臺灣博碩士論文加值系統

近年來全球貿易持續成長，海洋運輸需求增加，雖然航海助航設備發展迅速，但是船舶大型化、船舶數量和航行速度的增加，導致船舶在繁忙的水域航行時碰撞和擱淺仍時常發生。其中大多數碰撞情況被歸類於人為的錯誤決策。傳統的海上運輸由航行員主導與操作，海事案件報告數據顯示；75-96％的海上事故和傷亡肇因於人為疏失，其中56％的海事碰撞案件為違反國際海事組織(IMO)發布的國際海上避碰規則(International Regulations for Preventing Collisions at Sea; COLREGs)或區域航行法規。人為疏失造成的海難事故都伴隨著人員傷亡和環境災難，因此為了提高航行安全，開發智慧航海 (Intelligent navigation)系統以輔助航行員執行避碰決策與操作是目前的重要關鍵。
船舶在狹窄繁忙水域或港區航行時，多船會遇(multi-ship encounter)避碰演算及安全路徑規劃(path planning)是發展智慧航海系統的關鍵技術。由於船舶操縱性、海氣象環境與地理環境的限制，使得其發展成為較複雜且困難的課題。由於資訊產業成長帶動相關技術的快速發展，例如大數據(Big Data)、人工智慧(Artificial Intelligent)機器/深度學習(Machine/Deep learning)、物聯網(Internet of Things)、虛擬/擴充實境(Virtual/Augmented Reality)等技術，使得智慧航海系統成為可行的發展趨勢。智慧船舶避碰系統與汽車避碰系統不同，不僅是要獲得安全的目標路徑，且必須符合國際海上避碰規則和地區性航行規則。
發展智慧船舶避碰系統除了必須符合國際海上避碰規則以外，還要加入實務上的航行經驗。因此對避碰規則和航海經驗進一步做定量分析是賦予避碰系統“智慧”的基本條件。但國際海上避碰規則沒有提供具體的避碰措施與行動原則，如何在符合國際海上避碰規則的原則下自動做出避碰決策並完成避碰操作程序便需要進一步的研究，特別是在多船會遇情形時的避碰決策。
船舶的避碰系統運作時必須具備高可靠性(reliability)、容錯性(fault tolerance)和安全性(safety)，同時必須對船舶周圍環境的即時偵測，以避免與其他船舶或障礙物碰撞及擱淺。本研究以人工勢場法(artificial potential fields method)為基礎發展適用於複雜水域與多船會遇情況下的分散式避碰演算法(decentralized collsion avoidance algorithm)，結合船舶操縱與國際海上避碰規則開發智慧船舶避碰系統。並進行船舶會遇與避碰決策問題之定量分析，探討在多船會遇情況下，每艘船都可能同時扮演避讓船(give-way ship)以及直航船(stand-on ship)兩種角色，因此個別船舶如何自主決定應該在何時採用哪種俥與舵的操作組合來順利完成避碰。
論文的主要內容包括：(1)船舶會遇情況識別，(2)智慧船舶避碰演算法，(3)船舶操縱與航行模擬系統，(4)碰撞風險評估模式，(5)船舶避碰案例分析。分述如下：
船舶會遇情況識別：
根據國際海上避碰規則，將在水域航行時相遇的船舶分為避讓船和直航船兩種，並且提到直航船是應保持其速度和航向盡快通過該區域的船。同時為了方便直航船能安全通行，避讓船應改變其速度和航向以淨空該區域。避碰規則中提到船舶會遇的情況有三種：對遇(head-on)、交會(crossing)及追越(overtake)會遇情況。避碰規則第14條中解釋船舶對遇的情況為：當兩船在相反航向或近似相反航向的路線上相遇導致可能發生碰撞危險時，每艘船都應改變其向右舷的路線，以便都能朝他船的左舷通過。規則第15條中對交會情況的說明是：船舶交叉相遇而有碰撞危險時，見直航船在其右舷者，避讓船應避讓且避免橫越直航船之船艏，如果情況許可，要求避讓船保持在直航船的左舷，避免從直航船前穿過。規則第13及17條對船舶追越的情況說明為：任何船舶追越其他船舶時，應避讓被追越之船舶，凡船舶自直航船正橫(abeam)之後22.5度以上之方位駛近時，應視為追越船。本研究依據前述船舶避碰規則建立船舶會遇情況識別系統。
由本研究建立三維船舶操縱之運動模擬系統可以獲得本船與目標船的實時動態模擬結果，利用兩船距離、兩船方位(舷角)、兩船的速度作為導航資訊，計算最近會遇距離(distance to closet point of approach; DCPA)與到達最近會遇點時間(Time to the Closest Point of Approach; TCPA)等的實時數據，定量結合國際海上避碰規則建立船舶會遇情況識別系統。當船舶會遇情況發生時，判別本船的身分是否為避讓船或直航船，並確定本船與目標船的會遇情況。
智慧船舶避碰演算法：
人工勢場法(artificial potential fields)是由Khatib 1986年提出的，其構想是藉由障礙物的斥力場和目標位置的引力場共同作用形成一個虛擬的人工勢場，再搜索勢函數下降的方向，尋找一條無碰撞的最優路徑，早期主要應用於機器人運動導引，屬於反應式機器人避障算法。人工勢場因為在計算中不涉及隨機變量可產生確定性解(deterministic solution)，輸出完全可以預測。具有即時處理能力的分散式(decentralize)結構對於局部環境狀況的即時變化和符合國際海上避碰規則的特徵，是發展智慧船舶避碰系統的重要關鍵。
本研究基於人工勢場的確定性結構，導入流體動力學理想流理論(ideal flow theory)的速度勢場(velocity potential field)發展船舶避碰演算法。本研究發展的演算法是採用渦流速度勢(vortex potential)建構的變換航向(course-changing)模式和偶極子速度勢(dipole potential)的保持航跡(track-keeping)模式所組成。變換航向模式於船舶航行水域設置虛擬的渦旋速度場；根據國際海上避碰規則引導避讓船避開障礙物或目標船。保持航跡模式則產生虛擬的偶極流速度場，使避讓船在執行避碰操作後能回到原先設定的航線以及維持直航船沿著原始設定航線航行。依據本船與目標船的航行狀態，兩船距離、兩船方位(舷角)、最近會遇距離與到達最近會遇時間等數據，建立兩種操縱模式作用與變換準則，符合航海實務經驗。
船舶操縱與航行模擬系統：
假設船在水平面上作為剛體運動，忽略俯仰(pitch)，起伏(heave)和橫搖(roll)，因此數學模型被簡化為縱移(surge)，橫移(sway)和偏航(yaw)的三個自由度。本研究採用日本造船協會數學模式群組(Mathematical Modeling Group; MMG)發展的船舶運動數學模式，運用組件構成原理將船舶水動力與力矩分離成船體，螺槳與舵三部分個別建構，計算出慣性參數與阻尼參數再考慮互相干擾量進行修正，獲得作用於船舶的總力與總力矩。再以Runge-Kutta法計算每一時間步船舶運動參數。
本研究建立船速的模糊邏輯控制(fuzzy logic control)模組，以船舶操縱動態模擬獲得的本船速度與速度變化率做為控制模組的輸入，船速與速度變化率的歸屬函數轉換為模糊值並且由模糊推論得到轉速的模糊值，再由對應的轉速歸屬函數解模糊化後得到控制主機轉速的命令。同時以比例-積分-微分(proportional-integral-derivative; PID)控制演算法建立航向控制器，以動態模擬獲得的本船航向偏差量、累積偏差量與偏差量變化率計算舵角的大小。結合MMG船舶運動數學模式、船速模糊控制模式與PID控制演算法，建立執行多船會遇避碰操作模擬的平台。
碰撞風險評估模式：
本研究利用兩種方法，即模糊推論(fuzzy inference)及模糊綜合評估法(fuzzy comprehensive evaluation method)建立碰撞風險評估模式。其中模糊推論法以動態模擬獲得之舷角變化率、最近會遇距離與到達最近會遇點時間等參數，透過模糊化(fuzzification)的過程，以模糊IF-THEN法則推論碰撞風險模糊值，再經過解模糊化(defuzzification)程序獲得風險值。模糊評估法以動態模擬獲得之兩船距離、兩船方位(舷角)、最近會遇距離與到達最近會遇點時間等數據，由對應的歸屬函數模糊化後經過加權運算獲得風險值。本研究建立之碰撞風險評估模式用於船舶會遇情況時啟動與結束船舶避碰操作程序，以及多船會遇時評估各目標船的相對風險值，採取適當的避讓優先順序。
案例測試結果顯示：基於舷角變化率、最近會遇距離與到達最近會遇點時間等參數建立之模糊推論法無論在對遇、交會與追越等會遇情況均能夠充分表現出船舶會遇與避碰操作過程的風險指標。
船舶避碰案例分析：
為了驗證與確認本文發展的分散式避碰演算模式與智慧船舶避碰系統，設計多種船舶會遇情況，如固定障礙物案例、兩船會遇案例(包括對遇、交會與追越)以及多船會遇案例(四船對稱、序列與交錯)情況。固定障礙物案例測試結果顯示；本船具備自主導航避開障礙能力，並測試啟動距離與變換航向-保持航跡操作模式轉換參數對避讓過程影響的敏感性，顯示避碰演算法具備相當的穩定性。兩船會遇案例測試結果顯示所有避碰操作均符合國際海上避碰規則，且變換航向-保持航跡操作模式依據本船與目標船狀態自主轉換。同時也驗證本研究所建立的碰撞風險評估模式的適用性。四船對稱案例測試結果顯示避碰演算法的對稱性與穩定性。四船序列案例則測試本船同時避讓多船，結果顯示避碰演算法可以穩定執行避碰操作。四船交錯案例測試結果顯示：本船同時對目標船避讓船也對其他目標船直航，且對遇、交會與追越會遇情況同時發生的現象。由兩船與多船會遇情況的模擬結果顯示，本研究結合船舶會遇情況識別、智慧船舶避碰演算法、船舶操縱模擬系統與碰撞風險評估等模式所提出之智慧船舶避碰系統能有效並安全地執行避讓決策與操作。各案例本船均以預先設定的安全距離通過障礙物或是會遇的船舶。
結語：
本研究發展多船會遇情況下自動避碰和路徑規劃演算法的設計與初步應用。通過案例模擬與分析，速度勢場模型與航海實務操作以及船舶操縱概念一致，適合多船會遇情況下的避碰演算，具有符合國際海上避碰規則、考量即時環境變化，以及可安裝於各船舶上的分散式系統等優良特性。同時本研究建立之碰撞風險評估模式用於多船會遇時評估各目標船的相對風險值，採取適當的避讓優先順序，提供為智慧避碰系統的學習知識，以提升船舶航行、避碰決策與操作的效率和安全性。本文相關的研究成果更可以提供實務工作者的參考。

In recent years, global trade has continued to grow, and the demand for marine transportation has increased. Although the navigation aids have developed rapidly, the increase in the size of ships, the number of ships and the speed of navigation have caused collisions and grounding in busy waters area frequently. Most of these collisions are classified as human error. Traditional maritime transport is dominated and operated by pilots. Maritime data shows that 75-96\% of marine accidents and casualties are classified as human error, of which 56\% of collision cases are in violation of the International Regulations for Preventing Collisions at Sea (COLREGs) by International Maritime Organization (IMO) or regional navigation regulations. Casualties and environmental disasters accompany shipwrecks accident caused by human error. Therefore, in order to improve navigation safety, it is important to develop an intelligent navigation system to assist the pilots in performing decision-making and operation of collision avoidance .

When ships are sailing in narrow and busy waters or port areas, multi-ship collision avoidance and safe path planning are key technologies for developing intelligent navigation systems. Due to the limitations of ship maneuverability, sea meteorological environment and geographical environment, Developing the intelligent navigation systems has become a more complicated and difficult subject. However, the growth of Information Technology (IT) has led to the rapid development of Intelligent Systems, such as Big Data, machine/deep learning, Artificial Intelligence (AI), Internet of Things (IoT), virtual/augmented reality, etc. The technologies make intelligent navigation systems a viable development trend. The intelligent ship collision avoidance system is different from the car collision avoidance system in that it is not only to obtain a safe target path but also to comply with COLREGs and local navigation rules.

In addition to the international collision avoidance rules, the development of intelligent ship collision avoidance systems must also incorporate practical navigation experience. Therefore, further quantitative analysis of collision avoidance rules and navigation experience is “intelligence” to giving collision system "wisdom". However, the international collision avoidance rules do not provide specific collision avoidance measures. How to automatically make collision avoidance decisions under the COLREGs to complete the collision avoidance operation requires further research, especially in the case of multi-ship encounters.

The ship's collision avoidance system must operate with high reliability, fault tolerance and safety for detecting the surrounding environment of the ship in real time to avoid collision and grounding with other ships or obstacles. Based on the artificial potential field method, this study develops a decentralized collision avoidance model suitable for complex waters and multi-ship encounters based on ship maneuvering and COLREGs. Moreover, carry out quantitative analysis of the decision-making problems of multi-ship ship collision avoidance by the COLREGs. And discuss that in the case of multi-ship encounters, each ship gives way to certain target ships while stands on to other target ships. When to use a proper action of the engine orders and rudder orders to carry out collision avoidance of the ship.

The main contents of this thesis include: (1) ship encounter situation identification, (2) intelligent ship collision avoidance algorithm, (3) ship maneuvering and navigation simulation system, (4) collision risk assessment model, and (5) cases analysis of ship encounter. The description is as follows:

\noindent
{\bf Identification of ship encounter situation:}

According to the COLREGs, ships encounter during navigation in the waters are classified as give-way ship and stand-on ship. And it is mentioned that the stand-on ship should maintain its speed and course to pass the area as soon as possible. At the same time, the give-way ship should change its speed and course to clear the area. There are three situations mentioned in the collision avoidance rules, that are head-on, crossing, and overtake encounter situations. Rule 14 states head-on situation; when two power-driven vessels are meeting on reciprocal or nearly reciprocal courses so as to involve risk of collision each shall alter her course to starboard so that each shall pass on the port side of the other. Rule 15 states crossing situation; when two power-driven vessels are crossing so as to involve risk of collision, the vessel which has the other on her own starboard side shall keep out of the way and shall if the circumstances of the case admit avoiding, crossing ahead of the other vessel. This situation would frequently arise and it is always better to avoid a close quarter situation and go right around the stern of the other vessel rather than cross ahead of the other vessel. Rules 13 and 17 describe the overtake situation as follows: any vessel overtaking any other shall keep out of the way of the vessel being overtaken. Meaning: It does not make a difference whether any ship ahead of the own vessel has permitted by signalling that overtaking may take place. As long as the give-way vessel takes action well in time there is no problem and the stand-on vessel follows the above Rule, and the stand on the vessel is required not to take action, but it does not mean that she would not be alert and monitor the situation.

From this study, a three-dimensional ship maneuvering simulation system can be used to obtain real-time dynamic results of the ship and the target ship. The distance between the ships, the bearing angle, and the speeds are used to calculate real-time data, such as the distance to the closest point of approach ($DCPA$) and time to the closest point of approach ($TCPA$), and quantitatively combined with the COLREGs to create an identification system of ship encounter situation. When ships encounter, the system determines whether the own ship is a give-way ship or a stand-on ship, and determines the encounter situation between the ship and the target ships.

\noindent
{\bf Intelligent collosion avoidance algorithm:}

The artificial potential fields method was proposed by Khatib in 1986. The concept is to form a virtual artificial potential field by the interaction between the repulsive field of the obstacle and the attractive field of the target position. Also, then search for the direction of the potential function to find a collision-free optimal path, which was mainly applied to robot motion guidance in the early stage, and belongs to the reactive robot obstacle avoidance algorithm. The artificial potential fields can create a deterministic solution because it does not involve random variables in the calculation, and the output is completely predictable. The decentralized structure with immediate processing capability is an important key to the development of intelligent ship collision avoidance systems for the real-time changes of local environmental conditions and comply the COLREGs.

Based on the deterministic structure of the artificial potential fields, this study introduces the velocity potential field of the ideal flow theory to develop the ship collision avoidance algorithm. The algorithm developed in this study consists of a course-changing mode constructed by vortex potential and a track-keeping mode of dipole potential. The course-changing mode sets a virtual vortex velocity field in the ship's navigational waters, and guides the give-way ship to avoid obstacles or target ships according to the COLREGs rules.

Track-keeping mode produces a virtual velocity field of dipole flow that allows the give-way ship to return to the originally set route after performing the collision avoidance operation and to keep the stand-on ship sailing along the original set route. According to the navigation status of the ship and the target ships, the distance between the two ships, the position two ships (the bearing angle), the $DCPA$ and the $TCPA$, the two modes of operation and the transformation criteria are established, which is in line with the experience of navigation practice.

\noindent
{\bf Ship Maneuvering and Navigation Simulation System:}

Assuming that the ship acts as a rigid body in the horizontal plane, ignoring the pitch, heave and roll motions, the mathematical model is reduced to the sway, sway and yaw of three degrees of freedom. In this study, using the mathematical model of ship motion developed by Maneuvering Mathematical Modeling Group (MMG) in Japan to separate the ship's hydrodynamic force and moment into the parts of hulls, propeller and rudder and to be constructed separately. The inertia and the damping effects are computed and corrected by considering the mutual interference among hull, propeller and rudder to obtain the total forces and moments acting on the ship. Then the Runge-Kutta method is used to calculate the ship motion parameters for each time step.

In this study, the fuzzy control module of the ship speed is established. The speed and speed change rate of the ship obtained by the ship dynamic simulation is used as the input of the control module. The membership functions of the ship speed and speed change rate are converted into the fuzzy value. The fuzzy value of the rotating speed of propeller is obtained by fuzzy inference, and then the corresponding rotating speed membership function is used to defuzzify to obtain the command to control the rotating speed of the main engine. At the same time, the proportional-integral-derivative (PID) control algorithm is used to establish the course controller. The rudder angle is calculated by the dynamic course deviation, cumulative deviation and deviation rate of the ship. Combined with MMG ship motion mathematics mode, ship speed fuzzy control mode and PID control algorithm, a platform for implementing multi-ship encounter collision avoidance operation simulation is established.

\noindent
{\bf Collision risk assessment model:}

This study uses two methods, fuzzy inference and fuzzy comprehensive evaluation methods, to establish a collision risk assessment model. Among them, the fuzzy inference method uses the fuzzy IF-THEN rule to infer the fuzzy value of the collision risk through the parameters such as the rate of change of the bearing angle, the $DCPA$ and the $TCPA$ obtained by the dynamic simulation. Through a defuzzification procedure, the system evaluated the collision risk index.
The fuzzy evaluation method obtains the risk index by input the distance between the two ships, the ship position (bearing angle), the $DCPA$ and the $TCPA$. The corresponding membership function fuzzifies the input parameters and then a weighted evaluation or fuzzy comprehensive evaluation were used to obtain the collision risk index. The collision risk assessment model established in this study is used as a criterial to switch on or off the ship collision avoidance operation procedure when the ship in encounter situations, and also used to assess the relative risk index of target ships when multi-ships encounter to adopt appropriate avoidance priority order.

The results show that the fuzzy inference method based on the parameters such as the rate of change of the bearing angle, $DCPA$ and $TCPA$ can fully demonstrate the risk characteristics including head-on, crossing and overtake encounter of ships.

\noindent
{\bf Cases analysis of ship encounter: }

In order to verify and confirm the decentralized collision avoidance algorithm and the intelligent ship collision avoidance system developed in this paper, we design a variety of ship encounter situations, such as cases of fixed single/multiple obstacles, cases of two ship encounter situation(including head-on, crossing and overtake) and cases of multi-ships encounter situation(four ships symmetry, sequence and staggered) will be studied.
The results of the static obstacle test show that the ship can avoid obstacles by autonomous navigation. Moreover, the sensitivity test of the starting distance and the criterial of switching the course-changing and track-keeping mode was performed, and the results show stability performance of the collision avoidance algorithm. The results of the cases of the two ships encounter show that all the collision avoidance operations are in line with the COLREGs, and the course-changing and track-keeping mode are acting exactly according to the situations of the ship and the target ship. It also verifies the applicability of the collision risk assessment model established by the present study. The results of the four-ship symmetry case show the symmetry and stability of the collision avoidance algorithm. In the case of the four-ship sequence, the ship was tested to avoid multiple ships at the same time. The result shows that the collision avoidance algorithm can stably perform the collision avoidance operation. The results of the four-ship staggered case show that the ship is synchronous execution give-way and stand-on operations to target ships, and the head-on, crossing and overtake encounters situations simultaneously occured.

The simulation results of the cases of two ships and multi-ships show that the intelligent ship collision avoidance system combined with the ship encounter situation identification, intelligent ship collision avoidance algorithm, ship maneuver simulation system and collision risk assessment can be effectively and safely perform avoidance decisions and operations. In each case, the ship passes the obstacle or the encounter ship at a preset safe distance.

\noindent
{\bf Epilogue:}

This study addresses the design and preliminary application of intelligent collision avoidance and path planning algorithms for multi-ship encounters, that must find a safe route and decide on the control actions required to ensure compliance with COLREGs rules and minimize hazard to an acceptable level.
Through the cases studies, the velocity potential fields model is consistent with the navigation practice and the concept of ship maneuvering. It is suitable for collision avoidance evaluation under multi-ship encounter conditions and for real-time environmental variations in accordance with the COLREGs rules. The system is decentralized and can be installed to each ship and onshore office of watch. At the same time, the collision risk assessment model is used to evaluate the relative risk value of each target ship when multi-ships encounter, and adopt appropriate action priority order to provide learning knowledge for intelligent collision avoidance system to improve the widom of the system. The relevant research results in this paper also can be provided as a reference for practical workers.

Abstract ix
Contents xvii
List of Figures xix
List of Tables xxv
chapter 1 Introduction 1
1.1 Motivation 1
1.2 Literature review 3
1.3 Objectives 7
1.4 Contributions 9
chapter 2 COLREGs Rules and Regulation 11
2.1 COLREGs Rules and Regulation 11
2.2 Encounter Situation and Collision Avoidance 17
chapter 3 The Potential Field Method 21
3.1 The Artificial Potential Field Method 21
3.2 The Collision Avoidance Algorithm 28
3.2.1 Conceptual Basis 28
3.2.2 The Velocity Potential for Path Planning 30
3.2.3 The Track-keeping mode 31
3.2.4 The Course-changing mode 33
chapter 4 Collision Risk Evaluation 35
4.1 Fuzzy inference method 36
4.1.1 Input parameters 36
4.1.2 Fuzzy inference system 37
4.2 Fuzzy Comprehensive Evaluation Method 47
4.2.1 Lin-Yao’s Membership Function 48
4.2.2 Zheng’s Membership Function 49
4.2.3 Zheng-Wu’s Membership Function 51
chapter 5 The Ship Dynamical Simulation 53
5.1 The Dynamical Model for Ship Navigation 53
5.1.1 Equations of ship motion 53
5.1.2 Mass and Inertial Coefficients 56
5.1.3 Force and Moment on Hull 56
5.1.4 Thrust by Propeller 58
5.1.5 Force and Moment by Rudder 58
5.1.6 Formulation of The Problem 60
5.2 Fuzzy Controller for Propulsion System 61
5.3 The PID controller for Steering System 64
5.4 Dynamical Simulations 68
5.4.1 Test Ship 68
5.4.2 Turning Circle Test 72
5.4.3 20/20 degrees Zig-zag Test 73
chapter 6 Simulations of Two-Ship Encounter 79
6.1 Virtual vessel 80
6.2 Static Obstacle I 81
6.3 Static Obstacle II 90
6.4 Head-on encounter situation 91
6.5 Overtake Encounter Situation 97
6.6 Crossing Encounter Situation 103
chapter 7 Simulations of Multi-Ship Encounter 109
7.1 Four-ship Symmetry Encounter 109
7.2 Four-ship Sequence Encounter 116
7.3 Four-ship Staggered Encounter 121
chapter 8 Conclusion Remarks 133
Bibliography 137

[1] Ahn J. H., Rhee K. P. and You Y. J., A study on the collision avoidance of a ship using neural networks and fuzzy logic, Apply Ocean Research, 37, 2012, 162-173.
[2] Antão, P., Guedes Soares, C., Causal factors in accidents of high speed craft and conventional ocean going vessels, Reliability Engineering and System Safety 2008, 93, 1292–1304.
[3] Benjamin, M.R., Leonard, J.J., Curcio, J.A., Newman, P.M., A method for protocol-based collision avoidance between autonomous marine surface craft. J. Field Robot, 2006a, 23 (5), 333–346.
[4] Benjamin, M.R., Curcio, J.A., Leonard, J.J., & Newman, P.M., Navigation of unmanned marine vehicles in accordance with the rules of the road. In: Proceedings of the IEEE International Conference on Robotics and Automation,
2006b, pp. 3581–3587.
[5] Belenky V., Falzarano J., Rating-based maneuverability standards, ABS TECHNICAL PAPERS 2006, Originally presented at the SNAME 2006 Annual Meeting Conference held in Ft. Lauderdale, Florida, October 10-13, 2006, 227-246.
[6] Bukhari A., Tusseyeva I., Lee B. and Kim Y., An intelligent real-time multivessel collision risk assessment system from VTS view point based on fuzzy inference system, Expert System with Applications, 40(4), 2013, 1220-1230.
[7] Burgos E. and Bhandari S., Potential flow field navigation with virtual force field for UAS collision avoidance, 2016 International Conference on Unmanned
Aircraft Systems (ICUAS), June 7-10, 2016, Arlington, VA USA
[8] Conventions on the International Regulations for Preventing Collision at Sea (COLREGs). 1972. The International Maritime Organization(IMO).
[9] Fossen, T. I., Guidance and Control of Ocean Vehicles, John Wiley & Sons Ltd., 1994.
[10] Fujii J. and Tanaka K., Traffic capacity, The Journal of Navigation, 24(4), 1971, 543-552.
[11] Goodwin E. M., A statistical study of ship domains, The Journal of Navigation, 28(3), 1975, 328-334.
[12] Hasegawa K., Kouzuki A., Muramatsu T., Komine H., and Watabe Y., Ship autonavigation fuzzy expert system (SAFES), Journal of the Society of Naval Architecture of Japan, 1989(166), 1989, 445-452.
[13] He, Yixiong, Jin, Yi, Huang, Liwen, Xiong, Yong, Chen, Pengfei, Mou, Junmin, Quantitative analysis of COLREG rules and seamanship for autonomous collision avoidance at open sea. Ocean Engineering, 2017, 140, 281-291.
[14] Hu X. P., Li Z. Y. and Cao J., A Path Planning Method Based on Artificial Potential Field Improved by Potential Flow Theory, 2017 2nd International Conference on Computer Science and Technology (CST 2017), 617-625, ISBN:
978-1-60595-461-52016
[15] Hwang C. N., The Integrated Design of Fuzzy Collision-Avoidance and H1-Autopilots on Ships, Journal of Navigation, 55(1), 2002, 117-136.
[16] Ito, M., Zhang F. F., and Yoshida, N. Collision avoidance control of ship with genetic algorithms, Proceedings of the 1999 IEEE International Conference on
Control Applications, Kohala Coast, Hawaii, USA, August 1999, Vol. 2, pp.1791-1796.
[17] Khatib O., Real-time obstacle avoidance for manipulators and mobile robots. Robotics Res., 1986, 5(1), 90-98.
[18] Kijima K., Prediction method for ship manoeuvring performance in deep and shallow waters, Presented at the Workshop on Modular Manoeuvring Models, The Society of Naval Architects and Marine Engineering, (Nov. 13, 1991).
[19] Kim, D., Hirayama, K. and Park, G., Collision Avoidance in Multiple-Ship Situations by Distributed Local Search. Journal of Advanced Computational Intelligence and Intelligent Informatics, 2014, 18(5), 839–848.
[20] Kim, D., Hirayama, K. and Okimoto, T., Ship Collision Avoidance by Distributed Tabu Search. The International Journal on Marine Navigation and Safety of Sea Transportation, 2015, 9(1), 23–29.
[21] Kim, D., Hirayama, K., and Okimoto, T., Distributed Stochastic Search Algorithm for Multi-ship Encounter Situations. THE JOURNAL OF NAVIGATION, 2017, 70, 699–718.
[22] Kovács, B., Szayer, G., et al., A novel potential field method for path planning of mobile robots by adapting animal motion attributes, Robotics & Autonomous
Systems, 82(C), 2016, 24-34.
[23] Lee H. J. and Rhee K. P., Development of collision avoidance system by using expert system and search algorithm, International Shipbuilding Progress, 48(3),
2001, 197-212.
[24] Lee S. M., KWON K. Y. and Joh J., A fuzzy autonomous navigation of marine vehicles satisfying COLREGS guidelines, Control Autom., 2004, 2(2), 171-181.
[25] Liu Y. and Sun H. (1998). A Risk-Degree Model of Collis ion Based on Fuzzy Theory. Navigation of China. 43(2):23-29.
[26] Macktoobian, M., Shoorehdeli, M. A., Time-variant artificial potential field (TAPF): a breakthrough in power-optimized motion planning of autonomous space mobile robots, Robotica, 2016, 35(5), 1-23.
[27] Mamdani, E. H. and Assilian, S. A. (1975). An experiment in linguistic synthesis with a fuzzy logic controller. Intern. J. Man - Mach. Studies, 7:1-13, 1975.
[28] Mamdani, E. H. and Assilian, S. A. (1999). Experiment in linguistic synthesis with a fuzzy logic controller. International Journal of Human Computer Studies,
2(51), 135–147.
[29] Naeem, W., Henrique, S. C. and Hu, L. (2016). A Reactive COLREGs-Compliant Navigation Strategy for Autonomous Maritime Navigation. IFACPapersOnLine,
49(23), 207 -213.
[30] Noto N., Okuda H., Tazaki Y., Suzuki T., Steering Assisting System for Obstacle Avoidance Based on Personalized Potential Field, IEEE International
Conference on Intelligent Transportation Systems, 2012, 1702-1707.
[31] Perera, L. P., Carvalho, J. P., Guedes Soares, C., Fuzzy-logic based decision making system for collision avoidance of ocean navigation under critical collision
conditions. J. Mar. Sci. Technol., 2011, 16(1), 84–99.
[32] Perera, L. P., Carvalho, J. P., Guedes Soares, C., Intelligent ocean navigation and fuzzy-Bayesian decision/action formulation. IEEE J. Ocean. Eng. 37(2),
04–219.
[33] Perera, L. P., Carvalho, J. P., Guedes Soares, C., Solutions to the failure and limitations of mamdani fuzzy inference in ship navigation. IEEE Trans. Veh.
Technol., 2014, 63(4), 1539–1554.
[34] Pietrzykowski Z., Ship’s Fuzzy Domain – a Criterion for Navigational Safety in Narrow Fairways, Journal of Navigation, 61(3), 2008, 499-514.
[35] Roberts G.N., Trends in marine control systems, Annual Reviews in Control 32 (2008) 263-269
[36] Rong, H. , Teixeira, A., Soares, C. G., (2015) Evaluation of near-collisions in the Tagus River Estuary using a marine traffic simulation model, Scientific
Journals of the Maritime University of Szczecin, 43 (115), 68-78
[37] Rothblum, A.M.,Wheal, D.,Withington, S., Shappell, S. A.,Wiegmann, D. A., Boehm,W., Chaderjian, M., Key to successful incident inquiry. Proceedings 2nd International Workshop on Human Factors in Offshore Operations (HFW),
Houston, Texas, 2002, 1–6.
[38] Saboya Jr. F., Alves, M. G. and Pinto, W. D. (2006) Assessment of failure susceptibility of soil slopes using fuzzy logic. Eng. Geol., 86:211-224.
[39] Shibata N., Sugiyama S. & Wada T., Collision avoidance control with steering using velocity potential field, 2014 IEEE Intelligent Vehicles Symposium (IV) June 8-11, 2014, 438-443, Dearborn, Michigan, USA
[40] Smierzchalski R., Evolutionary trajectory planning of ships in navigation traffic areas. J. Mar. Sci. Technol., 1999, 4(1), 1–6.
[41] Smierzchalski R. and Michalewicz Z., Modeling of ship trajectory in collision situations by an evolutionary algorithm, IEEE Transactions on Evolutionary Computation, 2000, 4(3), 227-241.
[42] Statheros, T., Howells, G., McDonald-Maier, K., Autonomous ship collision avoidance navigation concepts, technologies and techniques, The Journal of Navigation, 2008, 61, 129-142.
[43] Szlapczynski R., A Unified Measure Of Collision Risk Derived From The Concept Of A Ship Domain, Journal of Navigation, 59(3), 2006, 477-490.
[44] Szlapczynski, R., Evolutionary Sets of Safe Ship Trajectories: A New Approach to Collision Avoidance. The Journal of Navigation, 2011, 64(1), 169–181.
[45] RAFAŁ SZŁAPCZYN’SKI, JOANNA SZŁAPCZYN’SKA, Customized
crossover in evolutionary sets of safe ship trajectories. Int. J. Appl. Math. Comput. Sci., 2012, 22(4), 999–1009.
[46] Szlapczynski, R., Evolutionary Planning of Safe Ship Tracks in Restricted Visibility. The Journal of Navigation, 2015, 68(1), 39–51.
[47] Tam, C., Bucknall, R., Path-planning algorithm for ships in close-range encounters. J. Mar. Sci. Technol., 2010, 15 (4), 395-407.
[48] Tam, C., Bucknall, R., Cooperative path planning algorithm for marine surface vessels. Ocean Eng., 2013, 57, 25–33.
[49] Tsou, M. C., Hsueh, C. K., The study of ship collision avoidance route planning by antcolony algorithm. J. Mar. Sci. Technol., 2010, 18(5), 746–756
[50] Tsou, M.C., Kao, S.L., Su, C.M., Decision support from genetic algorithms for ship collision avoidance route planning and alerts. J. Navig., 2010, 63 (01),
167–182.
[51] Wang, T. F. , Yan, X. P. and Wang, Y. (2017). Ship Domain Model for Multiship Collision Avoidance Decision-making with COLREGs Based on Artificial Potential Field. TransNav, The International Journal on Marine Navigation
and Safety of Sea Transportation, 11(1), 85-92.
[52] Wu Z. and Zheng Z. (2001). Time collision risk and its model. Journal of Dalian Maritime University, 27(2):1-5.
[53] Xiao F. L., Ligteringen H., van Gulijk C. and Ale B., Artificial force fields for multi-agent simulations of maritime traffic: a case study of Chinese waterway,
Procedia Engineering 45 ( 2012 ) 807-814.
[54] Xu Q. and Wang N. (2014). A Survey on Ship Collision Risk Evaluation. PROMET-Traffic & Transportation, vol. 26, pp. 475-486.
[55] Xue Y., Lee B.S. and Han D., Automatic collision avoidance of ships, Proceedings of the Institution of Mechanical Engineers, Part M: Journal of Engineering for the Maritime Environment, 223 (1), 2009, 33-46, ISSN 1475-0902
[56] Xue Y., Clelland D., Lee B. S. and Han D.F., Automatic simulation of ship navigation, Ocean Engineering, Volume 38, Issues 17–18, December 2011, Pages
2290- 2305