Collision risk assessment and collision avoidance of vessels have always been an important topic in ocean engineering. Decision support systems are increasingly becoming the focus of many studies in the maritime industry today as vessel accidents are often caused by human error. This study proposes an anticollision decision support system that can determine surrounding obstacles by using the information received from radar systems, obtain the position and speed of obstacles within a certain time period, and suggest possible routes to prevent collisions. In this study we use a neural network to predict the subsequent positions of surrounding vessels, a fuzzy logic system to obtain the risk of collision, and a particle swarm optimization algorithm to find the safe and shortest path for collision avoidance.

<p>Estimation of the possible position of the moving targets after a few steps has great significance especially in terms of defense industry. If a shoot aiming at the target is planned, the issue of estimation of forward position of the target gains importance in terms of accurate strike of the bullet at the target. In target tracking, impact of three different methods as motion estimation method on various motion types has been examined in our study. Motion types have been examined in four different types, which are rectilinear motion, circular motion, sinusoidal motion and curvilinear motion.  On the other hand, estimation methods have been examined under three different titles. These are Kalman estimation method, curve fitting method and Anfis method. Different motion types have been examined with different estimation methods and the results obtained have been presented. </p><p>Keywords: robotics, image processing, trajectory tracking, target tracking, estimation, Anfis, Kalman, curve fitting.</p>

 CSGA for collision avoidance  Fuzzy for collision risk assesment  Neural network for position estimation Collision risk assessment and collision avoidance of vessels have always been an important area of research in the field of ocean engineering. Decision support systems constitute the focus of many studies in the maritime industry as vessel accidents are often caused by human errors. In this study, an anti-collision decision support system is proposed. The proposed system can determine surrounding obstacles by using the information it receives from the AIS and Radar equipments, obtain the position of obstacles within a certain time period, calculate the TCPA(Time to the Closest Point of Approach) and DCPA(Distance at Closest Point of Approach) using a fuzzy system in the light of COLREGs (Collision prevention regulations at sea) and suggest the optimal route to prevent collisions using a hybrid cuckoo search-genetic algorithm. Figure A. General structure of the proposed system Purpose: As collision of ships are because of the human errors. Decision support systems have become very popular for ship collision avoidance. We propose an on board collision risk assesment and collision avoidance system to prevent ships from collisions. Theory and Methods: We used three methods in this study. A neural network to predict next positions of the targets. A fuzzy system to obtain risk assesment and a hybrid method based collision avoidance algorithm. Results: The proposed system has been tried for some scenarios and found to be successful. Also the improvements of the study have been shown both numerically and graphically. Conclusion: A collision risk assesment and collision avoidance system has been proposed. The system has given fast and reliable solution for the examined scenarios. The usage of the CSGA algorithm for collision avoidance is found to be high.

Shortest path algorithms are frequently used in important sectoral areas such as maritime, aeronautical and land transport, and are still highly popular. The different side of our work is that it is on the ports of Greece and Turkey on the Aegean Sea and that the nodes are prepared on the actual map based on the actual coordinates and depths. Our study differs from previous studies in that it focuses on Greek and Turkish ports on the Aegean Sea and the nodes are prepared on an actual map based on actual coordinates and depths. The study is part of an intelligent system capable of actually planning the route of a sailing vessel. It is designed to evaluate the methods that an intelligent system can use. The intelligent system will be a system that can dynamically change and optimize the route depending on changing conditions such as weather, current and navigational information. This is the reason why the node map is obtained using real and detailed information. The Breadth-First, Bellman-Ford and Dijkstra’s algorithms were used to calculate the shortest routes between ports and the results were evaluated in terms of the route, distance and calculation time used.

Öz Genetik algoritma, evrimsel bir algoritma olup, en sık kullanılan problem çözümleme algoritmalarından biridir. En kısa yol bulma problemi ise denizcilik, havacılık, savunma ve yük taşımacılığı gibi önemli alanlarda çokça çalışılan bir konudur. Bu çalışmada Ege denizi kıyısında bulunan limanların arasındaki en kısa yolun bulunmasında genetik algoritma kullanılmıştır. Ege denizi üzerinde bulunan 61 liman ve bu limanlar arasında yapılacak olası seyir için rota belirlemede tanımlanan 604 adet düğüm en kısa yol probleminin genetik algoritma yardımıyla belirlenmesi amacıyla kullanılmıştır. Çalışmayı farklı yapan kısımlar; limanların ve düğümlerin gerçek harita üzerinde ve gerçek koordinatlara göre kayıt altına alınmış olması ve düğüm haritasının büyük ölçekli olmasıdır. Ege denizi üzerinde bulunan 61 adet liman arasında yapılabilecek olası seyir durumunda izlenilebilecek rotalar genetik algoritma yardımıyla hesaplanarak sonuçlar ortaya konulmuştur.