609

1 INTRODUCTION

Maritime safety and operational efficiency are critical

challenges within the global maritime industry,

particularly in regions susceptible to piracy, natural

disasters, mechanical failures, and collisions [1-3]. Ship

robbery and piracy present significant risks to human

life and valuable cargo, underscoring the need for

robust emergency reporting systems [4-6]. Traditional

systems, such as the Global Maritime Distress and

Safety System (GMDSS) and Vessel Management

Systems (VMS), primarily rely on satellite

communications [7-10]. A major vulnerability of

satellite-based systems lies in their potential for signal

interception by malicious actors [11,12]. During

hijacking incidents, pirates can intercept distress

signals, compromising hostage safety and potentially

escalating dangerous situations. Additionally, the high

communication costs of satellite systems deter many

fishing ships, particularly those from economically

constrained communities, from adopting these

systems.

Numerous fishing ships utilize receive-only AIS

Class B systems to maintain the confidentiality of their

fishing grounds, inadvertently reducing the efficiency

of distress signaling during emergencies such as

piracy, fires, collisions, grounding, and sinking

incidents [13,14]. Historical events, including the

hijacking of fishing ships by pirates, highlight the

urgent need for secure and cost-effective emergency

communication solutions [15,16]. AIS is widely used in

Genetic Algorithm for Ship Robbery Emergency

Reporting System

T.H. Chang, S.L. Kao , C.C. Chou& H.C. Chang

National Taiwan Ocean University, Keelung, Taiwan

ABSTRACT: In contemporary maritime navigation, ships in distress primarily rely on satellite systems in

conjunction with radio systems within the framework of the Global Maritime Distress and Safety System

(GMDSS) to transmit distress signals. However, the insufficient confidentiality of satellite data enables pirates

engaged in ship hijacking to intercept these signals, potentially endangering the safety of hostages on board.

Additionally, the high communication costs associated with satellite information transmission often discourage

fishing ships from incurring these expenses. Given these cost constraints, this study seeks to develop an intelligent

emergency distress notification method integrated with the Automatic Identification System (AIS). Specifically,

this study introduces an innovative intelligent radio emergency notification system by incorporating the concept

of radio relay stations. The proposed system integrates the Genetic Algorithm (GA) with the Maritime Geographic

Information System (MGIS) as an alternative rescue method for ships in distress. The system collects all relevant

information from the distressed ship through shore stations, enabling it to respond to the ship and verify the

receipt of distress messages transmitted via AIS. The proposed method functions as an intermediary for distress

signal transmission and confirmation. By gathering ship positions, it establishes a mobile network for message

dissemination, thereby enhancing the reliability and efficiency of emergency distress communications at sea.

http://www.transnav.eu

the International Journal

on Marine Navigation

and Safety of Sea Transportation

Volume 19

Number 2

June 2025

DOI: 10.12716/1001.19.02.33

610

maritime operations for real-time tracking and

monitoring of ship movements [17]. It provides

essential data such as ship identity, position, speed,

and course, contributing to enhanced maritime

situational awareness. When integrated with advanced

algorithms, AIS data can significantly improve

maritime safety by identifying anomalies and

predicting potential threats.

Genetic Algorithms (GAs) are search heuristics

inspired by the principles of natural selection and

genetics [18,19]. These algorithms are particularly

effective in solving complex optimization problems

where traditional methods may fall short [20]. GAs

simulate evolutionary processes by employing

operators such as selection, crossover, and mutation to

evolve solutions over generations [21]. Through this

iterative process, GAs are capable of exploring large

search spaces efficiently and finding near-optimal

solutions, making them suitable for dynamic and

uncertain environments like maritime emergency

communication networks.

The main objective of this study is to develop a

secure and cost-effective emergency distress

notification system by integrating AIS, Genetic

Algorithm (GA), and Maritime Geographic

Information System (MGIS) to form an Intelligent

Radio Emergency Notification system as an alternative

to vulnerable and costly satellite-based systems. This

system employs a Genetic Algorithm (GA) to identify

the most effective distress relay return path by

selecting the route with the optimal Ship Escape Time

Index (SETI). GAs operate through processes such as

selection, crossover, and mutation to evolve solutions

over generations, aiming to find the optimal result

within a large search space.

In the context of the Intelligent Radio Emergency

Notification system, GA helps evaluate numerous

potential relay paths. It dynamically adjusts to

changing maritime conditions. It also optimizes the

route based on SETI. This approach ensures that the

shore station verifies whether the distress message was

mistakenly sent. Once verified, the system maintains

optimal communication timing within the mobile relay

communication network. By avoiding repeated

communication path searches, the system ensures that

acknowledgment messages are accurately delivered to

distressed ships. This reduces the risk of lost distress

messages.

The rest of this study is organized as follows.

Section 2 provides an overview of related research.

Section 3 presents research design and system

architecture. Section 4 discusses experimental results

and analysis. Finally, conclusions are presented in

Section 5.

2 METHOD

2.1 Automatic Identification System (AIS) relay

communication

The Automated Mutual-Assistance Vessel Rescue

System (AMVER) operates through the voluntary

participation of ships worldwide. Its primary purpose

is to provide rescue support during distress situations.

Ship position reporting systems are generally classified

into satellite-based and radio-based position reporting

systems. Currently, two primary satellite systems serve

as data transmission devices for ship monitoring.

These systems include the Advanced Research and

Global Observation Satellite (ARGOS) and the

International Maritime Satellite Organization

(INMARSAT) systems. Additionally, radio systems

such as AIS and Digital Selective Calling (DSC) provide

real-time reporting of ship identity codes and

positions.

AIS has been a mandatory navigation device since

2002. The International Maritime Organization (IMO)

established this requirement under the International

Convention for the Safety of Life at Sea (SOLAS) [22].

All newly constructed ships must be equipped with

AIS. This equipment assists watchkeepers and

enhances safety by effectively managing emergency

situations. The integration of the Global Positioning

System (GPS) with Very High Frequency (VHF) radio

technologies forms the foundation of AIS technology.

Both Class A and Class B systems are included. The

system employs Self-Organized Time Division

Multiple Access (SOTDMA) and Carrier Sense Time

Division Multiple Access (CSTDMA) technologies, as

shown in Table 1. The transmission frequency of

navigational information is dynamically adjusted

based on ship speed. It transmits both dynamic and

static data, including the Maritime Mobile Service

Identity (MMSI), call sign, ship name, coordinates,

course, and speed.

In this study, AIS is proposed as a relay station for

distress and acknowledgment messages. By collecting

the positions of various ships, the system forms a

mobile network for distress message transmission.

This approach expands the communication range of

radio systems. The system broadcasts radio signals

through base stations to ships within the effective

reception range. These ships can then forward

previously received data, creating a maritime mobile

communication relay network. When ship positions

meet specific spatial and temporal conditions, the AIS

relay function supports long-distance communication.

This capability ensures that ships in distress can

quickly send distress signals to shore-based command

centres or other ships. The relay network significantly

enhances rescue response speed and efficiency,

improving the safety of personnel at sea.

Table 1. SOTDMA and CSTDMA technologies in AIS

systems

Feature

SOTDMA

CSTDMA

Access Method

Time-slot reservation

Listen-before-talk

Synchronization

Requires GPS

synchronization

Does not require GPS

synchronization

Collision Avoidance

High, due to pre-

assigned time slots

Medium, relies on

carrier sensing before

transmission

Application

Primarily used in

Class A AIS systems

Primarily used in

Class B AIS systems

Transmission

Frequency Adaptation

Automatic based on

speed

Manually

configurable

2.2 Genetic Algorithm (GA)

This study employs a GA to optimize emergency

reporting routes. By analyzing AIS distress messages

collected from shore stations, the locations of relay

ships along different transmission routes can be

611

determined, facilitating the construction of a relay

network. The GA is then applied to identify the optimal

emergency reporting route based on the SETI. Since

GA mitigates the risk of local optima, it demonstrates

superior performance compared to other algorithms

within the communication range.

The GA operates through an encoding scheme,

where each chromosome represents a potential

solution, and its genes correspond to problem-specific

parameters. Given that the raw data in this study

primarily consist of real numbers, real-number

encoding is employed. Tournament selection is utilized

to randomly select two or more chromosomes from the

previous generation, comparing their fitness values

and incorporating the fittest chromosome into the next

generation.

Considering the constraints of radio transmission

distance, each crossover operation must ensure that the

resulting chromosome remains within the

communication range. To enhance computational

efficiency, single-point crossover is adopted. In this

method, a random crossover point is selected in two

chromosomes of equal length, and genes beyond that

point are exchanged. Additionally, to prevent

premature convergence to local optima, mutation is

introduced. This enables the exploration of a broader

solution space by generating individuals with distinct

characteristics. The flowchart of the genetic algorithm

is illustrated in Figure 1.

In real-life operational scenarios, the GA

computations would be performed by shore-based

rescue centers or command stations, rather than

onboard ships at sea. The hyperparameters, including

population size, crossover rate, and mutation rate, are

determined through empirical testing. In practical

applications, these parameters can be dynamically

adjusted based on factors such as ship density and the

urgency level of emergency situations.

Figure 1. Genetic algorithm process flowchart

2.3 Maritime Geographic Information System (MGIS)

This study employs genetic algorithms to determine

the optimal relay transmission path. It uses AIS to

obtain relevant information such as the location of

relay ships. However, if only the latitude and longitude

of relay ships are known and additional data is not

integrated, more time is required to search for and

organize information. In emergency situations, this

delay could result in missing critical rescue

opportunities. Therefore, this study utilizes MGIS for

its robust visualization and data integration

capabilities. It is incorporated into a ship emergency

response path system. This approach enhances

algorithmic efficiency by addressing deficiencies in

visualization and information display.

MGIS is a specialized system designed to collect,

store, analyze, and visualize marine spatial data. Based

on data representation methods, MGIS data can be

categorized into vector data—represented by points,

lines, and polygons to describe geographical features—

and raster data, which partitions marine spatial data

into a fixed grid, making it suitable for representing

continuous numerical information. Due to the unique

characteristics of the marine environment, including

hydrological variations and tidal influences, MGIS

requires the integration of more complex spatial data

and dynamic analysis compared to traditional

Geographic Information Systems (GIS). The system

development process includes digitizing traditional

nautical charts, transforming spatial data, and

constructing attribute databases, as shown in Figure 2.

This study employs ESRI’s ArcGIS 10.0 as the system’s

demonstration platform due to its widespread

application in geographic monitoring and its extensive

tool library for data analysis and visualization.

Figure 2. MGIS development process

3 RESEARCH DESIGN

3.1 Encoding

The first step in the genetic algorithm is encoding the

imported data to facilitate subsequent computations.

Ship data is imported into the genetic algorithm system

and assigned numerical encoding. Since ships can

receive and decrypt AIS messages within a 20-nautical-

mile range, precautions are necessary to prevent

crossover-generated chromosomes from aligning at

grid boundaries. This misalignment could cause some

genetic points to fall outside the AIS communication

range. To address this issue, the study divides the area

surrounding the distressed ship and shore station into

10-nautical-mile intervals. This ensures that scattered

ship positions are assigned to distinct grid cells. Each

ship is then given a spatial position encoding based on

its grid location. This method ensures that data

remains within the AIS communication range during

612

the crossover process while maintaining a consistent

chromosome length. It also prevents chromosomes

from drifting beyond the communication range after

crossover, as shown in Figure 3.

Figure 3. A chromosome with a Length of 7

3.2 Ship Escape Time Index (SETI)

SETI represents the duration a chromosome remains

within the AIS communication range (20 nautical

miles). Specifically, the chromosome maintains

communication for the time specified by SETI. This

ensures uninterrupted signal transmission and enables

wireless message exchange. This study employs SETI

as the fitness value in the genetic algorithm. The

communication loss time function f(t) and the

communication range are used to determine the time t

at which adjacent genes in a chromosome lose

communication. The ship's position coordinates after t

seconds provide the basis for constructing the

communication loss time function, as expressed in

equation (1).

( )

( ) ( )

1 1 1 2 2 2

sin sin

cos cos

x s t x s t

y s t y s t





+   − +   +





+   − +  



(1)

where x represents longitude, y represents latitude,



denotes course, and 𝑠 represents speed. The subscripts

1 and 2 respectively indicate two different ships.

The communication range constraint is given by

f(t)=20 nautical miles. This function determines the

time t at which two ships, each following their

respective courses and speeds, reach the AIS

communication boundary of 20 nautical miles.

A chromosome consists of i genes, where the

separation time between adjacent genes is determined

using equation (1). This calculation yields the

separation times t1,t2,t3,…,ti-1. The smallest separation

time among them is selected as the SETI for the

chromosome, as expressed in equation (2).

 

1 2 3 1

S ,ETI Min , , ,

t t t t

−

(2)

In this study, the fitness value of each chromosome

is determined by first computing the separation times

between its genes. The shortest separation time among

them is then selected as the fitness value of that

chromosome. The separation time is calculated based

on the known latitude and longitude coordinates of

two ship positions, along with their respective course

and speed.

3.3 Constraints and algorithm mechanism

A radio communication system is employed in this

study for distress message transmission. To ensure

effective message delivery, specific constraints are

introduced. These constraints also mitigate the risk of

transmission failures caused by radio communication

limitations. The key constraints are as follows.

1. The length of all chromosomes in the gene pool is

uniform.

2. The transmission distance between any two points

must not exceed the AIS communication range of 20

nautical miles.

3. Given the fixed chromosome length, the search for

the next relay ship does not consider movement

away from the final destination. The final

destination is the distressed ship.

4. All relay points must be among the ships detected

via AIS within the defined geographic area.

This formalization enhances the rigor of the

algorithmic approach and ensures that the

optimization process aligns with real-world AIS

communication limitations.

In the genetic algorithm, the fitness value

determines whether a chromosome is selected for the

next-generation gene pool. Chromosomes with higher

fitness values have a greater probability of being

included in the new gene pool. This study employs a

tournament selection method, in which two

chromosomes are randomly selected from the current

generation, and the one with the higher fitness value is

retained in the next-generation gene pool. This

approach ensures that the genetic algorithm not only

performs extensive searches but also progressively

improves the quality of solutions over successive

generations.

A single-point random mutation approach is

employed. Each chromosome undergoes mutation

based on a predefined mutation probability. If a

mutation occurs, a gene is randomly selected for

modification, as illustrated in Figure 4. Since the

encoded genes represent the spatial positions of relay

ships in the transmission route, altering the spatial

encoding during mutation may cause the transmission

path to exceed the radio communication limit of 20

nautical miles, resulting in communication failure. To

address this issue, this study retains the original spatial

encoding while randomly selecting an alternative ship

within the same spatial encoding region as the new

relay ship. This approach ensures that although the

spatial encoding remains unchanged, the actual relay

ship for message transmission is modified.

613

Figure 4. Mutation mechanism diagram

4 EXPERIMENTAL RESULTS

4.1 Research scope and data sources

The analysis focuses on the South China Sea and the

Philippine Sea. It examines the overlapping exclusive

economic zones of Taiwan and the Philippines. The

scope extends to the territorial waters of both Taiwan

and the Philippines. The main subjects of analysis are

ships equipped with AIS that navigate within Taiwan’s

territorial waters, Taiwan’s exclusive economic zone,

the Philippine exclusive economic zone, and the

territorial waters of the Philippines. Some fishing ships

in the territorial waters and exclusive economic zones

between Taiwan and the Philippines lack AIS

transponders. This absence prevents the acquisition of

real ship data for simulation purposes. As a result, the

simulation data relies on assumed values derived from

international sources such as MarineTraffic. Key

assumptions include ship course and speed. Ships

included in the model navigate between Taiwan and

the Philippines within the geographical range of 19°N

to 22°N latitude and 120°E to 122°E longitude. In this

study, the GA used for emergency response routing

follows specific parameter settings. The algorithm

maintains a chromosome population of 20 per

generation. It runs for a total of 1500 iterations to

ensure sufficient optimization. The crossover

probability is set to 1.0, enabling full crossover

operations in each generation. The mutation

probability is defined as 0.01 to introduce genetic

diversity while minimizing disruption. Selection is

performed using the tournament selection method,

which enhances competitive survival among

chromosomes. Mutation is applied using a single-point

random mutation technique, and crossover is executed

through a single-point crossover mechanism.

4.2 Results

The numerical analysis is shown in Table 2.

Table 2. Numerical results

Parameter

Value

Distressed ship location

Lat: 19.5326° N,

Long: 121.3760° E

Shore station location

Lat: 21.7996° N,

Long: 120.7890° E

Initial generation fitness value

0 seconds

Final generation fitness value

56.8468 seconds

Computation time

48.1658 seconds

The initial ship relay transmission route is shown in

Figure 5. After 1500 generations of computation using

the GA, the fitness value of the final-generation

chromosome, known as the SETI, reaches 56.8468

seconds. This represents a significant improvement

compared to the initial-generation chromosome. The

relay transmission process of the final-generation ship

distress message is depicted in Figure 6. This figure

illustrates the optimized communication path and

intermediary nodes after computational processing.

The trend of the best fitness value across generations is

shown in Figure 7. The simulation results indicate that

when the distress signal from the distressed ship is

received, an appropriate parameter configuration with

GA computation determines the SETI. This index

remains within a reasonable and acceptable time

frame. The SETI is derived from the GA solution.

Completing the verification process within this

timeframe ensures that the acknowledgment signal

from the shore station reliably reaches the distressed

ship. This process is illustrated in Figures 8 and 9.

Figure 5. Initial distress message relay route

Figure 6. Final distress message relay route

Figure 7. Best fitness value trend per generation

614

Figure 8. Distress signal transmission

Figure 9. Acknowledgment signal transmission

5 CONCLUSIONS

Based on the concept of relay stations in wireless

communication systems, this study proposes an

innovative intelligent radio emergency notification

system as an alternative mechanism for maritime

distress signals. Traditional emergency positioning

systems primarily rely on satellite systems for distress

message transmission. The proposed approach

integrates AIS with genetic algorithms and MGIS. This

integration allows ships encountering force majeure

incidents in offshore waters to utilize this distress

notification mechanism effectively and securely. The

system operates without concerns of information

leakage or high costs. Moreover, it ensures that the

distress message received by coastal stations is

acknowledged and confirmed.

The system uses AIS as the transmission medium

for distress messages and confirmation signals. The

system constructs a network for distress message relay

by collecting ship positions and integrating GA. It

determines the optimal relay route. Coastal stations

send confirmation messages via the AIS network and

notify nearby ships of the distress situation. This

process ensures that the distressed ship receives the

confirmation message and activates the rescue

mechanism. The distressed ship sends distress

messages and receives acknowledgment signals from

relay stations. This system reduces uncertainty in

distress message transmission. This approach is

particularly suitable for fishing ships as it enhances

data security and improves operational safety in

international fishing grounds.

This study assumes a moderately dense maritime

environment in which nearby ships are equipped with

operational AIS transponders, enabling effective

maritime communication networks. However, in real-

world scenarios, particularly in regions with sparse

ship density or among fleets operating inactive or

receive-only AIS transponders, the effectiveness of the

system could be significantly impacted. Such

constraints may limit the capability of maintaining

stable and continuous communication.

Additionally, due to limited access to real-time

maritime ship data, this study relies on hypothetical

AIS data for simulations. Consequently, the actual

system performance may differ according to genuine

maritime traffic conditions.

The communication network established by this

system serves as a general communication network

under normal conditions. It provides nearshore ship

communication services and extends the wireless radio

communication network. This enhances the

accessibility and effectiveness of maritime

communication, making information transmission

more convenient and efficient.

REFERENCES

[1] Hemmati, A., Stålhane, M., Hvattum, L. M., & Andersson,

H. (2015). An effective heuristic for solving a combined

cargo and inventory routing problem in tramp shipping.

Computers & Operations Research, 64, 274–282.

https://doi.org/10.1016/j.cor.2015.06.011

[2] Ben Farah, M., Ahmed, Y., Mahmoud, H., Shah, S. A., Al-

kadri, M. O., Taramonli, S., Bellekens, X., Abozariba, R.,

Idrissi, M., & Aneiba, A. (2024). A survey on blockchain

technology in the maritime industry: Challenges and

future perspectives. Future Generation Computer

Systems, 157, 618–637.

https://doi.org/10.1016/j.future.2024.03.046

[3] Alqurashi, F. S., Trichili, A., Saeed, N., Ooi, B. S., &

Alouini, M.-S. (2023). Maritime communications: A

survey on enabling technologies, opportunities, and

challenges. IEEE Internet of Things Journal, 10(4), 3525–

3547. https://doi.org/10.1109/JIOT.2022.3219674

[4] Thombre, S., Zhao, Z., Ramm-Schmidt, H., Vallet García,

J. M., Malkamäki, T., Nikolskiy, S., Hammarberg, T.,

Nuortie, H., Bhuiyan, M. Z. H., Särkkä, S., & Lehtola, V.

V. (2022). Sensors and AI techniques for situational

awareness in autonomous ships: A review. IEEE

Transactions on Intelligent Transportation Systems, 23(1),

64–83. https://doi.org/10.1109/TITS.2020.3023957

[5] Li, H., Çelik, C., Bashir, M., Zou, L., & Yang, Z. (2024).

Incorporation of a global perspective into data-driven

analysis of maritime collision accident risk. Reliability

Engineering & System Safety, 249, 110187.

https://doi.org/10.1016/j.ress.2024.110187

[6] Singh, S. K., & Heymann, F. (2020, April 20–23). Machine

learning–assisted anomaly detection in maritime

navigation using AIS data. In Proceedings of the

IEEE/ION Position, Location and Navigation

Symposium. IEEE.

https://doi.org/10.1109/PLANS46316.2020.9109806

[7] Kontopoulos, I., Varlamis, I., & Tserpes, K. (2021). A

distributed framework for extracting maritime traffic

patterns. International Journal of Geographical

Information Science, 35(4), 767–792.

https://doi.org/10.1080/13658816.2020.1792914

615

[8] Wei, T., Feng, W., Chen, Y., Wang, C.-X., Ge, N., & Lu, J.

(2021). Hybrid satellite-terrestrial communication

networks for the maritime Internet of Things: Key

technologies, opportunities, and challenges. IEEE

Internet of Things Journal, 8(11), 8910–8934.

https://doi.org/10.1109/JIOT.2021.3056091

[9] Soldi, G., Gaglione, D., Forti, N., Di Simone, A., Daffinà,

F. C., Bottini, G. (2021). Space-based global maritime

surveillance. Part I: Satellite technologies. IEEE

Aerospace and Electronic Systems Magazine, 36(9), 8–28.

https://doi.org/10.1109/MAES.2021.3070862

[10] Wang, H., Liu, Z., Liu, Z., Wang, X., & Wang, J. (2022).

GIS-based analysis on the spatial patterns of global

maritime accidents. Ocean Engineering, 245, 110569.

https://doi.org/10.1016/j.oceaneng.2022.110569

[11] Riveiro, M., Pallotta, G., & Vespe, M. (2018). Maritime

anomaly detection: A review. WIREs Data Mining and

Knowledge Discovery, 8(5), e1266.

https://doi.org/10.1002/widm.1266

[12] Han, Y., & Chu, L. (2025). A systematic review and

bibliometric analysis for maritime emergency

management. Journal of Sea Research, 205, 102585.

https://doi.org/10.1016/j.seares.2025.102585

[13] Ma, Q., Zhang, D., Wan, C., Zhang, J., & Lyu, N. (2022).

Multi-objective emergency resources allocation

optimization for maritime search and rescue considering

accident black-spots. Ocean Engineering, 261, 112178.

https://doi.org/10.1016/j.oceaneng.2022.112178

[14] Ribeiro, C. V., Paes, A., & de Oliveira, D. (2023). AIS-

based maritime anomaly traffic detection: A review.

Expert Systems with Applications, 231, 120561.

https://doi.org/10.1016/j.eswa.2023.120561

[15] Karahalios, H. (2018). The severity of shipboard

communication failures in maritime emergencies: A risk

management approach. International Journal of Disaster

Risk Reduction, 28, 1–9.

https://doi.org/10.1016/j.ijdrr.2018.02.015

[16] Xing, B., Zhang, L., Liu, Z., Sheng, H., Bi, F., & Xu, J.

(2023). The Study of Fishing Vessel Behavior

Identification Based on AIS Data: A Case Study of the East

China Sea. Journal of Marine Science and Engineering,

11(5), 1093. https://doi.org/10.3390/jmse11051093

[17] M.-C. Tsou, S. L. Kao, and C.-M. Su, "Decision support

from genetic algorithms for ship collision avoidance route

planning and alerts," J. Navig., vol. 63, no. 1, pp. 167-182,

2010. doi: 10.1017/S037346330999021X.

[18] J. H. Holland, Adaptation in Natural and Artificial

Systems. University of Michigan Press, 1975.

[19] D. E. Goldberg, Genetic Algorithms in Search,

Optimization, and Machine Learning. Addison-Wesley,

1989.

[20] M. Mitchell, An Introduction to Genetic Algorithms. MIT

Press, 1998.

[21] K. Deb, Multi-Objective Optimization Using

Evolutionary Algorithms. John Wiley & Sons, 2001.

[22] International Maritime Organization (IMO), "Adoption

of amendments to the International Convention for the

Safety of Life at Sea (SOLAS)," MSC 99(73),

2000.https://wwwcdn.imo.org/localresources/en/Knowle

dgeCentre/IndexofIMOResolutions/MSCResolutions/MS

C.99%2873%29.pdf.