CN-121999638-A - Ship abnormal operation prediction method, system, equipment and storage medium

CN121999638ACN 121999638 ACN121999638 ACN 121999638ACN-121999638-A

Abstract

The application relates to a ship abnormal operation prediction method, a system, equipment and a storage medium, wherein the method comprises the steps of acquiring real-time operation parameters and real-time environment data of a ship; the method comprises the steps of constructing a digital twin body of a ship, training a reinforcement learning model in the digital twin body, optimizing a safety decision strategy of the digital twin body, acquiring a simulated track of a deducted ship based on the optimized digital twin body according to real-time operation parameters and real-time environment data of the ship, predicting a future operation track of the ship according to the real-time operation parameters of the ship by adopting a preset time sequence prediction model to acquire a predicted track, comparing the simulated track with the predicted track in a deviation mode, and generating a ship abnormal operation prediction result according to a comparison result. The method provided by the application can realize high-precision active prediction and self-adaptive decision, and obviously improve the safety and intelligent level of ship navigation.

Inventors

LUO FUQIANG
SONG ZHIJIE
CHANG XIN

Assignees

广州航海学院

Dates

Publication Date: 20260508
Application Date: 20260129

Claims (10)

1. A ship abnormal operation prediction method, comprising: acquiring real-time operation parameters and real-time environment data of a ship; Constructing a digital twin body of the ship; Training a reinforcement learning model in the digital twin, and optimizing a security decision strategy of the digital twin; based on the optimized digital twin body, acquiring a simulated track of the deducted ship according to the real-time operation parameters and the real-time environment data of the ship; predicting a future running track of the ship by adopting a preset time sequence prediction model according to the real-time running parameters of the ship to obtain a predicted track; and carrying out deviation comparison on the simulated track and the predicted track, and generating a ship abnormal operation prediction result according to the comparison result.
2. The ship abnormal operation prediction method according to claim 1, wherein the constructing a digital twin body of a ship comprises: Acquiring historical operation data, historical environment parameters and ship parameters of a ship; Preprocessing the historical operation data and the historical environment parameters; based on the preprocessed historical operation data and the historical environmental parameters and the ship parameters, a preset multi-body dynamics model is adopted to construct a kinematic model and a dynamics model of the ship; constructing an environment model according to the historical environment parameters; and combining the kinematic model, the dynamic model and the environment model to construct a digital twin body of the ship.
3. The method of claim 1, wherein training a reinforcement learning model in the digital twin optimizes a security decision strategy of the digital twin, comprising: deploying a preset reinforcement learning agent in the digital twins; Setting the real-time operation parameters and the real-time environment data of the ship as a state space of the reinforcement learning model, wherein the state space comprises longitude and latitude, navigational speed, heading, surrounding ship density and wind speed of the ship; Setting the safety decision of the ship as the action space of the reinforcement learning model; Training the model based on a preset reward function, optimizing the model by adopting the reinforcement learning agent, and outputting an optimized safety decision strategy when the training result meets a preset convergence condition.
4. The method for predicting abnormal operation of a ship according to claim 1, wherein the obtaining, based on the optimized digital twin body, a simulated trajectory of the deduced ship according to real-time operation parameters and real-time environmental data of the ship comprises: Inputting the real-time operation parameters and the real-time environment data into the digital twin body for simulation, and adjusting the operation state of the ship based on the optimized safety decision strategy; Based on the adjusted running state, obtaining the simulated track of the deducted ship.
5. The method for predicting abnormal operation of a ship according to claim 1, wherein the comparing the deviation of the simulated trajectory and the predicted trajectory, and generating the predicted abnormal operation of the ship according to the comparison result, comprises: Calculating position deviation, course deviation and speed deviation between the simulated track and the predicted track according to the simulated track and the predicted track; comparing the position deviation, the heading deviation and the speed deviation with corresponding preset thresholds respectively; And generating an abnormal operation prediction result of the ship according to the threshold comparison result.
6. The method for predicting abnormal operation of a vessel according to claim 5, further comprising, after generating the predicted abnormal operation of the vessel based on the threshold comparison result, the steps of: identifying abnormal risk levels of the ship according to the ship abnormal operation prediction result; and carrying out early warning intervention on the ship by adopting a preset abnormal early warning strategy according to the abnormal risk level.
7. The ship abnormal operation prediction method according to claim 6, wherein the abnormal risk class is classified into three risk classes; the identifying the abnormal risk level of the ship according to the ship abnormal operation prediction result comprises the following steps: when the position deviation and/or the heading deviation of the ship meet a first abnormal risk condition, identifying that the abnormal risk level of the ship is a medium risk level; And when the speed deviation of the ship meets a second abnormal risk condition, identifying the abnormal risk level of the ship as a high risk level.
8. A ship abnormal operation prediction system, comprising: The data acquisition module is used for acquiring real-time operation parameters and real-time environment data of the ship; The digital twin body construction module is used for constructing a digital twin body of the ship; the digital twin training module is used for training a reinforcement learning model in the digital twin and optimizing a safety decision strategy of the digital twin; the simulation track acquisition module is used for acquiring a simulation track of the deducted ship based on the optimized digital twin body according to the real-time running parameters and the real-time environment data of the ship; The predicted track acquisition module is used for predicting the future running track of the ship by adopting a preset time sequence prediction model according to the real-time running parameters of the ship to acquire a predicted track; And the abnormal operation prediction module is used for carrying out deviation comparison on the simulated track and the predicted track, and generating a ship abnormal operation prediction result according to the comparison result.
9. A computer device comprising a processor, a memory and a computer program stored in the memory and executable on the processor, characterized in that the processor implements the steps of the ship abnormal operation prediction method according to any one of claims 1 to 7 when the computer program is executed.
10. A computer-readable storage medium storing a computer program, wherein the computer program when executed by a processor implements the steps of the ship abnormal operation prediction method according to any one of claims 1 to 7.

Description

Ship abnormal operation prediction method, system, equipment and storage medium Technical Field The invention relates to the technical field of intelligent navigation, in particular to a ship abnormal operation prediction method, a system, equipment and a storage medium. Background With the vigorous development of the global shipping industry, ship navigation safety is becoming an increasingly important concern for maritime management. Abnormal operation behaviors, such as track deviation, navigational speed abnormality and the like, which are main causes of accidents, such as offshore collision, stranding and the like, can be caused by mechanical faults, misoperation or severe environments and the like in the navigation process of the ship. Therefore, the ship abnormal operation prediction technology has great significance for guaranteeing the safety of life and property of the sea and improving shipping efficiency. However, in the prior art, the ship is monitored through a single model, so that the ship monitoring requirement in complex and changeable navigation scenes cannot be met. Disclosure of Invention Based on the above, the invention aims to provide a ship abnormal operation prediction method, a system, equipment and a storage medium, which are used for realizing high-precision active prediction and self-adaptive decision by combining a simulation track obtained by deduction of a ship digital twin body and a comprehensive comparison result of a prediction track obtained by prediction of a time sequence prediction model, and remarkably improving the safety and the intelligent level of ship navigation. In a first aspect, the present application provides a method for predicting abnormal operation of a ship, including: acquiring real-time operation parameters and real-time environment data of a ship; Constructing a digital twin body of the ship; Training a reinforcement learning model in the digital twin, and optimizing a security decision strategy of the digital twin; based on the optimized digital twin body, acquiring a simulated track of the deducted ship according to the real-time operation parameters and the real-time environment data of the ship; predicting a future running track of the ship by adopting a preset time sequence prediction model according to the real-time running parameters of the ship to obtain a predicted track; and carrying out deviation comparison on the simulated track and the predicted track, and generating a ship abnormal operation prediction result according to the comparison result. In a second aspect, the present application provides a ship abnormal operation prediction system, comprising: The data acquisition module is used for acquiring real-time operation parameters and real-time environment data of the ship; The digital twin body construction module is used for constructing a digital twin body of the ship; the digital twin training module is used for training a reinforcement learning model in the digital twin and optimizing a safety decision strategy of the digital twin; the simulation track acquisition module is used for acquiring a simulation track of the deducted ship based on the optimized digital twin body according to the real-time running parameters and the real-time environment data of the ship; The predicted track acquisition module is used for predicting the future running track of the ship by adopting a preset time sequence prediction model according to the real-time running parameters of the ship to acquire a predicted track; And the abnormal operation prediction module is used for carrying out deviation comparison on the simulated track and the predicted track, and generating a ship abnormal operation prediction result according to the comparison result. In a third aspect, the present application provides a computer device comprising a processor, a memory and a computer program stored in the memory and operable on the processor, the processor executing the computer program to perform the steps of the method for predicting abnormal operation of a vessel as described above. In a fourth aspect, the present application provides a computer readable storage medium storing a computer program which when executed by a processor implements the steps of a ship abnormal operation prediction method as described above. The ship abnormal operation prediction method, the system, the equipment and the storage medium are characterized by constructing a digital twin body of a ship, training a reinforcement learning model in the digital twin body based on real-time operation parameters and environment data to optimize a safety decision strategy of the digital twin body, deducing a simulation track of the ship by using the optimized digital twin body, simultaneously predicting a future operation track of the ship by adopting a time sequence prediction model to obtain a prediction track, and finally comparing the simulation track with the prediction track in a multi-dimensional