CN-121981540-A - Ship safety potential field modeling risk assessment method and system

CN121981540ACN 121981540 ACN121981540 ACN 121981540ACN-121981540-A

Abstract

The invention relates to the technical field of intelligent ship and marine traffic safety, in particular to a ship safety potential field modeling risk assessment method and system, comprising the steps of firstly synchronously acquiring multi-source data of ship motion, environment, rules and human factor states; the method comprises the steps of respectively constructing static environment, dynamic ship interaction, sailing rules and human factor safety potential fields, carrying out uncertainty correction, normalizing and fusing various potential fields into a unified total safety potential field, and finally integrating the total safety potential field based on a future track predicted by the ship to obtain risk situation indexes, dividing risk grades and identifying dominant risk sources. The invention realizes quantitative fusion and collaborative evaluation of human, ship, environment and regular multidimensional risk factors under a unified frame, can provide continuous and prospective navigation risk situation awareness, and provides reliable basis for collision prevention decision and safety control of intelligent ships.

Inventors

YANG SHENHUA
HUANG ZEYANG
WANG PU
CHEN GUOQUAN
Xie Daoshun
Yang Chenzheng

Assignees

集美大学

Dates

Publication Date: 20260505
Application Date: 20260119

Claims (10)

1. A method for modeling risk assessment of a marine safety potential field, comprising: step S1, acquiring and synchronizing motion state data, environment data, navigation rule data and human factor state data of a ship and other ships to determine a collaborative state set at a moment t under a navigation coordinate system; s2, constructing a static environment safety potential field changing along with the space position based on the environment data; s3, based on the motion state data, constructing a dynamic interaction safety potential field for each target ship by utilizing a ship field model related to the direction and relative motion information; s4, identifying meeting situations based on the collaborative state set, and mapping rule violation degrees into continuous costs according to the triggered navigation rule set to construct a navigation rule safety potential field; S5, constructing a human factor safety potential field based on fatigue, workload, attention or control delay indexes in the human factor state data; s6, carrying out safety margin correction on the static environment safety potential field, the dynamic interaction safety potential field and/or the navigation rule safety potential field according to the perception error, the execution error or the environment disturbance, or constructing an independent uncertainty correction field; Step S7, respectively normalizing the static environment safety potential field, the dynamic interaction safety potential field, the navigation rule safety potential field, the artificial safety potential field and/or the uncertainty correction field which is independently constructed into risk values of the same scale, and carrying out weighted fusion to obtain a total safety potential field; Step S8, predicting a motion track in a future time window based on the current motion state of the ship, integrating the total safety potential field along the predicted track to obtain a risk situation index, comparing the risk situation index with a preset threshold to determine a risk level, and determining a dominant risk source based on contributions of each potential field.
2. The ship safety potential modeling risk assessment method according to claim 1, wherein in step S2, the environmental data includes static obstacle and channel boundary information, and wherein in constructing the static environmental safety potential field, calculation is performed in a function form that decays with increasing distance based on a signed distance of an obstacle.
3. The ship safety potential modeling risk assessment method according to claim 1, wherein in step S3, the dynamic interaction safety potential is related to at least a normalized distance determined based on a target ship heading coordinate system, the normalized distance being calculated based on an elliptical domain or a super-elliptical domain in the target ship heading coordinate system, and a time factor based on a latest meeting time or a predicted minimum distance margin.
4. The ship safety potential modeling risk assessment method according to claim 1, wherein in step S4, the sailing rule safety potential field is constructed by mapping a rule violation degree to a continuous cost by a smoothing penalty function.
5. The method of modeling risk assessment of a marine safety potential field according to claim 1, wherein in step S7, the values of each potential field are mapped to risk values between zero and one using an exponential mapping function prior to weighted fusion.
6. A ship safety potential modeling risk assessment method according to claim 1, wherein in step S8, the risk situation index comprises at least an integrated value along a predicted trajectory and a maximum value within a predicted time window.
7. A ship safety potential field modeling risk assessment method according to claim 1, wherein in step S8, the dominant risk source is determined by calculating the contribution duty ratio of each fractional potential field along a predicted trajectory integral and arranging them side by side.
8. The ship safety potential modeling risk assessment method according to claim 1, wherein in step S6, the safety margin correction is achieved by expanding or scaling the distance of static obstacles or the normalized distance in the dynamic interaction safety potential.
9. A marine vessel safety potential field modeling risk assessment system, comprising: The data acquisition module is used for acquiring and synchronizing motion state data, environment data, navigation rule data and human factor state data of the ship and other ships so as to determine a collaborative state set at a moment t under a navigation coordinate system; The static potential field construction module is used for constructing a static environment safety potential field which changes along with the space position based on the environment data; The dynamic potential field construction module is used for constructing a dynamic interaction safety potential field for each target ship by utilizing the ship field model related to the direction and the relative motion information based on the motion state data; the navigation potential field construction module is used for identifying meeting situations based on the collaborative state set, and mapping rule violation degrees into continuous costs according to the triggered navigation rule set so as to construct a navigation rule safety potential field; The artificial potential field construction module is used for constructing an artificial safety potential field based on fatigue, workload, attention or control delay indexes in the artificial state data; the uncertainty correction module is used for carrying out safety margin correction on the static environment safety potential field, the dynamic interaction safety potential field and/or the navigation rule safety potential field according to the perception error, the execution error or the environment disturbance, or constructing an independent uncertainty correction field; The potential field fusion module is used for respectively normalizing the static environment safety potential field, the dynamic interaction safety potential field, the sailing rule safety potential field, the artificial safety potential field and/or the uncertainty correction field which is independently constructed into risk values of the same scale and carrying out weighted fusion to obtain a total safety potential field; the risk situation assessment module is used for predicting a motion track in a future time window based on the current motion state of the ship, integrating the total safety potential field along the predicted track to obtain a risk situation index, determining a risk level according to comparison of the risk situation index and a preset threshold value, and determining a dominant risk source based on contribution of each potential field.
10. The marine safety potential modeling risk assessment system of claim 9, wherein the dynamic interaction safety potential is related to at least a normalized distance determined based on a target marine heading coordinate system calculated based on an elliptical or super-elliptical domain in the target marine heading coordinate system and a time factor based on a most recent encounter time or a predicted minimum distance margin.

Description

Ship safety potential field modeling risk assessment method and system Technical Field The invention relates to the technical field of intelligent ships and maritime traffic safety, in particular to a ship safety potential field modeling risk assessment method and system. Background Along with the development of intelligent ships and autonomous navigation technologies, the perception and understanding capability of the ships to the surrounding marine traffic environment is continuously improved, but how to quantitatively evaluate navigation security risks and characterize risk situation evolution processes in complex and changeable marine environments is still a key problem for restricting the application of intelligent ships to real ships. Currently, a risk characterization method based on a field theory/potential field has been applied to the field of land traffic and ship traffic to some extent. Taking land traffic as an example, patent CN 117894180 describes the influence of static traffic environment, dynamic management and control and surrounding moving targets on the running risk of a vehicle by using a risk field, and realizes real-time risk assessment by overlapping the risk field, and patent CN 120599866 constructs a static obstacle potential energy field, a non-motor vehicle kinetic energy field and a driver behavior field in a slow-going shared space scene, and describes collision risk by using scalar risk field potential energy. In the field of ship traffic, patent CN 119642831 researches that a collision risk field is established through a ship field theory and a path is dynamically planned by combining a grid map and an ant colony algorithm, so that route optimization under collision risk constraint can be realized to a certain extent, but static environment and other ship collision risks are mainly focused, and the consideration of crew states and rule constraints is relatively less. CN 120910798 discloses an intelligent ship navigation risk evaluation energy field modeling method integrating multiple ship interaction characteristics, the method constructs a water traffic complexity index based on ship AIS data, further designs a navigation energy field and classifies the navigation scene in a risk manner by combining with an ALARP criterion, but the method mainly measures the risk from the angle of the traffic complexity of a ship group, and does not unify a human factor, a static environment and a navigation rule into a cooperative navigation safety potential field. In general, the prior art suffers from the following disadvantages: (1) The intelligent ship-oriented integrated navigation safety modeling framework is concentrated on a ship-environment or ship-ship level, and lacks a human-ship-environment-rule-oriented integrated navigation safety modeling framework for intelligent ships; (2) The rule factors are usually treated separately in the form of hard constraint or logic judgment, and lack of a rule safety potential field which can be uniformly overlapped and quantitatively compared with other risk sources; (3) The influence of human factors such as crew fatigue and workload on navigation safety is not included in the potential field modeling process, and the risk difference of the intelligent ship in different operation modes such as on-duty/off-duty remote control is difficult to reflect; (4) The existing method provides risk indexes at a certain moment or a certain position, and lacks a risk situation assessment method for aggregating risks along the ship prediction track along with time evolution. Therefore, it is necessary to propose a man-ship-environment-rule collaborative navigation safety potential field modeling and risk situation assessment method for intelligent ships to solve the above problems. Disclosure of Invention In order to achieve the purpose of the application, the application provides a ship safety potential field modeling risk assessment method, which comprises the following steps: step S1, acquiring and synchronizing motion state data, environment data, navigation rule data and human factor state data of a ship and other ships to determine a collaborative state set at a moment t under a navigation coordinate system; s2, constructing a static environment safety potential field changing along with the space position based on the environment data; s3, based on the motion state data, constructing a dynamic interaction safety potential field for each target ship by utilizing a ship field model related to the direction and relative motion information; s4, identifying meeting situations based on the collaborative state set, and mapping rule violation degrees into continuous costs according to the triggered navigation rule set to construct a navigation rule safety potential field; S5, constructing a human factor safety potential field based on fatigue, workload, attention or control delay indexes in the human factor state data; s6, carrying out safety margin correct