CN-121999639-A - Real-time ship anchor walking intelligent monitoring and early warning method and system

CN121999639ACN 121999639 ACN121999639 ACN 121999639ACN-121999639-A

Abstract

The invention provides a real-time ship anchor walking intelligent monitoring and early warning method and system, wherein track data in a normal anchoring state is firstly obtained based on a ship operation simulator, deflection width, ground speed and heading change rate are extracted as anchor walking monitoring indexes, and a Gaussian function is adopted to construct a standard cloud model of each index; the method comprises the steps of inputting indexes acquired in real time into a corresponding cloud model to obtain membership degrees, calculating real-time anchor walking probability by combining an entropy weight method, calculating net resultant force based on wind, flow and anchor chain parameters, estimating a drift velocity vector range after anchor walking, fitting real anchor positions by AIS data, generating a drift endpoint set by Monte Carlo simulation, dividing four drift danger level intervals of danger, bias safety and safety, generating a ship space density distribution map covering the intervals by combining nuclear density estimation, finally combining anchor walking probability and density information to calculate a comprehensive risk value, triggering sound light grading early warning according to a preset threshold, and realizing intelligent anchor walking risk prevention and control of high-precision early warning.

Inventors

ZHU YU
XI YONGTAO
ZHU QINGHUA
HAN BING
CHEN PENGJIE
HU SHENPING
FAN CUNLONG
WU JIANJUN

Assignees

上海船舶运输科学研究所有限公司
上海海事大学

Dates

Publication Date: 20260508
Application Date: 20251230

Claims (10)

1. The intelligent monitoring and early warning method for the ship anchor walking in real time is characterized by comprising the following steps in sequence: The method comprises the steps of constructing a standard cloud model, namely, carrying out anchoring simulation experiments under various sea conditions based on a ship operation simulator, obtaining track data of a ship in a normal anchoring state as historical normal sample data, extracting deflection width, ground speed and heading change rate from the historical normal sample data as anchor running monitoring indexes; Respectively inputting the deflection width, the ground speed and the heading change rate acquired at the current moment into corresponding standard cloud models to obtain a first membership degree of the deflection width in a normal anchoring state, a second membership degree of the ground speed in the normal anchoring state and a third membership degree of the heading change rate in the normal anchoring state; The method comprises the steps of determining an actual anchor position, namely acquiring real-time AIS data of a ship, ship type parameters, anchor chain parameters and wind parameters and flow parameters in the current environment, wherein the real-time AIS data comprise ship position coordinates and course angles, the ship type parameters comprise ship length, draft, forward projection area of the ship on a waterline and side projection area of the ship on the waterline, the anchor chain parameters comprise anchor weights and anchor chain azimuth angles, the wind parameters comprise relative wind speeds and relative wind sponsons, and the flow parameters comprise relative flow speeds of water to an anchored ship; Calculating net resultant force and drift velocity, namely calculating wind power by using a Hughes formula according to relative wind speed, relative wind side angle, forward projection area of a ship body on a waterline and side projection area of the ship body on the waterline, calculating hydrodynamic force by using a hydrodynamic formula according to the length of the ship body, draft and relative flow velocity of water to an anchored ship, and calculating anchor chain residual force by using a static anchor chain tension formula according to the anchor weight, course angle and anchor chain azimuth angle; The method comprises the steps of generating a drift danger level region, namely randomly sampling a drift process of a ship after the ship is anchored by adopting a Monte Carlo method for a plurality of times based on a drift speed vector range and combining with a preset reaction time required by the ship to be controlled again, obtaining a plurality of drift terminals taking actual anchor position coordinates as drift starting points to form a ship drift terminal position set, and dividing a sea area around the ship into a danger zone, a bias safety zone and a safety zone as drift danger level regions according to drift distance and direction distribution characteristics of the ship drift terminal position set and taking the actual anchor position coordinates as the center; The method comprises the steps of acquiring position coordinates of all ships in an anchor area in a preset time window from real-time AIS data, generating a real-time ship space density distribution map covering a drift danger level area based on the acquired position coordinates of all the ships by adopting a nuclear density estimation method; And the comprehensive anchor risk value calculation and early warning step comprises the steps of calculating the comprehensive anchor risk value according to the real-time anchor probability and combining a drift risk level region and a real-time ship space density distribution map, comparing the comprehensive anchor risk value with a preset four-level risk threshold interval, and triggering an audible and visual alarm device to output an early warning level corresponding to the risk threshold interval when the comprehensive anchor risk value falls into any risk threshold interval.
2. The method for intelligent monitoring and early warning of ship anchor walking in real time according to claim 1, wherein in the step of constructing the standard cloud model, the historical normal sample data includes a historical normal data sample set of drift width, a historical normal data sample set of ground speed and a historical normal data sample set of heading change rate, and the step of constructing the standard cloud model of each anchor walking monitoring index by using a gaussian function comprises the following steps: Based on a historical normal data sample set of the deflection width, the ground speed and the heading change rate, calculating the digital characteristics of each anchor walking monitoring index by adopting a Gaussian function, and respectively constructing a standard cloud model of the deflection width, a standard cloud model of the ground speed and a standard cloud model of the heading change rate based on the digital characteristics of each anchor walking monitoring index, wherein the digital characteristics comprise an expected value, entropy and super entropy.
3. The method for intelligent monitoring and early warning of ship anchor running in real time according to claim 1, wherein in the step of calculating and early warning the comprehensive anchor running risk value, calculating the comprehensive anchor running risk value specifically includes: Space matching is carried out on all drift end positions and drift dangerous level areas, the number of drift end points falling into each section in the drift dangerous level areas is counted, the drift probability of each section is calculated according to the number of the drift end points and the total number of the drift end points of each section, the average ship density value of each section in a real-time ship space density distribution diagram is queried, the result weight of each section is obtained based on the average ship density value, and the anchor passing result severity coefficient is calculated based on the drift probability and the result weight, wherein the following formula is shown: , Wherein, the Is the severity coefficient of the anchor-moving result; probability of drifting into the kth interval for the vessel; The result weight of the kth interval; and calculating a comprehensive anchor walking risk value according to the real-time anchor walking probability and the anchor walking result severity coefficient, wherein the comprehensive anchor walking risk value is shown in the following formula: R = P × S, Wherein R is a comprehensive anchor risk value, P is a real-time anchor probability, and S is an anchor result severity coefficient.
4. The method for intelligent monitoring and early warning of ship anchor running in real time according to any one of claims 1 to 3, wherein in the step of calculating net resultant force and drift velocity, the hydrodynamic formula is as follows: , Wherein, the For the purpose of wind power, the air-powered vehicle is provided with a wind power generation device, In order to achieve an air density of the air, Is the coefficient of wind power, For the relative wind speed, Is the orthographic projection area of the ship body on the waterline, Is the projection area of the side of the ship body on the waterline, Is the relative wind side angle; the hydrodynamic formula is shown as follows: , Wherein, the Is used as the water power, Is the density of the seawater, and the seawater is the density of the seawater, Is the moment coefficient of the hydrodynamic rotating ship, For the length of the hull, d is the draft, Is the relative flow rate of water to the vessel; the static chain tension formula is as follows: , Wherein, the In order for the anchor chain to remain in force, In order to be a coefficient of the gripping force, In order to be an anchor weight, In order to be the heading angle, Is the anchor chain azimuth.
5. The method for intelligent monitoring and early warning of ship anchor walking in real time according to any one of claims 1 to 3, wherein in the step of calculating and early warning the comprehensive anchor walking risk value, the early warning level includes four levels of early warning, three levels of early warning, two levels of early warning and one level of early warning, wherein: four-level early warning, namely, comprehensive anchor risk value R epsilon (0.20, 0.40) corresponds to blue early warning, and the anchor risk value R epsilon is low in anchor probability and no ship or obstacle exists in dangerous areas or biased dangerous areas; Three-stage early warning, namely, synthesizing anchor running risk values R epsilon (0.40, 0.60) corresponding to yellow early warning, wherein the anchor running probability is medium and ships or barriers exist in a partial danger zone; The second-level early warning, namely, comprehensive anchor risk value R epsilon (0.60, 0.80) corresponds to orange early warning, and represents that the anchor probability is high and ships or barriers exist in dangerous areas or partial dangerous areas; and the first-stage early warning is to integrate the anchor risk value R epsilon (0.80, 1.00) and correspond to the red early warning, which shows that the anchor probability is extremely high and ships or barriers exist in dangerous areas.
6. The real-time ship anchor walking intelligent monitoring and early warning system is characterized by comprising a standard cloud model construction module, a real-time anchor walking probability calculation module, an actual anchor position determination module, a net resultant force and drift speed calculation module, a drift danger level area generation module, a ship space density distribution diagram generation module and a comprehensive anchor walking risk value calculation and early warning module which are connected in sequence, The standard cloud model construction module is used for carrying out anchoring simulation experiments under various sea conditions based on a ship operation simulator, obtaining track data of a ship in a normal anchoring state as historical normal sample data, and extracting deflection width, ground speed and heading change rate from the historical normal sample data to be used as anchor running monitoring indexes; The real-time anchor walking probability calculation module is used for respectively inputting the deflection width, the ground speed and the heading change rate acquired at the current moment into the corresponding standard cloud models to obtain a first membership degree of the deflection width belonging to a normal anchoring state, a second membership degree of the ground speed belonging to the normal anchoring state and a third membership degree of the heading change rate belonging to the normal anchoring state; The real anchor position determining module is used for acquiring real-time AIS data of the ship, ship type parameters, anchor chain parameters and wind parameters and flow parameters in the current environment, wherein the real-time AIS data comprise ship position coordinates and course angles, the ship type parameters comprise ship length, draft, forward projection area of the ship on the waterline and side projection area of the ship on the waterline, the anchor chain parameters comprise anchor weight and anchor chain azimuth angle, the wind parameters comprise relative wind speed and relative wind sponson angle, and the flow parameters comprise relative flow velocity of water to the moored ship; the net resultant force and drift velocity calculation module calculates wind power by using a Hughes formula according to relative wind speed, relative wind sponson angle, forward projection area of a ship body on a waterline and side projection area of the ship body on the waterline, calculates water power by using a water power formula according to the length of the ship body, draft and relative flow velocity of water to an anchored ship, calculates anchor chain residual force by using a static anchor chain tension formula according to the anchor weight, course angle and anchor chain azimuth angle, calculates net resultant force based on the wind power, the water power and the anchor chain residual force, and calculates drift velocity vector range of the ship after anchor walking according to the net resultant force; The drift danger level region generation module is used for randomly sampling the drift process of the ship after the ship is anchored by adopting a Monte Carlo method for a plurality of times based on the drift velocity vector range and combining with the preset reaction time required by the ship to be controlled again, so as to obtain a plurality of drift end points taking actual anchor position coordinates as drift start points to form a ship drift end point position set; The ship space density distribution map generation module acquires position coordinates of all ships in an anchor area in a preset time window from real-time AIS data, generates a real-time ship space density distribution map covering the drifting danger level area based on the acquired position coordinates of all ships by adopting a nuclear density estimation method; The comprehensive anchor risk value calculation and early warning module calculates a comprehensive anchor risk value according to the real-time anchor probability and by combining a drift risk level region and a real-time ship space density distribution map, compares the comprehensive anchor risk value with a preset four-level risk threshold interval, and triggers an audible and visual alarm device to output an early warning level corresponding to the risk threshold interval when the comprehensive anchor risk value falls into any risk threshold interval.
7. The intelligent monitoring and early warning system for ship anchor walking in real time according to claim 6, wherein the standard cloud model construction module is characterized in that the historical normal sample data comprises a historical normal data sample set with a drift width, a historical normal data sample set with a speed to ground and a historical normal data sample set with a heading change rate, and the standard cloud model construction module is characterized in that the standard cloud model construction module is used for respectively constructing each anchor walking monitoring index by adopting a Gaussian function and comprises the following specific steps: Based on a historical normal data sample set of the deflection width, the ground speed and the heading change rate, calculating the digital characteristics of each anchor walking monitoring index by adopting a Gaussian function, and respectively constructing a standard cloud model of the deflection width, a standard cloud model of the ground speed and a standard cloud model of the heading change rate based on the digital characteristics of each anchor walking monitoring index, wherein the digital characteristics comprise an expected value, entropy and super entropy.
8. The intelligent monitoring and early warning system for ship anchor running in real time according to claim 6, wherein in the comprehensive anchor running risk value calculating and early warning module, calculating the comprehensive anchor running risk value specifically includes: Space matching is carried out on all drift end positions and drift dangerous level areas, the number of drift end points falling into each section in the drift dangerous level areas is counted, the drift probability of each section is calculated according to the number of the drift end points and the total number of the drift end points of each section, the average ship density value of each section in a real-time ship space density distribution diagram is queried, the result weight of each section is obtained based on the average ship density value, and the anchor passing result severity coefficient is calculated based on the drift probability and the result weight, wherein the following formula is shown: , Wherein, the Is the severity coefficient of the anchor-moving result; probability of drifting into the kth interval for the vessel; The result weight of the kth interval; and calculating a comprehensive anchor walking risk value according to the real-time anchor walking probability and the anchor walking result severity coefficient, wherein the comprehensive anchor walking risk value is shown in the following formula: R = P×S, Wherein R is a comprehensive anchor risk value, P is a real-time anchor probability, and S is an anchor result severity coefficient.
9. The intelligent monitoring and early warning system for ship anchor running in real time according to any one of claims 6 to 8, wherein in the net resultant force and drift velocity calculation module, the hydrodynamic formula is as follows: , Wherein, the In order to achieve an air density of the air, For the relative wind speed, Is the orthographic projection area of the ship body on the waterline, The projection area of the side of the ship body on the water line, In order to be a relative wind-port angle, Is the wind power coefficient; the hydrodynamic formula is shown as follows: , Wherein, the Is the density of the seawater, and the seawater is the density of the seawater, Is the moment coefficient of the hydrodynamic rotating ship, The length of the hull, d is the draft, The relative flow rate of water to the vessel; the static chain tension formula is as follows: , Wherein, the In order for the anchor chain to remain in force, In order to be a coefficient of the gripping force, In order to be an anchor weight, In order to be the heading angle, Is the anchor chain azimuth.
10. The intelligent monitoring and early warning system for ship anchor handling in real time according to any one of claims 6 to 8, wherein the early warning level comprises four levels of early warning, three levels of early warning, two levels of early warning and one level of early warning, wherein: four-level early warning, namely, comprehensive anchor risk value R epsilon (0.20, 0.40) corresponds to blue early warning, and the anchor risk value R epsilon is low in anchor probability and no ship or obstacle exists in dangerous areas or biased dangerous areas; Three-stage early warning, namely, synthesizing anchor running risk values R epsilon (0.40, 0.60) corresponding to yellow early warning, wherein the anchor running probability is medium and ships or barriers exist in a partial danger zone; The second-level early warning, namely, comprehensive anchor risk value R epsilon (0.60, 0.80) corresponds to orange early warning, and represents that the anchor probability is high and ships or barriers exist in dangerous areas or partial dangerous areas; and the first-stage early warning is to integrate the anchor risk value R epsilon (0.80, 1.00) and correspond to the red early warning, which shows that the anchor probability is extremely high and ships or barriers exist in dangerous areas.

Description

Real-time ship anchor walking intelligent monitoring and early warning method and system Technical Field The invention relates to the technical field of ship navigation safety and intelligent monitoring, in particular to a real-time ship anchor-moving intelligent monitoring and early warning method and system. Background Anchoring is a necessary and conventional operation taken by a ship when performing tasks such as platform avoidance, tide waiting, berthing, cargo passing, health quarantine and the like on the sea. However, in the anchoring process, the ship is in a static state or a low-speed swinging state for a long time, is extremely easy to continuously act by severe weather such as wind, waves, currents and the like and marine environment factors, and is possibly caused by the influence of uncertainty such as uneven seabed substrate, attenuation of anchor holding force, loose anchor chains or improper arrangement and the like, so that the fluke loses effective ground holding capacity, and the anchor running phenomenon is caused. Once the ship is anchored, the ship drifts along with wind flow, not only can deviate from a preset anchor position, but also can collide with other anchoring ships, navigation ships or offshore facilities, even serious marine accidents such as stranding, capsizing or oil spilling are caused, and the structural safety of the ship, the lives of crews and the marine ecological environment are seriously threatened. Therefore, the effective and timely monitoring and early warning of the state of the moored ship are particularly important for guaranteeing the navigation safety. Currently, the judgment of anchor walking in the industry mainly depends on the traditional manual experience method. For example, the "observation anchor chain judgment method" presumes the anchor by visually judging whether the anchor chain is straight or severely shaked, and the "anchoring warning circle judgment method" defines a circular area with a certain radius with a preset anchor position as the center, and the anchor chain judgment method is regarded as the anchor when the AIS or the radar displays that the ship position exceeds the circle. However, the method essentially belongs to a post-event or in-event monitoring means, can only be perceived after the ship has significantly displaced, lacks early sensing capability on the anchor tendency, cannot provide prospective early warning, and is difficult to meet the requirement of modern intelligent shipping on active safety prevention and control. With the development of automation and informatization technologies, some automatic anchor feeding monitoring alarm systems based on sensors and data processing appear in recent years, and the first type of the alarm systems can be roughly divided into the following two types, namely alarm technologies based on stress analysis. According to the method, the tension of an anchor chain is measured in real time by installing a tension sensor, theoretical anchor holding power is calculated by combining environmental parameters such as wind, flow and the like, and when the measured tension exceeds the limit holding power of an anchor, the risk of anchor walking is judged. However, the method has multiple challenges in practical application, namely, on one hand, the high-precision, corrosion-resistant and long-term stable anchor chain tension sensor has high cost and is easily interfered by marine environment, and on the other hand, the accurate acquisition of real-time wind speed, flow direction, wave spectrum and seabed substrate parameters is extremely difficult, and the anchor-soil interaction model has high nonlinearity and uncertainty, so that the theoretical grasping force estimation error is large, and finally, the system has high false alarm rate and poor robustness and is difficult to stably operate under complex sea conditions. The second type is an alarm technique based on state detection. The method mainly utilizes AIS, GNSS or radar equipment to continuously monitor the ship position, ground Speed (SOG), heading and change rate and other motion state parameters, and when the SOG continuously exceeds a certain threshold (e.g. 0.5 section) or the ship position deviates from the initial anchor position by more than a set distance, the anchor moving alarm is triggered. Although this approach enables automated identification, its nature is still to confirm the fact of anchor failure that has occurred, rather than to predict the tendency of anchor failure. Because the time delay exists from the weakening of the holding power of the ship to obvious displacement, the method usually gives an alarm when the ship is not anchored in a reversible way, the early warning window is extremely short, and the adequate emergency treatment time is difficult to be reserved for the crew. In summary, with the application of artificial intelligence, big data and other technologies in the shipping industry, the development