A Construction Method for a Dynamic Assessment Model of Water Traffic Risks

Abstract

The present invention belongs to the technical field of water traffic risk assessment, and discloses a method for constructing a dynamic risk assessment model for water traffic. By collecting water traffic elements, on the basis of analyzing the internal structural characteristics of the water traffic system, the establishment of the internal information set of the water traffic system is completed; based on the internal information flow of the water traffic system, the dynamic characteristics of water traffic risks are analyzed, and the risk operation indicators of water traffic are extracted; the risk operation indicators are used as Bayesian network nodes to determine the Bayesian network structure; then, the mutual information method is used to rank the correlation of the risk operation indicators and screen the Bayesian network nodes; finally, the Bayesian network node parameters are input to obtain the dynamic risk assessment model for water traffic. The present invention provides a method for establishing a real-time dynamic assessment model, which can provide theoretical and technical support for water traffic safety early warning and pre-control, and effectively improve water traffic safety and water traffic management efficiency.

Description

Technical Field

[0001] The present invention belongs to the technical field of water traffic risk assessment, and particularly relates to a method for constructing a dynamic assessment model of water traffic risk. Background Art

[0002] At present, with the development of ship specialization and large-scale, the density of ships in important waters and waterways is increasing, the water traffic environment is becoming increasingly complex, and water traffic accident risks such as ship collisions, groundings, fires, and leaks occur frequently. According to the national water traffic safety situation, accident risks occur frequently in complex waters such as the Three Gorges Reservoir Area, bridge areas, and busy waters. When ships navigate in complex waters, they are affected by factors such as waterway conditions, traffic flow, hydrometeorological conditions, and navigation obstacles, and there are relatively large water traffic safety risks. Traditional water traffic safety risk assessment methods such as the Analytic Hierarchy Process (AHP) and Fuzzy Comprehensive Evaluation (FCE) are better applicable to the sea, but the application of AHP and FCE in the identification and evaluation of water traffic safety risks in complex waters has great limitations. As a result, there are large deviations and lags in the early warning and pre-control of water traffic safety risks based solely on the evaluation results of AHP and FCE during the maritime supervision process, which causes the maritime supervision decision-making department to be unable to effectively identify the traffic safety risks in the supervised waters and reduces the energy efficiency of the maritime supervision department in traffic safety risk early warning. The early warning and pre-control of water traffic safety risks are key issues that need to be solved in the development process of ship-shore integrated management, and the dynamic evaluation method of traffic comprehensive risk applicable to complex waters is the bottleneck restricting the realization of early warning and pre-control.

[0003] In the prior art, the research on the water traffic safety system is not deep enough, and the research on the characteristics and structure of the water traffic safety system is lacking, resulting in the construction of existing water traffic risk assessment models mostly based on static indicators, and the dynamic and regional applicability of the evaluation models is poor. The structure of the water traffic system is abstract and the dynamic characteristics are complex. To establish a generally applicable water traffic risk evaluation model, it is necessary to deeply explore the internal causes and action laws of risk generation. Summary of the Invention

[0004] Aiming at the problems existing in the prior art, the present invention provides a method for constructing a dynamic assessment model of water traffic risk. Based on the information flow theory and Bayesian network, the present invention creatively cuts in from the perspective of internal information interaction of the human-ship-environment system, deeply explores the internal relationship between abnormal information interaction of the water traffic system and water traffic risk, and conducts research from three aspects: the internal structure characteristics and measurement of the water traffic system, the extraction of dynamic characteristics of water traffic risk, and the establishment of a dynamic evaluation model of water traffic system risk, forming a systematic theory and method for comprehensive dynamic evaluation of complex water area traffic, providing theoretical and technical support for early warning and pre-control of water traffic safety risks. The present invention has important theoretical and practical significance in breaking through the limitations of traditional AHP and FCE in static evaluation of water traffic safety risks, innovating water traffic risk theories and methods, promoting the cross-integration of water traffic engineering disciplines and systems science, and boosting ship-shore integrated intelligent management, etc.

[0005] The specific technical solution of the present invention is as follows:

[0006] A method for constructing a dynamic assessment model of water traffic risk, the method comprising:

[0007] 1) Collecting statistical data and investigation reports of ship accidents in a certain water area to obtain the elements of the water traffic system;

[0008] 2) Analyzing the internal structure characteristics of the elements of the water traffic system to establish an internal information set of the water traffic system;

[0009] 3) Analyzing abnormal scenarios of internal information flow of the water traffic system to extract dynamic operation indicators of water traffic system risk;

[0010] 4) Using the dynamic operation indicators of the risk as Bayesian network nodes to determine the Bayesian network structure, and then sorting the correlation of the dynamic operation indicators of the risk by the mutual information method and screening the Bayesian network nodes;

[0011] 5) Inputting the Bayesian network node parameters to obtain a dynamic assessment model of water traffic risk.

[0012] Furthermore, the elements of the water traffic system include:

[0013] Navigation environment elements: hydrology, meteorology, waterway conditions, hydraulic structures;

[0014] Traffic flow elements: waterway traffic capacity, ship traffic flow direction, ship traffic flow width, ship traffic flow speed, ship traffic flow density, ship traffic flow structure;

[0015] Elements of ship behavior: including ship elements and driver elements, where the ship elements include: ship type, ship tonnage, ship equipment, ship status, ship speed, and ship draft; the driver elements include: driver's professional quality, driver's age, driver's emergency response ability, driver's safety awareness, driver's awareness of rules, and driver's illegal scoring situation;

[0016] Social environment and management elements: national policies, shipping development, navigation technology, maritime management laws and regulations, navigation management systems, company internal management systems, and bridge management systems.

[0017] Furthermore, the internal structural features include:

[0018] The connections among the four elements of people, ships, environment, and management within the water traffic system;

[0019] The dynamics, multi-directionality, and spatio-temporal heterogeneity of the internal structure of the water traffic system;

[0020] Furthermore, the internal information set of the water traffic system includes: human-human interaction information, human-ship interaction information, ship-ship interaction information, and ship-shore interaction information.

[0021] Among them, the human-human interaction information includes bridge instructions, engine room instructions, handover information, and daily communication information; the human-ship interaction information includes radar information, electronic chart information, log information, AIS information, GPS information, compass information, cargo stowage information, steering instructions, automatic engine room equipment information, deck equipment information, fault information, and ship movement information; the ship-ship interaction information includes voyage information, cargo information, ship movement information, collision avoidance information, VHF information, and distress information; the ship-shore interaction information includes ship declaration, arrival at anchor confirmation, planned distribution information, dispatching instructions, ship real-time information, voyage plan, help information, navigational notice, and VTS supervision information.

[0022] Furthermore, the abnormal scenarios of internal information flow in the water traffic system include:

[0023] Information transmission delay, information blockage, information overflow, information error, information misunderstanding, and information concealment.

[0024] Furthermore, the analysis of abnormal scenarios of internal information flow in the navigation system includes the following steps:

[0025] The first step is to use the classification method to statistically analyze different forms of interaction information and analyze the impacts of navigation environment, navigation rules, crew quality, and management system factors on internal information interaction in the navigation system;

[0026] In the second step, complex network analysis method is adopted to obtain the information spatial distribution characteristics of the water traffic system, and to explore the generation, transformation, dissipation process and external triggering conditions of the uneven information distribution;

[0027] In the third step, the standard deviation measurement method is used to measure the unevenness of the information flow in the navigation system; the unevenness includes the information flow fluctuation amplitude, the information imbalance rate, and the information anomaly threshold;

[0028] In the fourth step, based on the water traffic information transfer map, the dynamic display and feature extraction of regional traffic anomaly scenarios are carried out.

[0029] Furthermore, the steps for extracting the dynamic operation indicators of navigation risks include the following: using wavelet transform and spectrum analysis to study the volatility characteristics of system risks and determine the volatility threshold of system risks; aiming at the differential characteristics of the spatial distribution of information flow, determining the risk spatial boundary; using the data distribution theory, combining the system safety mode and the uneven characteristics, and proposing risk operation indicators that characterize the dynamic characteristics of system risks.

[0030] Furthermore, the extraction of the dynamic operation indicators of navigation risks also includes:

[0031] Ranking of risk operation indicators based on the mutual information method, and the specific calculation method is:

[0032]

[0033] where I(X;Y) is the joint probability distribution of index X and index Y, and P(x) and P(y) respectively represent the marginal probability distributions of index X and index Y;

[0034] The larger the I(X;Y) value of the risk indicator, the stronger the correlation and the higher the ranking.

[0035] Furthermore, the screening of the Bayesian network nodes includes the following processing process:

[0036] For specific risk operation indicators, if the calculated I(X;Y) value is less than 0.05, then the risk operation indicator will be deleted, and the corresponding Bayesian network node will also be removed from the Bayesian network;

[0037] For specific risk operation indicators, if the calculated I(X;Y) value is greater than or equal to 0.05, then the risk operation indicator will be retained, and the corresponding Bayesian network node will also be retained in the Bayesian network.

[0038] Furthermore, in the Bayesian network structure, it also includes the optimization of the Bayesian network structure: on the basis of the screening of Bayesian network nodes, the Bayesian network structure is dynamically adjusted based on the network structure adaptive learning method.

[0039] The advantages and positive effects of the present invention are as follows: By constructing a dynamic evaluation model for water traffic risks, the present invention realizes the dynamic evaluation of the risks of the water traffic system. Starting from the internal structural characteristics of the water traffic system, by deeply exploring the abnormal interaction of information flows in the internal structure of the water traffic system, finding the internal relationship between the abnormal information interaction and water traffic risks, and extracting dynamic operation indicators of risks, a dynamic evaluation model is established in combination with a Bayesian network. Methodologically, the dynamic evaluation model for water traffic risks solves the problem of dynamic evaluation of water traffic risks and enriches the theory and methods of water traffic risks. In terms of application, the invention results can provide technical support for early warning and pre-control of water traffic safety risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] To more clearly illustrate the technical solutions of the embodiments of the present application, the following briefly introduces the drawings required to be used in the embodiments of the present application. Obviously, the following described drawings are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0041] Figure 1 It is a flowchart of a method for constructing a dynamic evaluation model for water traffic risks provided by an embodiment of the present invention.

[0042] Figure 2 It is a schematic diagram of an index system for ship navigation elements provided by an embodiment of the present invention.

[0043] Figure 3 It is a schematic diagram of an index system for the internal information set of the total navigation system provided by an embodiment of the present invention.

[0044] Figure 4 It is a schematic diagram of a Bayesian network model for water traffic risk assessment provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0045] In order to make the objectives, technical solutions and advantages of the present invention clearer, the following further details the present invention in combination with embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0046] The present invention conducts a detailed statistics and analysis on the historical data of water traffic accidents and ship traffic in the backwater change area of the Three Gorges Reservoir from 2010 to 2019, obtains the water traffic elements in the backwater change area of the Three Gorges Reservoir, establishes a dynamic operation index system for risks, proposes a method for constructing a risk evaluation model based on a Bayesian network, and evaluates the water traffic risks in the backwater change area of the Three Gorges Reservoir. As Figure 1 shown, the specific implementation steps are as follows:

[0047] S101: Obtain data and information on elements of the waterborne transportation system such as the navigable environment, traffic flow, ship behavior, social environment, and management in the research waters by using methods of traffic surveys, historical data collection, and expert consultation.

[0048] S102: Combine the internal structural characteristics of the waterborne transportation system in the research waters to establish an internal information set of the waterborne transportation system; the internal information set in the backwater fluctuation area of the Three Gorges Reservoir is shown in Table 1:

[0049] Table 1 Internal information set of the traffic system in the backwater fluctuation area of the Three Gorges Reservoir

[0050]

[0051] S103: Analyze the waterborne transportation accident data in the backwater fluctuation area of the Three Gorges Reservoir from 2010 to 2019, and extract dynamic operation indicators of the waterborne transportation system risk based on abnormal information flow, as shown in Table 2;

[0052] Table 2 Status parameter table of dynamic operation indicators of ship navigation risk

[0053]

[0054]

[0055] S104: Use the risk operation indicators in the backwater fluctuation area of the Three Gorges Reservoir as Bayesian network nodes to determine the Bayesian network structure; then use the mutual information method to rank the correlations of the risk operation indicators and screen the Bayesian network nodes; when the navigation risk is high, the status ranking of the risk operation indicators is shown in Table 3.

[0056] Table 3 Ranking table of the influence of node status on ship navigation when "navigation risk = high"

[0057]

[0058]

[0059] S105: By inputting the parameters of the Bayesian network nodes, obtain a dynamic assessment model for waterborne transportation risk, and the finally established Bayesian network model for the backwater fluctuation area of the Three Gorges Reservoir is as Figure 4 shown.

[0060] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any modification, equivalent replacement, and improvement made by those skilled in the art within the technical scope disclosed by the present invention, as long as they are made within the spirit and principle of the present invention, shall be covered by the protection scope of the present invention.

Claims

1. A method for constructing a risk dynamic assessment model for water traffic, characterized in that, The method includes the following steps: 1) Collect statistical data and investigation reports of ship accidents in a certain water area to obtain the elements of the water traffic system; 2) Analyze the internal structural characteristics of the water traffic system elements and establish an internal information set of the water traffic system; 3) Analyze the abnormal scenarios of information flow in the internal information set of the water traffic system and extract the dynamic operation indicators of the water traffic system risks; 4) Use the dynamic operation indicators of the risks as the nodes of the Bayesian network, determine the Bayesian network structure, and then sort the correlations of the dynamic operation indicators of the risks by the mutual information method and screen the nodes of the Bayesian network; 5) Input the parameters of the Bayesian network nodes to obtain a dynamic evaluation model for water traffic risks, wherein, the analysis of the abnormal scenarios of information flow in the navigation system includes the following steps: The first step is to statistically analyze different forms of interactive information by the classification method and analyze the impacts of navigation environment, navigation rules, crew quality, and management system factors on the internal information interaction of the navigation system; The second step is to obtain the spatial distribution characteristics of the water traffic system information by using the complex network analysis method and explore the generation, transformation, dissipation processes, and external triggering conditions of the uneven information distribution; The third step is to measure the unevenness of the information flow in the navigation system by the standard deviation measurement method; the unevenness indicators include the fluctuation amplitude of the information flow, the information imbalance rate, and the information anomaly threshold; The fourth step is to dynamically display and extract the characteristics of regional traffic abnormal scenarios based on the water traffic information transmission map.

2. The method for constructing a risk dynamic assessment model for an aquatic transportation system according to claim 1, wherein The elements of the water traffic system include: Navigation environment elements: hydrology, meteorology, waterway conditions, and hydraulic structures; Traffic flow elements: waterway traffic capacity, direction of ship traffic flow, width of ship traffic flow, speed of ship traffic flow, density of ship traffic flow, and structure of ship traffic flow; Ship behavior elements: including ship elements and driver elements. The ship elements include: ship type, ship tonnage, ship equipment, ship status, ship speed, and ship draft; the driver elements include: professional quality of the driver, age of the driver, emergency response ability of the driver, safety awareness of the driver, rule awareness of the driver, and driver's illegal score; Social environment and management elements: national policies, shipping development, navigation technology, maritime management laws and regulations, navigation management system, company internal management system, and bridge management system.

3. The method for constructing a risk dynamic assessment model for water traffic according to claim 1, characterized in that, The internal structural characteristics include: The connections among the four elements of people, ships, environment, and management in the water traffic system; The dynamics, multi-directionality, and spatio-temporal heterogeneity of the internal structure of the water traffic system; Furthermore, the internal information set of the water traffic system includes: person-person interaction information, person-ship interaction information, ship-ship interaction information, and ship-shore interaction information.

4. The method for constructing a risk dynamic assessment model for waterborne transportation according to claim 1, wherein The internal information set of the water traffic system includes: person-person interaction information, person-ship interaction information, ship-ship interaction information, and ship-shore interaction information; The person-person interaction information includes bridge instructions, engine orders, handover information, and daily communication information; The human-ship interaction information includes radar information, electronic chart information, log information, AIS information, GPS information, compass information, cargo stowage information, steering commands, automatic engine room equipment information, deck equipment information, fault information, and ship motion information; The ship-ship interaction information includes voyage information, cargo information, ship motion information, collision avoidance information, VHF information, and distress information; The ship-shore interaction information includes ship declaration, arrival at anchor confirmation, planned distribution information, dispatching instructions, real-time ship information, voyage plan, help request information, navigational notice, and VTS supervision information.

5. The method for constructing a risk dynamic assessment model for waterborne traffic according to claim 1, wherein The abnormal scenarios of internal information flow in the water traffic system include: Information transmission delay, information blockage, information overflow, information error, information misunderstanding, and information concealment.

6. The method for constructing a risk dynamic assessment model for water traffic according to claim 1, characterized in that, The extraction of dynamic operation indicators for navigation risks includes the following steps: using wavelet transform and spectral analysis to study the volatility characteristics of system risks and determine the system risk volatility threshold; determining the risk space boundary according to the differential characteristics of the spatial distribution of information flows; using data distribution theory and combining the system safety mode and imbalance characteristics to propose risk operation indicators that characterize the dynamic characteristics of system risks.

7. The construction method of the risk dynamic assessment model for waterborne traffic according to claim 1, characterized in that In the Bayesian network structure, it also includes: optimization of the Bayesian network structure: based on the screening of Bayesian network nodes, the Bayesian network structure is dynamically adjusted based on the network structure adaptive learning method.

Images