Since 2016, Korea has promoted the research & development project to build the SMART-Navigation system for the SOLAS ships as well as fishing boats and other small ships to provide the LTE-Maritime communication network for non-SOLAS ships in order to reduce maritime accidents. The services of the SMART-Navigation project include monitoring assistance service for high risk ships. This service focuses on the vulnerable situation awareness, risk assessment and accident response.
In this paper, a novel algorithm is proposed for the navigational Collision Risk (CR) solving system based on vulnerability of maritime accidents by using fuzzy logic for SMART-Navigation. Maritime vulnerability is defined as the possibility of a marine accident or degree of risk of marine accident. The factors of vulnerability are integrated with basic collision risk to expand the validity and enhance the utilization of collision risk assessment in the real scenario.
This navigational collision risk solving system is consisted of four components. In the first module, as two main factors, distance to closest point of approach (DCPA) and time of closest point of approach (TCPA) are inferred for basic collision risk using fuzzy logic.
The factors of vulnerability include bad weather, strong tidal current, accidents prone area, traffic congestion, operator fatigue and fishing boats operating area. The number of accidents occurred during 10 years in 159 waterways and the data of six factors of vulnerability in Korea are collected and analyzed by using fuzzy C-means clustering method. Based on the results from the project research team, variables related to six vulnerabilities are selected. Wind speed and wave height are the input variables for the bad weather factor. In order to consider the influence of weather condition for different sizes of ships. The singleton fuzzy reasoning logic is utilized, thus tuning algorithm is inserted to adjust the consequence of reasoning engine. The factor of strong tidal current refers to the tidal current speed and the accidents in the past 10 years. In the accident-prone areas, two input variables are selected for the reasoning engine using static ratio of accident number per hour in a day the past 10years and dynamic ratio of passing ship's number per hour. To solve the factor of congestion of shipping traffic, Static Congestion Degree (SCD) and Dynamic Congestion Degree (DCD) are considered at the same time. As we know, the operator fatigue is an important reason for collision accidents. According to the survey, time of working and navigation and tonnage of fishing vessel are chosen as the variables of the reasoning system. The distance to the center of fishing area and the size of fishing area are combined for the last factor of vulnerability. Additionally, the results of the six factors of vulnerability are integrated to calculate the combined vulnerability using Analytic Hierarchy Process (AHP) method. Then, two concepts of basic collision risk and vulnerability are incorporated to obtain the navigational collision risk.
In the simulation, ten areas are chosen from 159 narrow waterway areas to examine the validity of this navigational collision risk solving system. In each area, the input variables are used to infer the vulnerability for the factors. These results are combined by using weighted average method. The threshold value of the navigational collision risk solving system is set to alert the dangerous encounter for the service of accident response.