigational system, prepared within the framework of an intentional project by the scientific team of the Maritime University of Szczecin.

1. Introduction

Navigation in restricted waters is often referred to as pilotage or pilot navigation. In the process of navigating in restricted waters, because of the fast changes in the vessel’s posi-tion in relaposi-tion to objects ashore, the observed and the reckoned posiposi-tions are not marked on the chart, as in navigation in unrestricted and coastal areas. The vessel’s position is determined in the mind of the pilot or the master conducting the ship. In the process of con -ducting pilot navigation the pilot can be supported by the PNS (Pilot Navigation System). Currently, there are a few solutions of pilot navigational systems produced in the world. These systems are constructed on ECS basis (systems of electronic charts) or ECDIS (sys-tems of imaging electronic charts and navigational information), the latter being a detailed development of the former. Their common characteristic is the vessel imaged on the elec-tronic chart in the shape of an outline called “conventional waterline”. The accuracy of the PNS depends on the positioning system applied and ranges from 1m to 20m.

The basic deficiencies of the PNSs currently produced are:

 the information presented is not the optimal information which causes it not to be

taken advantage of in the utmost degree and there are difficulties with its being ab -sorbed by the pilot,

 lack of special images useful in pilotage navigation, like: in relation to the shore, in

 lack of optimal user interface,

 lack of a manoeuvre prediction system.

These deficiencies result from the systems being only modernizations of the systems func-tioning in unrestricted water areas (ECS or ECDIS) for the needs of pilotage, and were not worked out by scientific methods. A team of scientists from the Navigational Department of the Maritime University of Szczecin, within the framework of a project co-financed by the Ministry of Education and Science, undertook to work out the optimal solution for a pilotage navigational system, making use of scientific methods of constructing naviga-tional systems. As a result of research carried out, two PNS prototypes emerged:

 a stationary one, designed for sea ferries,

 a portable one, designed for pilotage.

At present these prototypes are undergoing experimental research and are being prepared for starting production. The following elements make up these systems:

 subsystem of electronic charts,  positioning subsystem,

 information processing and imaging subsystem.

The present article describes a method of optimising the PNS solution applied for the con-struction of this system’s prototype.

2. Assumptions of the navigational pilot system optimisation method

The pilot navigational system is designed for carrying out navigation in restricted water areas (pilotage navigation). Pilotage navigation consists in performing three following tasks by the PNS:

 planning a safe manoeuvre.

 determining the location of the vessel in a given area with definite accuracy.

 controlling the vessel’s movement providing for safe performance of the planned

ma-noeuvre.

The planned manoeuvre should fulfil the conditions (criteria) for safe navigation in re-stricted areas. The general condition of navigational safety in these areas can be written down as follows [4]:

dijk(1-)  D(t)

h(x, y, t)  T(x, y, t) + (x, y, t) (1) where:

D(t) – available navigational area (fulfilling the condition of available depth at moment t),

dijk(1-) – safe manoeuvring area (traffic lane) of the of the i-th vessel

perform-ing the j-th manoeuvre in k-th navigational conditions determined on

confidence level 1-,

h(x, y, t) – depth of the area at the point with coordinates (x, y) at moment t, T(x, y, t) – vessel’s draft at area point with coordinates (x, y) at moment t, ∆(x, y, t) – underkeel clearance at area point with coordinates (x, y) at moment t. This signifies that in order to meet the condition of safe navigation the width of the traffic lane (safe manoeuvring area) must be contained in the available navigational area fulfill-ing the condition of safe depth. This task boils down to:

 establishing the vessel’s safe speed,

 establishing safe underkeel clearance,

 planning the tactic of particular manoeuvres in the restricted area.

Establishing the vessel’s location boils down to establishing the location of her “conven-tional waterline” in relation to the safe isobath or coastline, or the fairway axis. The loca-tion of the vessel waterline in relaloca-tion to the safe isobath is determined in the navigator’s thinking process in the following stages:

 acquiring a good knowledge of local navigational conditions, including the content of the navigational chart of the area where navigation is performed,

 determining the vessel’s position identified with position of the observation point,  determining the location of the vessel’s “conventional waterline” in relation to the safe

isobath, after taking into consideration the vessel’s parameters and course.

Controlling the vessel’s movement facilitating the safe performance of a planned manoeu-vre consists in making decisions concerning the settings of the rudder, the engines, thrusters and tugs. The decision concerning these settings is made by the navigator in the following stages:

 establishing the current location of “conventional waterline” in the manoeuvring area

and the vessel’s movement parameters,

 determining the planned location of the vessel’s waterline in the manoeuvring area

and her parameters,

 determining the parameters of the manoeuvre performed with consideration of the navi-gational conditions prevailing there (hydrometeorological and bathymetric conditions),

 working out a decision concerning the settings of rudder, engines, thrusters and tugs.

Taking into consideration the tasks set for the PNS system and the experience concerning pilotage of similar systems in Poland and in the world, the assumptions for the construc-tion of a pilot navigaconstruc-tional system can be set forth as follows [3]:

1. The system is built specially for conducting navigation in restricted areas (pilotage nav-igation). It has characteristic features like: integration with the water area, integration with the vessel, optimal information, optimal user’s interface, predicting and support-ing manoeuvrsupport-ing.

2. The system is built in two versions: portable for pilots (small dimensions, autonomous power supply), stationary, designed for ferries and vessels constantly moving in defi-nite restricted areas.

3. Apart from standard imaging, the system also has special imaging designed for per-forming definite manoeuvres in restricted areas, like: shore orientation for the mooring manoeuvre, fairway orientation for the manoeuvre of passing on a fairway.

4. The information presented on indicators is optimal information, determined on the ba-sis of: research done by expert pilots and ferry masters, simulation research conducted with pilots and ferry masters, real research carried out on sea ferries and vessels navi-gating in restricted areas.

Information indispensable for the safe performance of a planned manoeuvre can be di-vided into three kinds:

 information about the area concerning safe isobaths of the closest obstacles to

naviga-tion and naviganaviga-tional marks (L1),

 information about the location of the vessel’s waterline in the area (L2),

 information about the predicted trajectory of the vessel in the area, that is determina-tion of the future movement trajectory of the vessel in the area, in a definite time scale with current settings (L3).

With a task so established and its limitations, the function of purpose, which is informa-tion used for performing the planned manoeuvre in definite navigainforma-tional condiinforma-tions, will be minimised:

Q

Q

min







Qxy  information used in the researched system for the safe performance

of manoeuvres in area (xy);

Qijk1xy  information concerning area (xy) for the i-th vessel, performing the

j-th manoeuvre in k-th navigational conditions;

Qijk2xy  information concerning position (xy) of the i-th vessel, performing

the j-th manoeuvre in k-th navigational conditions;

Qijk2xy  information concerning prediction (xy) for the i-th vessel,

perform-ing the j-th manoeuvre in k-th navigational conditions;

Rijkxy  navigational risk of performing the j-th manoeuvre by the i-th vessel

in k-th navigational conditions in area (xy);

dop lxy

R  admissible navigational risk in area (xy);

dop lxy

Q  maximum amount of information of l-th kind which the pilot or the master may take advantage of in area (xy);

Zxy  the cost of building a system in area (xy);

akc xy

5. The system has an optimal user’s interface constructed specially for pilots and ferry masters. This interface was constructed on the basis of simulation research conducted with pilots and real research conducted on the Świnoujście  Szczecin fair-way.

6. The system has a specially modified chart in its own format, containing ele-ments indispensable for conducting pilotage navigation, for instance, the possibility of building an accurate, safe, dynamic isobath. It is also possible to take advantage of standard charts in accordance with IHO57, as also other systems of electronic charts. 7. In both systems, stationary and portable, prediction of the performed

ma-noeuvre will be applied. In the stationary system mama-noeuvre prediction will be based on an accurate hydrodynamic model identified with 40 – 60 parameters of a given ves-sel. Prediction accuracy will amount to about 10% parameters of the manoeuvre per-formed (linear and temporal). In the portable system manoeuvre prediction will be based on extrapolation of current movement parameters of the vessel determined on the basis of the positioning system in the general hydrodynamic model of the vessel identified with 5 – 10 parameters. Prediction accuracy will amount to about 20% pa-rameters of the manoeuvre performed (linear and temporal).

8. Methods of supporting manoeuvres will be applied in the systems, e.g.: dynamic ves-sel domains.

2. The method of optimal solution for a pilot navigational system

The basic research problems that emerged at the construction of the PNS can be formu-lated as follows:

Configuration of a positioning subsystem ensuring the following parameters:

1. PNS stationary system:

 continuity of position fixing,

 accuracy of each point of conventional waterline not smaller than 3.0 m.

2. PNS portable system:

 continuity of position fixing,

 accuracy of each point of conventional waterline not smaller than 5.0 m.

Experimental research on the positioning system was conducted on the ferry m/f „Jan Śni-adecki” at the port of Świnoujście and on its approach. The base position of the vessel (the location of “conventional waterline”) was ensured by 2 RTK receivers installed at various established points of the vessel, cooperating with a reference RTK receiver installed at the Creative Work Centre at Świnoujście, owned by the Maritime University of Szczecin. The following configurations of the positioning system were subjected to research:

1. 2 independent DGPS receivers – large distance between antennas (of the order of 20 m);

2. 2 independent EGNOS receivers – large distance between antennas (of the order of 20 m);

3. 2 correlated DGPS receivers – small distance between antennas (of the order of 1 m); 4. 1 DGPS receiver plus gyrocompass;

5. 1 EGNOS receiver plus gyrocompass;

6. 1 DGPS receiver plus electronic (magnetic) compass.

A statistical analysis of results of experimental research conducted based on 15,000 mea-surements showed that the most favourable solution from the accuracy point of view is re-spectively:

 stationary PNS: positioning subsystem no. 4 made up of 1 DGPS receiver and gyro-compass (maximum directional error of conventional waterline points p =



 portable PNS: positioning subsystem no. 3 made up of 2 correlated DGPS receivers with distance between antennas in the range 1 m (p =



Fig. 1 presents the obtained distribution of directional error points of „conventional water-line” for the positioning system composed of two independent DGPS receivers. The relia-bility of satellite systems determined in the course of research amounts to respectively: DGPS system – 98.75 %, EGNOS system – 95.8 %.

Fig. 1. Distribution of directional error points of the conventional waterline of m/f „Jan Śniadecki” for the positioning system composed

of 2 independent DGPS receivers (confidence level 0.95)

The construction of an electronic chart in a format adapted for pilotage navigation, char-acterised by the following parameters:

1. It is a vector chart of restricted areas where pilotage takes place.

2. It has a data base adapted with respect to content, structure and format in-dispensable for conducting pilotage navigation.

3. The data base contains:

 standard navigational information used in the process of pilotage navigation,

 additional navigational information like current soundings, fairway axes, fairway

slopes etc.

4. The data base is updated only in the case of:

 the entry of new corrections concerning a given area, issued officially by an

autho-rized hydrographic office,

 performing new soundings of the area by a responsible maritime administration.

5. The chart has a special program servicing it, meeting the requirements of pilotage navigation.

The construction of a prediction system has been solved in different ways in both ver-sions of the system:

1. In the stationary PNS the prediction subsystem was based on the vessel’s precise hydro-dynamic model. This model is built individually for the vessel where the PNS is in-stalled, and the change of vessel requires a change in the model parameters.

2. In the portable PNS the prediction subsystem was based on extrapolation of the ves-sel’s movement parameters (speed, acceleration) determined with the application of the positioning subsystem.

The construction of an optimal system of information imaging and the user’s interface in the PNS has been reduced the following problems:

1. In which navigational coordinates the situation should be imaged and whether the kind of coordinates depends on the type of manoeuvre performed?

2. What scale should the situation be imaged in and whether it depends on the type of ma -noeuvre performed?

3. What chart contents should be presented on the indicator?

4. What alphanumeric information should be presented on the indicator?

5. What should be the size of the PNS indicator screen and where it should be located in the stationary and portable variant?

6. What should be the layout of information presented on the screen?

7. How should graphic information, and how should alphanumeric information be pre-sented?

8. How should optic and acoustic signalisation be solved? 9. How should the PNS system control be solved?

Research on optimising information presented on the PNS indicator and the optimal solu-tion for the user’s interface was carried out by using a specially prepared three-stage opti-misation method. These stages are: expert research, simulation research, real research. The idea of the prepared method was to construct a PNS prototype based on results ob-tained after completing each stage of the research. The constructed prototype was tested in

the next stage of the research. After the third stage the prototype research was completed and the PNS entered the design and production stage.

Stage I. Expert research was conducted in a group of 13 pilots working in the Szczecin and Świnoujście Port Team and of 8 masters of sea ferries from Euroafrica, Polish Baltic Ship-ping (PŻB) and Unity Line. The method of expert questionnaires was used, characterised by independence of the experts’ opinions and anonymity of judgments pronounced. The ques-tionnaire was constructed in the form of 14 questions intended to provide information con-cerning the required operational PNS parameters. The results of expert research have been presented in Table 1. They served the purpose of constructing a PNS prototype tested in the next stage (simulation research).

Table 1. Results of PNS expert research [3]

No. Description

of function

Portable system Stationary system

1 2 3 4

1. _{Screen size 10 in. 15 in.}

2. Indicator location Portable by pilot On the manoeuvring stand

Orientation Lack of unequivocal answers. The following imaging types should be

tested by simulation research (depending on manoeuvre performed): in relation to N, in relation to course, in relation to fairway axis, in relation to shore.

3. _{Work scale} variable, signalisation of sufficient accuracy.

4. There should be seen in

front of the vessel’s bow on the fairway

distance variable depending on speed, not less than 2 x L

5. Position shifting automatic position shifting from the screen centre to its end with

the possibility of manual shifting.

6. _{Indicator contents –}

con-stantly displayed

shore line, fairway axis, safe isobath, beacons, buoys, leading lines, cables, pipelines, optical waveguides, buoy or beacon num-ber or their symbolical markings.

7. Indicator contents

–dis-played on request

fairway slopes, light sectors, fairway axis kilometres.

8. Vessel marking on the

indicator

vessel marked with maximum outline, course line should pass through the whole vessel and the whole indicator up to the bow.

9. Signalisation moment of starting a turn on fairways, work scale too small.

11. Alphanumeric information – constantly displayed (maybe with on-and-off op-tion)

course, linear speed, lateral speed, angular speed.

12. Alphanumeric information

–displayed on request distance of bow, stern or board from fairway axis, distance of bow, stern or board from shore.

quest (time interval to be se-lected in a definite range)

(vessel silhouettes – maximum outline).

Stage II. Simulation research was carried out on a specially constructed interactive model consisting of three following elements: PNS system (without positioning and prediction subsystems), simulation model of vessel traffic, navigator (pilot, master). A general dia-gram of a simulation model construction and the information flow between its elements has been presented in Fig. 2.

Fig. 2. A general diagram of a simulation model construction and the information flow in the model

The simulation model was constructed with the use of two computers. One worked as a PNS system, and an electronic chart was imaged on its indicator with the vessel’s current silhouette. On the other computer a vessel traffic model was installed, and its indicator was the model control interface. This interface was designed in two versions, for a single-screw vessel and a double-single-screw vessel.

The range of simulation research covered the performance of the following manoeuvres: 1. Port approach manoeuvres, conventional 5,000 DWT vessel (single screw, no thruster):

manoeuvre of entering anchor position – imaging in relation to North.

2. Port entry manoeuvre: entering a deepened fairway (route) – imaged in relation to North and the fairway axis, conventional 5,000 DWT vessel; entering the port of Świnoujście – imaged in relation to North and fairway axis, conventional 40,000 DWT vessel.

3. Fairway passage manoeuvre – ferry m/f “Jan Śniadecki”: deepened unmarked fairway (designated fairway axis) – imaged in relation to North and fairway axis (axis change every 10 and every 20); fairway restricted by safe isobath with designated fairway axis, imaged in relation to North and fairway axis (axis change every 10 and every 20).

4. Turning manoeuvre – bulk carrier LC = 260 m, T = 8.5 m and 2 tugs 2500 BP: turning

basin designated with safe isobath and limited with two shores – the Northern Basin at Świnoujście – imaging in relation to North.

5. Mooring manoeuvre: starboard side mooring manoeuvre of a conventional 5,000 DWT vessel to rectilinear quay -imaged in relation to North and shore; port side mooring

ma-PNS without positioning and

prediction subsystem Simulation model of vessel

traffic

Navigator (pilot, master)

noeuvre of ferry m/f “Jan Śniadecki” to quay nr 3 Sea Ferry Base at Świnoujście – im-aged in relation to North and shore.

Those taking part in the research were pilots working on the Świnoujście – Szczecin fair-way and masters with experience in navigation in restricted areas. Each of them per-formed a full set of exercises and after each exercise and set of exercises filled separate expert questionnaires. The expert questionnaires were constructed as a set of questions concerning user’s interface parameters and preferences concerning the kind of imaging. For the assessment of particular manoeuvres performed at different kind of PNS imaging, a series of parameters were used depending on the kind of manoeuvres performed. These were respectively:

1. anchoring manoeuvre: distance of the dropped anchor from the assigned anchoring point, vessel speed at the moment of dropping the anchor, anchoring time;

2. fairway entry manoeuvre: distance of fairway entry from initial buoy, time of fairway entry;

3. fairway passage manoeuvre and port entry manoeuvre: width of traffic lane; 4. turning manoeuvre: size of manoeuvring area, manoeuvring time;

5. mooring manoeuvre: energy of first contact with the quay, distribution of points of the vessel’s first contact with the quay (or its stopping).

For example, in Fig. 3 the fairway passage manoeuvre by ferry m/f “Jan Śniadecki” has been compared using three different kinds of imaging of the PNS navigational situation. As assessment criterion, the traffic lane width determined on confidence level 0.95 has been applied. An analysis of the diagram unambiguously defines the most favourable so-lution, which is imaging in relation to the fairway axis (fairway coordinates) with axis change every 10.

Fig. 3. Traffic lanes of m/f “Jan safe isobath fairway axis

Śniadecki” on confidence level 0.95 during fairway passage limited by safe isobath with designated fairway axis

By application of the traffic lane width criterion, the performed statistical analysis of sim-ulation research on the manoeuvres of fairway passage and port entry has shown that (Fig. 4):

 when designating fairways on the indicator, not only the fairway axis but the safe

iso-bath should be used as well,

 the most favourable imaging is imaging in relation to the fairway axis (fairway

coordi-nates) with axis change every 10 [2].

This has also been corroborated by expert questionnaires conducted during simulation re-search.

Fig. 4. Frame diagram of the mean and standard deviation of traffic lane width for m/f „Jan Śniadecki” during fairway passage with designated fairway axis

Fig. 5. Frame diagram of the mean and standard deviation of traffic lane width for m/f „Jan Śni-adecki” during fairway passage limited with safe isobath with designated fairway axis

Applying the criterion of energy of the vessel striking against the quay, the performed sta-tistical analysis of simulation research results of mooring manoeuvres has shown that the most favourable imaging is imaging in relation to quay (shore coordinates). This has also been corroborated by expert questionnaires of simulation research [1].

Stage III.

1. Real research for the stationary PNS system was carried out on the ferry m/f „Jan Śni-adecki” for the following manoeuvres: fairway passage (Northern Fairway), entry and leaving the ports of Świnoujście and Ystad, mooring and unmooring in the ports of Świnoujście and Ystad.

2. Real research for the portable PNS system was carried out on:

 on the ferry m/f „Jan Śniadecki” for the manoeuvres: entry and leaving the ports of Świnoujście and Ystad, mooring and unmooring in the ports of Świnoujście and Ys-tad,

 various vessels entering or leaving the port of Świnoujście for the manoeuvres:

fair-way passage (Northern Fairfair-way), entry and leaving the ports of Świnoujście and Ys-tad, mooring and unmooring in the ports of Świnoujście and YsYs-tad,

 various vessels entering or leaving the port of Szczecin for the manoeuvres: passage on the Świnoujście – Szczecin fairway, entry and leaving the port of Świnoujście, mooring and unmooring in the port of Szczecin.

The range of the experiment covered about 500 hours and has corroborated the simulation research results concerning the basic parameters of the prototype of optimal system of imaging and user’s interface. The results of this research have led to only minor correc-tions of the user’s interface in the range of graphic markers.

3. Conclusions

The paper presents the results of research conducted by the scientific team of the Marine Traffic Engineering Institute at the Maritime University of Szczecin, within the frame-work of an intentional project, co-financed by the Ministry of Education and Science. A special three-stage method has been worked out for constructing the optimal system navi-gational information of orientation and the user’s interface of the pilot navinavi-gational sys-tem.

Using the method worked out, two PNS prototypes have been built: a stationary one de-signed for sea ferries, a portable one dede-signed for pilotage. The parameters worked out are significantly better than the parameters of devices of this type existing in the world, as those produced currently are merely modernised ECS and ECDIS systems serving the purpose of navigation in sea areas (unrestricted ones).

References

1. Gucma M., Ślączka W.: Badania symulacyjne wskaźnika PNS dla manewru cumowania (Simulation research of PNS for mooring manoeuvre). ZN nr 6(78), AM Szczecin, 2005.

2. Gucma M., Ślączka W.: Badania symulacyjne wskaźnika PNS dla manewru przejścia torem wodnym (Simulation research of PNS for fairway passage manoeuvre). ZN nr 6(78), AM Szczecin, 2005.

3. Gucma S.: Optymalne rozwiązanie pilotowego systemu nawigacyjnego (Opti-mal project of pilotage navigational system). ZN nr 6(78), AM Szczecin, 2005. 4. Gucma S.: Nawigacja pilotażowa (Pilotage navigation). Fundacja Promocji