Channel Safety Assessment in Ship Navigation Based on Fuzzy. Logic Model

Transcription

1 Wu et al Channel Safety Assessment in Ship Navigation Based on Fuzzy Logic Model Yuan Wu School of Naval Architecture, Ocean and Civil Engineering Shanghai Jiao Tong University 00 Dongchuan Road, Shanghai, P. R. China Tel: ()0-0, Fax: ()0- wuyuan@sjtu.edu.cn Hao Hu* School of Naval Architecture, Ocean and Civil Engineering and State Key Laboratory of Ocean Engineering Shanghai Jiao Tong University 00 Dongchuan Road, Shanghai, P. R. China TEL: ()0-0, FAX: ()0- hhu@sjtu.edu.cn *corresponding author Submitted to the Transportation Research Board rd Annual Meeting for Presentation and Publication Submission date: November, 0 Word Count: Text (0 words) + Figures (0) +Table (00) = words total 0

2 Wu et al. 0 ABSTRACT: With the progress of economic globalization, international shipping trade is developing fast, resulting in a gradual rise in the amount of transportation in channel. In turn maritime affairs occur more frequently due to that than before, and the circumstance of transportation in channel is getting more complex. It is a great influence to the safety of navigation as well as a problem must be solved about safety assessment. What is more, the development of a port cannot be sustained without the comprehensive assessment of the relative channel. Therefore, this paper tries to make an overall evaluation of channel safety. This paper establishes a multilevel index system and constructs a fuzzy logic model based on fuzzy set to give a complete assessment on the channel safety. Previous researches show that fuzzy logic model can be used to manage multi-factor in safety assessment. However, the application for channel safety is an experiment and innovation. The model is applied to Yangtze Estuary deep water channel, and reaches an conclusion roughly correspond to the practice. It has important value to make integrity planning of the channel. KEY WORDS: Channel, Safety Assessment, Fuzzy Logic, Assessment System

3 Wu et al INTRODUCTION Nowadays, the integrated and comprehensive trend of world ports development, the large-scale and high-speed urgent demand of ship and the increasing volume of cargo all bring more challenges for shipping safety which has a direct connection with the prospect of shipping and its surrounding industries. Meanwhile, it plays an important role in the protection of marine environment and the exploration of ocean resources. Despite the remarkable effort performed at different levels to achieve a safe Maritime Transport System(MTS), the occurrence of accidents and incidents at sea is still increasing(). In the shipping field, with the growing emphasis on the safety management, the methodology of safety analysis and the research and applications of safety assessment gain more and more attention, especially for the safety assessment which in recent years has even been put on the agenda. From the perspective of the safety management, safety assessment is not only of great academic and pragmatic importance, but is related to the future development of shipping industry. As a hot topic in the field, safety assessment is also called risk assessment. Risk assessment is one method to determine the likelihood and severity of accidents and risk, using the principles and methods of the systems engineering to identify and analyze the hazard factors existing in the systems (). It aims to assess the security of the overall system and provide preventive measures in accordance with scientific procedures and methods which need to do plenty of researches and analysis on the possibility of accidents and the extent of damage and injury. Channel is the navigation passage providing safety for the ships in lakes, bays, ports and other waters. And channel safety is an important subsystem of maritime safety. Most sea traffic accidents are ship collision and grounding in the port and channel. The safe navigation of ships, especially in narrow shipping waterways, is of the utmost concern to the maritime authorities (). At the same time, due to the changes of larger tonnage and faster speed of the modern ships, the increasingly heavy traffic density and the limitation of the fairways, the risk for navigating a ship in coastal waters is getting much higher than before (). The basic cause of the accidents is the human error..but the hidden reasons at the back of the accidents should not be neglected. This is the imperfection of the management system of channel. To reduce the types of accidents, a number of navigation assistance systems capable of assisting mates have been introduced, and the proposed systems either use the accumulated actual accident data for simulation purposes or assume real-time vessel data using programming tools (). In order to improve the security situation of shipping transportation, especially for the cargo transportation, channel safety management needs to be taken seriously. And the first step is to have a accuracy understanding about the safety level. Channel around the port and lane in the open sea are both the fundamental unit of route. However, compared with the sea lane, channel is more manageable and less uncertain. According to the result of channel safety assessment in ship navigation, the problems and potential safety hazard of the channel can be found in advance. To solve problems and eliminate hazard, relevant maintenance and desilting will be arrange which is of great significance in shipping

4 Wu et al safety. LITERATURE REVIEW Many researches of safety assessment have been done by lots of scholars of various countries who put forward some methodologies to study the issue in detail, such as the grey system theory, the Bayesian network, the artificial neural networks, the set pair analysis theory and fuzzy set theory. Fuzzy set theory is appropriate for making decisions with complex situation when the context of the problem is usually unclear. The three most familiar model of fuzzy set is fuzzy synthetic evaluation model, fuzzy evidence theory and fuzzy logic model. Htin and Xiao () developed a fuzzy synthetic evaluation method to assess the channel safety of the Yangon River. Zhang et al. () presented fuzzy synthetic evaluation method to find out the security status of the main channel navigation environmental of Lianyungang Port. Yang et al. () proposed a subjective security-based assessment and management framework using fuzzy evidential reasoning to deal with the security in situation of high-level uncertainty. Adam et al. () explored the application of the fuzzy logic for risk assessment of major hazards connected with transportation of flammable substances in long pipelines. The research of channel safety assessment is not many. And using fuzzy logic model to deal with the problem is less common. Nevertheless, fuzzy logic method has many applications in relative fields. How (0) established the development of safety model using fuzzy logic for modeling various design variables for maritime and offshore safety. Rainer () focused on risk assessment in container security with fuzzy Logic model and showed that this method could help decision makers to response quickly. Li et al. () presents a dynamic fuzzy logic model (DFLM) which is able to improve the long-term safety management situation of marine crude oil transportation. FUZZY LOGIC METHOD The concept of fuzzy sets was proposed by L. A. Zadeh in. Fuzzy logic theory is mainly used to study some ambiguous problems in the real world and make them clarity. The introduction of fuzzy logic can help complex system express human judgment and thought process in mathematical form directly and simply. On the basis of L. A. Zadeh's famous "incompatible theorem"(), complexity and accuracy are mutually contradictory. As the system complexity increases, people's ability of accurately description declines. Until it reaches such a limit, the accurate descriptions finally lose their meaning. In this case, the "fuzzy" description is desirable and meaningful, because at least it provides a possible solution for some problem which cannot be analyzed precisely. Fuzzy logic is a special form of expert system which uses fuzzy interval to describe sociotechnical problem. It is different from classical probability theory, which is exactly the greatest advantage and characteristic. A channel security system is a complex system that is not able to be accurately described. Fuzzy set theory () provides three useful tools to solve the problem, named the membership function, the linguistic variable and logical operator. A reasonably solve to the transition from concept of safe to unsafe is presented by using membership function

5 Wu et al instead of classical mathematics, which arbitrarily prescribes an absolute limit. The introduction of the linguistic variable makes the evaluation results of lexical expression in a colloquial way possible. Different fuzzy operators reflect different ways of thinking, so using different fuzzy operator can make an overall consideration of varieties of evaluation views. In the classical set theory, the relationship between any one element or individual and any one set only have two kinds which be represented as "belongs" and "not belong". The description is defined,and ambiguity is absolutely not allowed. However, in the fuzzy logic theory, an element can belong to different sets with different membership degrees which determine membership functions. The transition of element from one set to another is gradual rather than sudden. Fuzzy set can be entirely characterized by its membership functions, and the membership functions represent the degree of the elements belonging to the element of fuzzy subsets. The value of the membership function is continuous in a range of 0 and. Moreover, fuzzy logic model generally has the following properties, namely the degree of membership for each element in all fuzzy subsets sum to (, ).If-then rules is the core content of fuzzy logic model and the presentation of the logic, which needs a lot of data and theoretical support. SYNTHETIC ASSESSMENT SYSTEM During the navigation, there are a number of factors affecting safety directly or indirectly, and the final evaluation results depend on the combination of multiple impacts. According to different object of study, assessment system focused on different emphasis points and analysis angles. But overall, the factors can be summarized as three aspects: the human, the vessel and the environment. In order to make the assessment system become more simple and effective, it is of great importance to select the relational indicators and make a scientific assessment system which is the first step in the process of comprehensive evaluation in ship navigation. The Principle of Choosing The principles of independence There exists connection among the various factors affecting the safety, and some of them are interactive. So in the process of establishing index system, a relatively independent system should be proposed in case of the distortion of final evaluation results. The principles of scientific Assessment index system must be built on basis of scientific research, which can fully reflect the underlying mechanisms of maritime system. Therefore, when selecting evaluation factors, we should pay attention to the accuracy and reliability. The principles of practicality Practical principles should be fully considered while determining the system. It plays such a major role in the post-process work which can avoid increase unnecessary difficulty of the research and can simplify the calculation procedure.

6 Wu et al. 0 0 Safety Class In order to provide the evaluation standards for the safety assessment and analysis, the degree of safety need to be divided into several levels. As we all know, the more levels the division has, the more comprehensive and accuracy the study is. However, this practice should bring more workload. In current study, safely level is generally divided into - levels (, ). Combined with the former research experience, this paper has safety levels: security, basic security, more insecurity, unsafe and very unsafe. Safety Assessment Index System The aim of this paper is the channel, which means a more universal evaluation model should be put forward without the restriction of a specific port or particular sea area. So as to make the safety evaluation results more useful and provide more feasible suggestion, this article is based on a large number of predecessors researches on this object and feedbacks from experts. After being recombined of original and new factors, safety assessment index system for channel safety is composed of four sections: hydrometeorology, channel condition, traffic factor and management level. These four major factors can be divided into eleven sub-factors, such as wind, current, visibility, relative depth, relative width, angle of bend, traffic density, the rate of old ship, comprehensive quality of crews, navigational aids, management systems (). The complete system is shown as Figure. 0 FIGURE Safety Assessment Index System for Channel Safety APPLICATION IN CHANNEL SAFETY ASSESSMNNT Compared with other methods, especially for complex problem, fuzzy logic method has great advantages and the widespread application. And many researches show that fuzzy logic model is suitable for safety assessment. Weight Parameters In general, each element in safety assessment index system has the different status and importance for the evaluation. Respective weights are required in accordance with own

7 Wu et al. 0 0 degree of importance for factors. It is of great significance to take a scientific and effective way to get more accurate weights, which is the necessary condition for establishing correct logic and producing reasonable results. To sum up, there are three kinds of methods in determining weight and each method has its own advantage. They are subjective method, objective method and combinational method. To be more precise, expert investigation method, analytic hierarchy process (AHP), entropy method and variation coefficient method are all in common use. In many assessment index system, the weights of indexes are usually ascertained by the analytic hierarchy process which is also be used in this paper. AHP method has the following characteristics: First, it is a combination of quantitative and qualitative method to simulate the person's decision-making process. Secondly, AHP applies to the multi-index comprehensive evaluation system which is difficult to quantitatively describe. What s more, the scientificity and reliability of index system can be ensured through consistency checks and the assessment is more rational and more direct. According to the content of our study, 0 experts who work in relative field were enrolled in the questionnaire investigation and valid questionnaires were collected. The AHP method is used to analyze the research results and obtain the weight of each factor indicators. The weights of every index of the first level are showed in Table. TABLE Weights of Every Index of the First Level Channel Index Hydrometeorology condition Traffic factor Management level Weight The Membership Function In the process of safety assessment, the concept of membership degree is introduced to quantify the fuzzy information and to build a relatively simple mathematical model. Linguistic representation of parameters is listed in Table TABLE Linguistic Representations Parameter Linguistic representation Hydrometeorology Channel condition Traffic factor Management level Safety level Good Good Good High Security Normal Normal Normal Medium Basic security Bad Bad Bad Low More insecurity Unsafe Very unsafe Based on the questionnaires received from experts and the previous research

8 Wu et al. results, the constructed membership functions for the input parameters are trapezoidal. In order to unitize the different factors, the four inputs have the same membership function, so just take Hydrometeorology as an example which is presented as Figure. As the output parameter, the membership function of safety level is triangular showed as Figure. FIGURE Membership Functions Defined for Hydrometeorology 0 FIGURE Membership Functions Defined for Output Parameters If-Then Rules IF-then rules and logical operators which build relationships between the input and the output parameters are the foundation of fuzzy logic model. Based on the weights of influencing factors, the if-then rules used in this paper are showed as Appendix. Results output On the basis of If-then rules and defuzzification method, the surface viewers can be get which reflect the relationship between two inputs and the outputs. For example, one of the inputs is fixed as Hydrometeorology, and then three figures can be showed as Figure.

9 Wu et al. FIGURE Surface Viewers

10 Wu et al CASE STUDIED The Yangtze Estuary Deep Water Channel The Yangtze Estuary deep water Channel is an integral part of the Yangtze River waterway and the main channel of the waterway. With the increase of the traffic in the river, accidents such as collisions, groundings, overturns, oil-spills and fires occur repeatedly, causing serious consequences to the sustainable development of Yangtze River (0). Besides, the depth in some part of the channel is only about meters, which became a bottleneck and restricts the further development of Shanghai s economy, as well as the area of Yangtze Basin. Since, the depth of the Yangtze river deep water channel have continuous increased to. m, m and. m after three great dredging project. And the length for bidirectional channels is. km. The scale and intensity of the Yangtze Estuary deep channel dredging project is the largest in water transportation engineering history of China. Nowadays, the deep water channel can not only meet the two-way navigable need of the third and the fourth generation of container ship, and also satisfy the navigable requirement of the fifth and the sixth generation of container ship. The safety situation of Yangtze estuary deep water channel affects the amount of cargoes and containers in Shanghai port which is a huge support for the construction of international shipping center. In addition, the channel safety also impacts the port throughput of Jiangsu province and other regions. The sustainable safety situation of the Yangtze River can propels the Chinese economy development to a new level. Computational Results FIGURE Inputs and Output of Fuzzy Logic with Old Data Before the dredging project for the Yangtze deep water channel, the channel condition

11 Wu et al. and the traffic factor had an adverse impact both on the safety level and the economic benefit. The old scores of hydrometeorology, channel condition, traffic factor, management level is,, and, and the output is. shown as Figure which means that there is a potential danger in the channel safety. According to the latest data of Yangtze Estuary deep water Channel in 0, the scores of the channel are,.,. and, respectively, and the final output is. showed as Figure. 0 FIGUTRE Inputs and Output of Fuzzy Logic with Present Data We can see that because of the relatively bad situation of hydrometeorology and traffic factor, the safety level is not enough and it still can be improved. If the traffic factor and the channel condition can be improve to., and the output should be. showed as Figure.

12 Wu et al. 0 0 FIGUTRE Inputs and Output of Fuzzy Logic with Improved Data CONCLUSIONS In this paper, a fuzzy logic model aimed at channel safety assessment is put forward to export the safety level of the channel and provide data support for the channel maintenance and improvement. Fuzzy logic method is more of advantage compared to the traditional certainty probability model. It is relatively more accurate to evaluate the safety level through subordinating degree function discussed by experts and summarized from related data. Nowadays it is essential to achieve sustainable economic growth. Then the protection of original environment and the exploitation of modern channel are sometimes contradictory. This is not obvious boundary so that fuzzy logic can do a lot of help. Clear intuitive evaluation results and graphs can be gain after determining the corresponding input variable. In addition, a summary of the safety factors that influence the channel is also carried out. This is the basis of the fuzzy logic model. Through case analysis of the argument, it can be seen that the results is similar to the research results of former scholars through using this model to assess the safety of Yangtze estuary deep water channel. Besides, the fuzzy logic model is easy to be applied. After the channel safety index system is established and the weight of each factor is determined, the model is settled. When changing the inputs, the outputs have undergone a correspondingly change. According to the differences between the results, the defects and problems can be found which can impact negative influences on the performance and the efficiency of channel. In the future study, we still need to revise the if-then rules and the weight of factors. The

13 Wu et al procedure of ship navigation in the channel can be divided into more detail process. If the most risky process is known, then the safety assessment is more targeted. In conclusion, the introduction of fuzzy logic model can make the evaluation process more in line with the rules of development of objective things, and fuzzy logic model of the channel safety assessment in ship navigation is reasonable and feasible. REFERENCES. Chen, Z.H.,and Q.G, Fang. Application of FSA to the safety vessel traffic in Shanghai Harbor. Navigation of China, Journal of the China Institute of Navigation, Shanghai, No., 00, pp-.. Li, H.B., Y.L., Hao, and H.Y.,Yu. A study of ship integrated navigation system risk assessment based on fuzzy analytic hierarchy process. International Conference on Educational and Network Technology (ICENT) 00.. Li, S.Y., Q., Meng, and X.B.,Qu. An overview of maritime waterway quantitative risk assessment models. Risk Analysis, Volume, Issue, 0, pp... Hu, S.P., Q.G., Fang and C.Q., Cai. Formal risk assessment of vessel traffic at coastal waters. Fifth International Joint Conference on INC, IMS and IDC,00.. Kim, D.Y., G.K., Park, and M.S., Chang. A study on the management method of ship entity for intelligent navigation safety information system. Soft Computing and Intelligent Systems (SCIS) and International Symposium on Advanced Intelligent Systems (ISIS), 0.. Htin A. K., and Y.J.,Xiao. Assessment on the Yangon River channel based on fuzzy synthetic evaluation. Journal of Shanghai Maritime University, Vol., No., 00, pp Zhang, B.G., L.X.,Gan, L.L.,Ma, and X.M.,Luo. Risk Assessment of the Main Channel Navigation Environmental of Lianyungang Port Based on Fuzzy Comprehensive Assessment. Port Engineering Technology, Vol., No., 0,pp. -.. Yang, Z. L., J., Wang, S., Bonsall, and Q. G., Fang. Use of Fuzzy Evidential Reasoning in Maritime Security Assessment. Risk Analysis, Volume, Issue, 00, pp. 0.. Adam, S.M.,and M. Sam Mannanb. Fuzzy Logic for Piping Risk Assessment. Journal of Loss Prevention in the Process Industries,Volume, Issue, 00, pp.. 0. How, S. S., T., Ruxtona and J., Wang. A Fuzzy-logic-based Approach to Qualitative Safety Modeing for Marine Systems. Reliability Engineering & System Safety, Volume, Issue, 00, pp... Rainer M. Risk Assessment in Container Security Using Fuzzy Logic. International Conference on Computing Logistics (ICCL), 00.. Li, Y.Z., H., Hu, and D.Z., Huang. Dynamic fuzzy logic model for risk assessment of marine crude oil transportation. In transportation Research Record: Journal of the Transportation Research Board. Transportation Reasearch Board of the National Academics,Washingon,D.C.,0,pp.-.. Zadeh, L.A.. An Outline of a New Approach to the Analysis of Complex Systems and Decision Processes. Systems, Man and Cybernetics (SMC), Vol.,Issue,,pp. -.

14 Wu et al. 0. Zadeh, L.A.. Commonsense Reasoning based on Fuzzy Logic.Winter Simulation Conference,, pp.. Dubois, D., and H.,Prade. Operations on Fuzzy Numbers. International Journal of Systems Science. V0., No.,, pp. -.. Laarhoven, van P.J.M. and W., Pedrycz. A Fuzzy Extension of Satty s Priority Theory.Fuzzy Sets and Systems.Vol., No.,, pp... Yang,Y.Q., S.H.,He, and F.L., Qi, Method of fuzzy synthetic evaluation for safety of operating railway tunnel lining. Rock Mechanics: Achievements and Ambitions, 0.. Hu, S.P., Q.G.,Fang, H.B.,Xia and Y.T., Xi(00). Formal safety assessment based on relative risks model in ship navigation. Reliability Engineering and System Safety,Vol, Issue, 00,pp... Wang,Z.S., Y.X., Jin, and Z.H., Zhu. Sea-route navigation environment risk evaluation based on uncertainty AHP. Navigation of China, No., 00,pp Zhang, D., X.P.,Yan, Z.L.,Yang, and J.,Wang. Application of formal safety assessment to navigational risk evaluation of Yangtze River. International Conference on Ocean, Offshore and Arctic Engineering, Vol., 0 pp.-.

15 Wu et al. Appendix The Table of If-then Rules No. Hydrometeo Channel Traffic Management Safety -rology condition factor level Level Good Good Good High Security Good Good Good Medium Security Good Good Good Low Basic security Good Good Normal High Basic security Good Good Normal Medium Basic security Good Good Normal Low More insecurity Good Good Bad High More insecurity Good Good Bad Medium More insecurity Good Good Bad Low Unsafe 0 Good Normal Good High Security Good Normal Good Medium Basic security Good Normal Good Low More insecurity Good Normal Normal High Basic security Good Normal Normal Medium More insecurity Good Normal Normal Low Unsafe Good Normal Bad High Unsafe Good Normal Bad Medium Unsafe Good Normal Bad Low Very unsafe Good Bad Good High More insecurity 0 Good Bad Good Medium More insecurity Good Bad Good Low More insecurity Good Bad Normal High Unsafe Good Bad Normal Medium Unsafe Good Bad Normal Low Very unsafe Good Bad Bad High Very unsafe Good Bad Bad Medium Very unsafe Good Bad Bad Low Very unsafe Normal Good Good High Basic security Normal Good Good Medium Basic security 0 Normal Good Good Low More insecurity Normal Good Normal High Basic security Normal Good Normal Medium More insecurity Normal Good Normal Low Unsafe Normal Good Bad High Unsafe Normal Good Bad Medium Unsafe Normal Good Bad Low Very unsafe Normal Normal Good High Basic security Normal Normal Good Medium Basic security Normal Normal Good Low More insecurity 0 Normal Normal Normal High More insecurity

16 Wu et al. No. Hydrometeo Channel Traffic Management Safety -rology condition factor level Level Normal Normal Normal Medium More insecurity Normal Normal Normal Low Unsafe Normal Normal Bad High Unsafe Normal Normal Bad Medium Very unsafe Normal Normal Bad Low Very unsafe Normal Bad Good High Unsafe Normal Bad Good Medium Unsafe Normal Bad Good Low Very unsafe Normal Bad Normal High Very unsafe 0 Normal Bad Normal Medium Unsafe Normal Bad Normal Low Very unsafe Normal Bad Bad High Very unsafe Normal Bad Bad Medium Very unsafe Normal Bad Bad Low Very unsafe Bad Good Good High More insecurity Bad Good Good Medium More insecurity Bad Good Good Low Unsafe Bad Good Normal High Unsafe Bad Good Normal Medium Unsafe 0 Bad Good Normal Low Very unsafe Bad Good Bad High Very unsafe Bad Good Bad Medium Very unsafe Bad Good Bad Low Very unsafe Bad Normal Good High Unsafe Bad Normal Good Medium Unsafe Bad Normal Good Low Very unsafe Bad Normal Normal High Unsafe Bad Normal Normal Medium Very unsafe Bad Normal Normal Low Very unsafe 0 Bad Normal Bad High Very unsafe Bad Normal Bad Medium Very unsafe Bad Normal Bad Low Very unsafe Bad Bad Good High Very unsafe Bad Bad Good Medium Very unsafe Bad Bad Good Low Very unsafe Bad Bad Normal High Very unsafe Bad Bad Normal Medium Very unsafe Bad Bad Normal Low Very unsafe Bad Bad Bad High Very unsafe 0 Bad Bad Bad Medium Very unsafe Bad Bad Bad Low Very unsafe