CN-122022494-A - Marine transportation safety analysis method based on AI and ship AIS big data

CN122022494ACN 122022494 ACN122022494 ACN 122022494ACN-122022494-A

Abstract

The invention discloses a marine transportation safety analysis method based on AI and ship AIS big data, which belongs to the technical field of marine wind power plants and marine culture area site selection, and concretely comprises the steps of obtaining ship AIS data of a target sea area to generate a ship AIS traffic flow analysis chart, dividing the target sea area into analysis grid units and calculating external navigation risk scores of the grid units, screening and aggregating to generate a plurality of candidate site areas based on the risk scores, extracting site boundary parameters, determining candidate parent harbors corresponding to the candidate site areas, planning operation and maintenance routes from the parent harbors to sites based on electronic chart data, calculating navigation difficulty coefficients and selecting a minimum value as an operation and maintenance path conflict index, weighting and summing the external navigation risk scores and the operation and maintenance path conflict index to obtain a comprehensive score, and selecting the site area with the lowest score as the target site area. According to the method, the external navigation risk and the operation and maintenance path safety are brought into a unified decision frame, and the multi-dimensional collaborative optimization of wind farm site selection is realized.

Inventors

LIN CHAOMING
LI ZIHENG
WU WUTAO
HUANG XIAOJING
JIANG JINYUN
LIN LONGXIAN
WU YIXIN
LIN SHIRONG
Shen Ruizhe
FENG CHENGLIN
Ding Xinkun

Assignees

福建港航船舶工程咨询管理有限公司

Dates

Publication Date: 20260512
Application Date: 20260409

Claims (8)

1. The marine transportation safety analysis method based on AI and ship AIS big data is characterized by comprising the following steps: S1, acquiring ship AIS data of a target sea area, and performing track extraction and density analysis on the ship AIS data to generate a ship AIS traffic flow analysis chart covering the target sea area; S2, dividing the target sea area into a plurality of evaluation grid cells, extracting the normalized density value of each analysis grid cell based on the ship AIS traffic flow analysis graph, and calculating the external navigation risk score of each analysis grid cell; s3, determining a plurality of candidate site areas based on the external navigation risk scores of the evaluation grid units, and extracting site boundary parameters of the candidate site areas; s4, acquiring peripheral port information corresponding to each candidate site area, determining a plurality of candidate parent ports corresponding to each candidate site area, and planning operation and maintenance routes from each candidate parent port to the corresponding candidate site area based on electronic chart data to obtain all operation and maintenance routes corresponding to each candidate site area; s5, extracting the route characteristic parameters of each operation and maintenance route, calculating the navigation difficulty coefficient of each operation and maintenance route corresponding to each candidate site area based on the route characteristic parameters, and selecting the minimum navigation difficulty coefficient corresponding to each candidate site area as the operation and maintenance route conflict index of the candidate site area; and S6, calculating the corresponding comprehensive scores of the candidate site areas based on the external navigation risk scores and the operation and maintenance path conflict indexes of the candidate site areas, and selecting the candidate site area with the lowest comprehensive score as a target site area and outputting the target site area.
2. The marine transportation safety analysis method based on AI and ship AIS big data according to claim 1, wherein in S1, the specific generation process of the ship AIS traffic flow analysis chart is as follows: Acquiring ship AIS data of a target sea area, wherein the ship AIS data comprises static information and dynamic information of each ship, the static information comprises ship types, ship capturers and ship widths, and the dynamic information comprises ship positions, navigational speeds, heading and time stamps; Preprocessing the ship AIS data, and removing abnormal data with static information missing and positions exceeding the boundary of the target sea area to obtain preprocessed ship AIS data; extracting track point sequences of all ships based on the pretreated ship AIS data, wherein the track point sequences are arranged in ascending order according to time stamps to form ship track data of all ships; Dividing the target sea area into a plurality of analysis grid cells, and performing density calculation on ship track points in each analysis grid cell by adopting a nuclear density estimation algorithm to generate ship track point density values of each analysis grid cell; And overlapping the normalized density value of each analysis grid unit to the electronic chart data according to the space position by taking the electronic chart data as a base chart to generate a ship AIS traffic flow analysis chart.
3. The marine transportation safety analysis method based on AI and ship AIs big data according to claim 2, wherein in S2, the specific calculation process of the external navigation risk score is: Acquiring ship AIS data corresponding to each analysis grid unit, and extracting normalized density values of each analysis grid unit and the number of track points corresponding to each ship type in each analysis grid unit from the ship AIS data, wherein the ship types comprise a commercial ship, a fishing ship, a cargo ship, a passenger ship and an engineering ship; Calculating the track point composition proportion of each ship type in each analysis grid unit based on the ratio of the track point number corresponding to each ship type in each analysis grid unit to the total track point number of all ship types in the analysis grid unit; multiplying the normalized density value of each analysis grid unit by the track point composition ratio of each ship type to obtain the subentry density value of each ship type in each analysis grid unit; acquiring a preset ship risk weight coefficient set, wherein the ship risk weight coefficient set comprises risk weight coefficients corresponding to various ship types; The method comprises the steps of multiplying the subitem density value of each ship type in each analysis grid unit with the risk weight coefficient of the corresponding ship type to obtain the subitem risk value of each ship type, and summing the subitem risk values of all the ship types in the analysis grid units to obtain the external navigation risk score of each analysis grid unit.
4. The marine transportation safety analysis method based on AI and ship AIS big data according to claim 1, wherein in S3, the specific process of determining a plurality of candidate site areas is as follows: Obtaining the external navigation risk scores of all the analysis grid cells, and marking the analysis grid cells with the external navigation risk scores lower than a preset risk threshold as candidate grid cells; and carrying out connected domain aggregation on each candidate grid cell, merging adjacent candidate grid cells into a connected region, generating a plurality of candidate address areas, and extracting boundary coordinate sequences of each candidate address area to serve as the address boundary parameters of each candidate address area.
5. The marine transportation safety analysis method based on AI and ship AIs big data according to claim 1, wherein in S4, the specific planning process of the operation and maintenance route from each candidate parent port to the corresponding candidate site area is as follows: Acquiring electronic chart data of a target sea area, wherein the electronic chart data comprises a water depth distribution chart layer, a navigation obstacle distribution chart layer and a navigation forbidden area distribution chart layer; And determining a geometric center point of each candidate site area as a route end point by using a boundary coordinate sequence of each candidate site area, using port position coordinates of candidate parent ports corresponding to each candidate site area as route start points, planning a shortest navigation path which meets the constraint conditions that the water depth is not less than a preset water depth threshold and avoids a navigation obstacle distribution map layer and a navigation forbidden area distribution map layer based on the electronic chart data, and generating an operation and maintenance route from each candidate parent port to the corresponding candidate site area.
6. The marine transportation safety analysis method based on AI and ship AIS big data according to claim 1, wherein in S5, the specific extraction process of the route characteristic parameters of each operation and maintenance route is as follows: acquiring operation and maintenance routes from each candidate parent port to a corresponding candidate site area, wherein the operation and maintenance routes comprise a path point sequence arranged according to a navigation sequence, and each path point comprises position coordinates; Sequentially calculating Euclidean distances between adjacent path points along the operation and maintenance route, and accumulating distance values between each adjacent path point to obtain the length of the route; Extracting the water depth value of each path point on the operation and maintenance route, and selecting the minimum value in the water depth values of each path point to obtain the minimum water depth of the route; Obtaining a channel distribution map layer of a target sea area, carrying out space superposition analysis on the operation and maintenance route and the channel distribution map layer, and counting the number of crossing points of the operation and maintenance route and the channel distribution map layer to obtain the number of times of crossing the route through the channel; acquiring fishing zone distribution data of a target sea area, performing space superposition analysis on the operation and maintenance route and the fishing zone distribution data, and counting the number of crossing points of the operation and maintenance route and the fishing zone distribution data to obtain the number of times that the route passes through the fishing zone; And taking the length of the route, the minimum water depth of the route, the times of crossing the channel and the times of crossing the fishing zone of the route as the route characteristic parameters of the operation and maintenance route.
7. The marine transportation safety analysis method based on AI and ship AIS big data according to claim 1, wherein in S5, the specific calculation process of the operation and maintenance path conflict index is: acquiring all operation and maintenance routes corresponding to each candidate site area, and extracting route characteristic parameters of each operation and maintenance route, wherein the route characteristic parameters comprise route length, minimum water depth of the route, the number of times of crossing the route and the number of times of crossing the route in a fishing area; Counting the sum of the length of the routes, the sum of the water depth dangerous values, the sum of the times of the routes crossing the channels and the sum of the times of the routes crossing the fishing zone for all the operation and maintenance routes; acquiring a preset route characteristic weight coefficient set, wherein the route characteristic weight coefficient set comprises a route length weight coefficient, a water depth dangerous value weight coefficient, a route crossing channel frequency weight coefficient and a route crossing fishing zone frequency weight coefficient; For each operation and maintenance route, multiplying the ratio of the route length to the sum of the route lengths by a route length weight coefficient, multiplying the ratio of the water depth dangerous value to the sum of the water depth dangerous value by a water depth dangerous value weight coefficient, multiplying the ratio of the number of times of the route crossing the channel to the sum of the number of times of the route crossing the channel by a route crossing number of times weight coefficient, multiplying the ratio of the number of times of the route crossing the fishing zone to the sum of the number of times of the route crossing the fishing zone by a route crossing number of times weight coefficient, and then summing to obtain the navigation difficulty coefficient of the operation and maintenance route.
8. The marine transportation safety analysis method based on AI and ship AIS big data according to claim 1, wherein in S6, the specific generation process of the target site area is as follows: And obtaining the conflict indexes of the external navigation risk score and the operation and maintenance path of each candidate site area, carrying out weighted summation on the conflict indexes of the external navigation risk score and the operation and maintenance path of each candidate site area, obtaining the comprehensive score of each candidate site area, and selecting the candidate site area with the lowest comprehensive score as the target site area.

Description

Marine transportation safety analysis method based on AI and ship AIS big data Technical Field The invention relates to the technical field of offshore wind farms and maritime culture area site selection, in particular to a marine transportation safety analysis method based on AI and ship AIS big data. Background Offshore wind farms and offshore aquaculture area construction are important directions for ocean energy development, and as offshore wind farm resources tend to saturate, offshore wind projects gradually develop into offshore areas farther offshore and deeper in water depth. The offshore wind farm and the offshore culture area site selection are used as early key links of project development, and the construction cost, the operation efficiency and the navigation safety level of the offshore wind farm and the offshore culture area are directly determined. The current offshore wind farm and offshore aquaculture area site selection generally relies on a geographic information system and a multi-criterion decision-making method, multiple factors such as wind resource conditions, water depth geological conditions, ocean functional regions and the like are comprehensively considered, and candidate sites are determined through space superposition analysis and weighted scoring. In the aspect of navigation safety evaluation, the prior art generally adopts the data of the automatic ship identification system to analyze traffic flow. Specifically, historical ship track data of a target sea area is collected, a ship traffic flow density distribution map is generated by using methods such as nuclear density analysis and track band extraction, and main channel trend, habit channel distribution and a fishing boat operation dense area are identified. Based on the analysis result, the wind power plant is arranged in the sea area with lower traffic flow density of the ship by setting the safe distance in the site selection process, so that the risk of collision between the wind power plant and the passing ship is prevented. The method takes avoidance as a core principle, and focuses on the influence of wind power plant construction on the existing navigation environment. However, the prior art only focuses on the external navigation conflict between the offshore wind farm and the offshore maritime culture area and the passing ship, and does not incorporate the operation and maintenance activities after wind farm construction into the site selection evaluation system. The operation period of the offshore wind power plant and the offshore mariculture area is twenty to thirty years, during which operation and maintenance ships need to frequently operate from the female port navigation to the wind power plant, and the space overlapping of the operation and maintenance navigation line, the main navigation channel and the fishing area can directly cause new navigation safety risks. Because of the lack of quantitative evaluation on the safety of the operation and maintenance route, the site possibly selected in the prior art is far away from the main channel, but the operation and maintenance access path of the site is required to frequently pass through a high-risk water area, so that the external navigation risk and the operation and maintenance path risk are subjected to splitting treatment in the site selection decision. The split type evaluation mode enables the site selection result to expose the interaction risk of the operation and maintenance ship and the passing ship in the actual operation stage, increases the navigation safety hidden trouble of the operation and maintenance ship, and meanwhile, the selected site is subjected to the problem of increasing the fuel cost and the navigation time due to overlong navigation lines and limiting the ship passing window due to over-shallow water depth and increasing the avoidance frequency and the navigation risk due to crossing the navigation channel or the fishing zone because the parameters such as the length of the operation and maintenance path, the water depth condition, the navigation channel and the fishing zone crossing condition are not included in the site selection stage, so that the overall operation benefit of the wind power plant is finally influenced. Disclosure of Invention The invention aims to provide a marine transportation safety analysis method based on AI and ship AIS big data, which solves the following technical problems: The prior art only focuses on the external navigation conflict between the offshore wind farm and the maritime culture area and the passing ship, and does not bring the route safety of the operation and maintenance ship into site selection evaluation, so that the external navigation risk and the operation and maintenance path risk are subjected to splitting treatment in site selection decision, and the selected site faces the problems of increased potential safety hazards of the operation and maintenance ship navigation, increas