Dynamic evaluation method for ship navigation risk in arctic navigation channel based on dynamic Bayesian network

Abstract

The invention provides an arctic navigation channel ship navigation risk dynamic assessment method based on a dynamic Bayesian network, and belongs to the technical field of ship navigation safety. The arctic navigation channel ship navigation risk factors are identified from three aspects of environment, people and ship; establishing an arctic navigation channel ship navigation risk grade standard; constructing a static Bayesian network graph model of the ship navigation risk of the arctic navigation channel; determining parameters of a static Bayesian network model, and constructing a dynamic Bayesian network of the ship navigation risk of the arctic navigation channel; dynamically evaluating the ship navigation risk of the arctic navigation channel based on a dynamic Bayesian network; according to the method, environment, ship and human factors are considered, years of marine meteorological reanalysis data and expert knowledge are combined, updating can be carried out according to past, current and future conditions, and the risk of ice trapping and ship-ice collision of the arctic waterway is dynamically evaluated.

Description

Technical Field

[0001] The present invention belongs to the technical field of ship navigation safety, and in particular, relates to a method for dynamically assessing the risk of ship navigation in an Arctic waterway based on a dynamic Bayesian network. Background Art

[0002] With the reduction of sea ice area and the improvement of navigation conditions, more and more commercial ships use the Northeast Passage of the Arctic as their route. However, the complex ice conditions, extremely low temperatures, remote locations and lack of infrastructure conditions in the Arctic region pose great challenges to the safety of ship navigation. Ships sailing in the Arctic may face two main accident scenarios - ice entrapment and ship-ice collision. The risk of such accidents has the characteristics of dynamic evolution. At present, the main methods for assessing the risk of ship navigation in the Arctic region include fault tree analysis, event tree analysis, Bayesian network, bowtie model, hierarchical analysis method, etc. Many scholars have developed risk assessment models based on these methods to predict the probability of marine accidents for ships sailing in Arctic waters by combining subjective and objective data. However, ship navigation is a continuous process, and navigation risks show dynamic characteristics that evolve with environmental changes and the ship's operation process. However, most existing studies only consider the static probability of risk, without considering the correlation and complementarity of information that changes over time. Summary of the invention

[0003] In view of the above-mentioned defects, the present invention proposes a dynamic assessment method for ship navigation risks in the Arctic route based on a dynamic Bayesian network, which takes into account environmental, ship and human factors, and combines many years of marine meteorological reanalysis data and expert knowledge. It can be updated according to past, present and future conditions, and dynamically assess the risks of ice trapping and ship-ice collision in the Arctic route.

[0004] The present invention is achieved through the following technical solutions:

[0005] A dynamic assessment method for ship navigation risk in the Arctic route based on dynamic Bayesian network:

[0006] The method specifically comprises the following steps:

[0007] Step 1: Identify the risk factors affecting the navigation of ships in the Arctic route from the three aspects of environment, people and ships;

[0008] Step 2: Establish risk level standards for ships sailing in the Arctic route;

[0009] Step 3: Construct a static Bayesian network graph model of the risk of ship navigation in the Arctic route;

[0010] Step 4: Determine the static Bayesian network model parameters;

[0011] Step 5: Construct a dynamic Bayesian network for ship navigation risks in the Arctic route;

[0012] Step 6: Dynamically evaluate the navigation risks of ships in the Arctic route based on the dynamic Bayesian network.

[0013] Further, in step 1, the risk factors include environmental factors B1, human factors B2, and ship factors B3;

[0014] The corresponding indicators of environmental factor B1 are: wind N1, temperature N2, visibility N3, waves N4, sea ice density N5 and sea ice thickness N6;

[0015] The corresponding indicators of human factor B2 are: insufficient experience in ice navigation N7, insufficient professional skills N8, fatigue N9, improper route selection N 10 、Insufficient information 11 , Lookout FailureN 12 and the risk is underestimated 13 ;

[0016] The corresponding index of ship factor B3 is: speed N 14 、Operation equipment failure N 15 、Irrational loading N 16 、Ship ice class N 17 、Navigation equipment failure N 18 、Digital charts are wrong 19 .

[0017] Furthermore, in step 2, the risks of ship navigation in the Arctic route are divided into ice entrapment risk and ship-ice collision risk, and the risk level is determined by the probability of risk occurrence and the severity of risk consequences.

[0018] Furthermore, in step 3, according to the causal relationship of each risk factor, the hierarchical analysis diagram is converted into a directed acyclic graph, a static Bayesian network model is constructed, and the supplementary node ice condition N is introduced. 20 ;

[0019] In the Bayesian network model, if an arc points from node A to node B, A is considered to be the parent node and B is the child node. A node without a parent node is called a root node.

[0020] Furthermore, in step 4,

[0021] Step 4.1: Classify the identified risk factors into states;

[0022] Step 4.2: Determine the prior probability of all root nodes and the conditional probability table of child nodes and input them into the Bayesian network; the data comes from satellite remote sensing monitoring data of environmental factors such as meteorology and hydrology and expert experience knowledge;

[0023] The prior probability data of the root nodes “wind N1”, “temperature N2”, “visibility N3”, “sea ice density N5” and “sea ice thickness N6” are derived from meteorological and hydrological satellite remote sensing monitoring data, “lack of experience in ice navigation N7”, “lack of professional skills N8”, “unreasonable loading N9” and “inadequate navigation experience N10”. 16 " and "The digital chart is wrong 19 The prior probability of " comes from expert judgment. The conditional probability table data of sub-node wave N4 comes from meteorological and hydrological satellite remote sensing monitoring data, and the conditional probability table data of other sub-nodes comes from expert experience knowledge.

[0024] Assume that the node variable N i There are n states, and there are m parent nodes associated with them, denoted as parents j (N i ), parents(N i ) represents the vector composed of all parent node variables of the variable, σ represents the state of the variable, and γ represents the vector parents(N i ), the conditional probability of the variable can be calculated as follows:

[0025]

[0026] Furthermore, in step 5,

[0027] Step 5.1: Construct a static Bayesian network with the navigation risk of ships in the Arctic route as the target node through steps 1 to 3;

[0028] Step 5.2: Identify the nodes in the constructed static Bayesian network whose states change over time, and record them as dynamic nodes;

[0029] Step 5.3: Obtain the time interval for the risk assessment of ship navigation in the Arctic route, and divide the time interval into T time slices according to the number of days. Each time slice corresponds to a static Bayesian network with the same structure. Connect the dynamic nodes of the previous time slice with the next time slice to form a dynamic Bayesian network.

[0030] Step 5.4: Calculate the state transfer matrix of dynamic nodes between different time slices based on Markov chain; Assume that the dynamic node N has three states, namely: State1, State2 and State3; Formula (2) gives the state distribution M of N at time t t , the three states satisfy formula (3); as time goes by, the state of N in each time slice transfers between the three states, forming a Markov chain; η, γ represents the probability of transition between any two states; Equations (4)-(6) represent the calculation process of the probability of occurrence of the three states at time t+1; the state transition matrix A of N is shown in Equation (7);

[0031] M t =(S1(t),S2(t),S3(t)) (3)

[0032] S1(t)+S2(t)+S3(t)=1 (4)

[0033]

[0034] Further, in step 6,

[0035] Step 6.1: Input the calculated state transfer matrix of dynamic nodes into the dynamic Bayesian network, and obtain the navigation risk R of ships in the Arctic route through automatic updating. t The risk probability that changes over time enables preliminary reasoning about navigation risks when there is no evidence input;

[0036] Step 6.2: When the status of a risk factor node in the past, present and future is known, it is input into the dynamic Bayesian network as evidence to achieve real-time update of navigation risk.

[0037] An evaluation system for the dynamic evaluation method of ship navigation risk in the Arctic route based on dynamic Bayesian network:

[0038] The evaluation system includes a risk factor identification module, a risk level standard evaluation module, a static-dynamic Bayesian network construction module and a dynamic evaluation module;

[0039] The risk factor identification module identifies risk factors affecting the navigation of ships in the Arctic route from three aspects: environment, people and ships;

[0040] The risk level standard assessment module establishes the risk level standard for ships sailing in the Arctic waterway;

[0041] The static-dynamic Bayesian network construction module first constructs a static Bayesian network graph model of the risk of ship navigation in the Arctic waterway, and then determines the parameters of the static Bayesian network model; finally, a dynamic Bayesian network of the risk of ship navigation in the Arctic waterway is constructed based on the parameters of the static Bayesian network model;

[0042] The dynamic assessment module dynamically assesses the navigation risks of ships in the Arctic waterway.

[0043] An electronic device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0044] A computer-readable storage medium is used to store computer instructions, and when the computer instructions are executed by a processor, the steps of the above method are implemented.

[0045] Beneficial effects of the present invention

[0046] 1. The dynamic assessment method for the navigation risk of ships in the Arctic route of the present invention comprehensively considers the environmental, ship and human factors, and combines many years of marine meteorological reanalysis data and expert knowledge. The method can effectively reflect the distribution of ship navigation risks in various sea areas and has reference significance for the safe navigation management of ships in the Arctic route.

[0047] 2. The method of the present invention combines a dynamic Bayesian network, which can be updated according to past, present and future conditions. It has the ability to update in real time, which improves the accuracy of the risk assessment of ships sailing in the Arctic route.

[0048] 3. The method for dynamic assessment of the risk of ship navigation in the Arctic route based on dynamic Bayesian network of the present invention has established and identified the risk factors affecting the ice entrapment and ship-ice collision of ships in the Arctic route, and provided the steps for dynamic assessment of the risk of ship navigation in the Arctic route, the risk classification level standard, the static Bayesian network of ship navigation risk and the dynamic Bayesian network of ship navigation risk, which can dynamically assess the risk of ship navigation in the Arctic route. The method can effectively reflect the change of the probability of ship navigation risk, and has the ability to update in real time, and can be used as a tool to guide ship operations and navigation decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 Schematic diagram of the dynamic assessment method for ship navigation risks in the Arctic route based on dynamic Bayesian networks.

[0050] Figure 2 Factors affecting the navigation risks of ships in the Arctic route.

[0051] Figure 3 It is the risk level standard for ships sailing in the Arctic route.

[0052] Figure 4 Static Bayesian network for ship navigation risk in Arctic shipping routes.

[0053] Figure 5 Dynamic Bayesian network for ship navigation risk in Arctic shipping routes.

[0054] Figure 6 This is a chart showing the changing probability of risk for ships sailing in the East Siberian Sea on the Arctic route.

[0055] Figure 7 This is a chart showing the changing probability of risk for ships sailing in the East Siberian Sea on the Arctic route. DETAILED DESCRIPTION

[0056] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0057] The experimental methods used in the following examples are conventional methods unless otherwise specified. The materials, reagents, methods and instruments used are conventional materials, reagents, methods and instruments in the art unless otherwise specified, and can be obtained through commercial channels by those skilled in the art.

[0058] A dynamic assessment method for ship navigation risk in the Arctic route based on dynamic Bayesian network:

[0059] like Figure 1 As shown, the method specifically comprises the following steps:

[0060] Step 1: Identify the risk factors affecting the navigation of ships in the Arctic route from the three aspects of environment, people and ships; Figure 2 As shown in the figure, environmental factors, human factors and ship factors are selected as risk assessment indicators for ship navigation in the Arctic route, and 19 risk factors that affect ship navigation are identified;

[0061] In step 1, the risk factors include environmental factors B1, human factors B2, and ship factors B3;

[0062] The corresponding indicators of environmental factor B1 are: wind N1, temperature N2, visibility N3, waves N4, sea ice density N5 and sea ice thickness N6;

[0063] The corresponding indicators of human factor B2 are: insufficient experience in ice navigation N7, insufficient professional skills N8, fatigue N9, improper route selection N 10 、Insufficient information 11 , Lookout FailureN 12 and the risk is underestimated 13 ;

[0064] The corresponding index of ship factor B3 is: speed N 14 、Operation equipment failure N 15 、Irrational loading N 16 、Ship ice class N 17 、Navigation equipment failure N 18 、Digital charts are wrong 19 .

[0065] Step 2: Establish risk level standards for Arctic route ship navigation; divide the Arctic route ship navigation risks into ice entrapment risks and ship-ice collision risks, and the risk level is determined by the probability of risk occurrence and the severity of the risk consequences.

[0066] In this embodiment, the probability of risk occurrence is expressed in five levels: frequent, possible, accidental, less likely, and impossible. The risk loss level standard is divided into five levels according to the severity of the loss: catastrophic, severe, significant, slight, and negligible. According to the probability and consequences of risk occurrence, the ice entrapment risk and ship-ice collision risk are divided into three levels: low risk, medium risk, and high risk. Figure 3 shown.

[0067] Step 3: Construct a static Bayesian network graph model of the risk of ship navigation in the Arctic route;

[0068] In step 3, according to the causal relationship of each risk factor, the hierarchical analysis diagram is transformed into a directed acyclic graph to construct a static Bayesian network model.

[0069] Through logical analysis, it is found that some factors are affected by both sea ice density and sea ice thickness. In order to simplify the model structure and reduce the workload of experts, an intermediate variable "ice condition" N is introduced into the model. 20 As the parent node of sea ice density and sea ice thickness. The static Bayesian network graph model of the risk of ship navigation in the Arctic route is as follows: Figure 4 shown.

[0070] In the Bayesian network model, if an arc points from node A to node B, A is considered to be the parent node and B is the child node. A node without a parent node is called a root node.

[0071] Step 4: Determine the static Bayesian network model parameters;

[0072] Step 4.1: Classify the identified risk factors into states; set the environmental factors "wind N1", "temperature N2", "visibility N3", "wave N4", "sea ice concentration N5" and "sea ice thickness N6" into three states;

[0073] Human factors include “lack of experience in ice navigation N7”, “lack of professional skills N8”, “fatigue N9”, “improper route selection N10”, “inadequate navigation experience N11”, “inadequate navigation experience N12”, “inadequate navigation experience N13”, “inadequate navigation experience N14”, “inadequate navigation experience N15”, “inadequate navigation experience N16”, “inadequate navigation experience N17” 10 "Insufficient information 11 ", "Lookout failed N 12 " and "The risk is underestimated 13 "、Ship factor" Speed ​​N 14 ", "Operation equipment failure N 15 ", " "Improper loading N 16","Ship Ice Class N 17 ","Navigation equipment failure N 18 ", "Digital charts are wrong 19 " and the supplementary node "Ice Condition N 20 " is set to two states, as shown in Table 1.

[0074] Table 1 Node status classification

[0075]

[0076] Step 4.2: Determine the prior probability of all root nodes and the conditional probability table of child nodes and input them into the Bayesian network; the data comes from satellite remote sensing monitoring data of environmental factors such as meteorology and hydrology and expert experience knowledge;

[0077] The prior probability data of the root nodes “wind N1”, “temperature N2”, “visibility N3”, “sea ice density N5” and “sea ice thickness N6” are derived from meteorological and hydrological satellite remote sensing monitoring data, “lack of experience in ice navigation N7”, “lack of professional skills N8”, “unreasonable loading N9” and “inadequate navigation experience N10”. 16 " and "The digital chart is wrong 19 The prior probability of " comes from expert judgment.

[0078] The conditional probability table data of sub-node wave N4 comes from meteorological and hydrological satellite remote sensing monitoring data, and the conditional probability table data of other sub-nodes comes from expert experience knowledge.

[0079] According to statistics, the navigation period of the Northeast Passage in the Arctic is generally from July to October, so the present invention collects the meteorological and hydrological satellite remote sensing monitoring data of the Northeast Passage in the Arctic from July to October 2013 to 2022 for calculation. Among them, the wind speed, wave height, sea ice density and temperature data come from the European Center for Medium-Term Weather Forecasts, and the sea ice thickness data uses the PIOMAS model grid data released by the Polar Science Center. The present invention uses the method of expert survey to collect subjective data from experts in this field to make up for the lack of objective data. It collects the opinions of a total of 10 experts engaged in this research field from China Ocean Shipping Group Co., Ltd., China Ship Science Research Center, Dalian University of Technology, etc.

[0080] Assume that the node variable N i There are n states, and there are m parent nodes associated with them, denoted as parents j (N i ), parents(N i ) represents the vector composed of all parent node variables of the variable, σ represents the state of the variable, and γ represents the vector parents(N i ), the conditional probability of the variable can be calculated by the following formula:

[0081]

[0082] The present invention takes the positions, educational background, working experience and working experience in polar shipping research of the experts participating in the survey as the main basis for judging the comprehensive level of the experts, constructs a weight evaluation index system for survey experts (see Table 2), obtains the weight coefficient of each expert according to formula (2), and takes the weighted average of the prior probability and conditional probability of the 10 experts' judgment as the model input.

[0083]

[0084] Table 2 Survey expert weight evaluation index system

[0085]

[0086] Step 5: Construct a dynamic Bayesian network for ship navigation risks in the Arctic route;

[0087] Step 5.1: Through steps 1 to 4, a static Bayesian network with the navigation risk of ships in the Arctic route as the target node is constructed, such as Figure 5 As shown;

[0088] Step 5.2: Identify the nodes in the constructed static Bayesian network whose states change over time, and record them as dynamic nodes;

[0089] A dynamic node is a node whose probability distribution changes over time. In the present invention, all variables related to the environment are considered dynamic nodes, including "temperature N2", "visibility N3", "waves N4", "sea ice density N5" and "sea ice thickness N6".

[0090] Step 5.3: Obtain the time interval for the risk assessment of ship navigation in the Arctic route, and divide the time interval into T time slices according to the number of days. Each time slice corresponds to a static Bayesian network with the same structure. Connect the dynamic nodes of the previous time slice with the next time slice to form a dynamic Bayesian network.

[0091] The time interval for the risk assessment of the Arctic route ships in this invention is 7 days. The proposed model has 7 time slices. The established dynamic Bayesian network is as follows: Figure 5 As shown in the figure. Each dynamic node has an arc pointing to itself, indicating that the risk indicator of the node in a certain time slice will have an impact on the next time slice, and this impact conforms to the Markov process. The single digit on the arc indicates the time delay of the effect, and the arc marked as 1 indicates that the effect requires one time step.

[0092] Step 5.4: Calculate the state transfer matrix of dynamic nodes between different time slices based on Markov chain; Assume that the dynamic node N has three states, namely: State1, State2 and State3; Formula (3) gives the state distribution M of N at time t t , the three states satisfy formula (4);

[0093] As time goes by, the state of N in each time slice transfers between three states, forming a Markov chain; η, γ represents the probability of transition between any two states; Equations (5)-(7) represent the calculation process of the probability of occurrence of the three states at time t+1; the state transition matrix A of N is shown in Equation (8);

[0094]

[0095]

[0096]

[0097] Step 6: Conduct a dynamic assessment of the risks of ship navigation in the Arctic route.

[0098] Step 6.1: Input the calculated state transfer matrix of the dynamic node into the dynamic Bayesian network, and obtain the risk probability of the Arctic route ship navigation risk Rt changing with time through automatic updating, so as to realize the preliminary reasoning of the navigation risk when there is no evidence input; the change of the probability of the navigation risk of the ship in the East Siberian Sea of ​​the Arctic route in this embodiment when there is no evidence input is as follows: Figure 6 As shown;

[0099] Taking the ice trap risk in the East Siberian Sea of ​​the Arctic route in July as an example, in the first time slice, the probability of high ice trap risk is 26.46%, the probability of medium ice trap risk is 33.70%, and the probability of low ice trap risk is 39.83%; in the seventh time slice, the probability of high ice trap risk is 21.23%, the probability of medium ice trap risk is 30.74%, and the probability of low ice trap risk is 48.04%; it can be found that the probability of low ice trap risk has an increasing trend, and the navigation risk is decreasing with time. This is caused by the melting of sea ice, which is consistent with the actual situation.

[0100] Step 6.2: When the status of a risk factor node in the past, present and future is known, it can be input into the dynamic Bayesian network as evidence to achieve real-time update of navigation risk.

[0101] Assume that the crew obtains data on certain risk factors for the next three days by observing the weather forecast. For example, it is known that the visibility status for the next three days is: status 1, status 2, status 1; the temperature status is: status 2, status 3, status 2; and the ice thickness status is: status 2, status 1, status 1. By inputting this information as new evidence into the dynamic Bayesian network, the navigation risk can be updated. The dynamic probability change diagram of ship-ice collision risk is shown in the figure below: Figure 7 shown.

[0102] A dynamic assessment system for ship navigation risks in the Arctic route based on dynamic Bayesian networks:

[0103] The evaluation system includes a risk factor identification module, a risk level standard evaluation module, a static-dynamic Bayesian network construction module and a dynamic evaluation module;

[0104] The risk factor identification module identifies risk factors affecting the navigation of ships in the Arctic route from three aspects: environment, people and ships;

[0105] The risk level standard assessment module establishes the risk level standard for ships sailing in the Arctic waterway;

[0106] The static-dynamic Bayesian network construction module first constructs a static Bayesian network graph model of the risk of ship navigation in the Arctic waterway, and then determines the parameters of the static Bayesian network model; finally, a dynamic Bayesian network of the risk of ship navigation in the Arctic waterway is constructed based on the parameters of the static Bayesian network model;

[0107] The dynamic assessment module dynamically assesses the navigation risks of ships in the Arctic waterway.

[0108] An electronic device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0109] A computer-readable storage medium is used to store computer instructions, and when the computer instructions are executed by a processor, the steps of the above method are implemented.

[0110] The memory in the embodiments of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory, ROM, a programmable read-only memory, PROM, an erasable programmable read-only memory, EPROM, an electrically erasable programmable read-only memory, EEPROM, or a flash memory. The volatile memory may be a random access memory, RAM, which is used as an external cache. By way of example but not limitation, many forms of RAM are available, such as static random access memory static RAM, SRAM, dynamic random access memory dynamic RAM, DRAM, synchronous dynamic random access memory synchronous DRAM, SDRAM, double data rate synchronous dynamic random access memory double data rate SDRAM, DDR SDRAM, enhanced synchronous dynamic random access memory enhanced SDRAM, ESDRAM, synchronous link dynamic random access memory synchlink DRAM, SLDRAM, and direct memory bus random access memory direct rambus RAM, DR RAM. It should be noted that memory of the methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0111] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented by software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, computer, server or data center through a wired method such as coaxial cable, optical fiber, digital subscriber line digital subscriber line, DSL or wireless such as infrared, wireless, microwave, etc. to another website site, computer, server or data center. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more available media integrated. The available medium can be a magnetic medium such as a floppy disk, a hard disk, a tape, an optical medium such as a high-density digital video disc digital video disc, DVD, or a semiconductor medium such as a solid state hard disk solid state disc, SSD, etc.

[0112] In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in a processor or an instruction in the form of software. The steps of the method disclosed in conjunction with the embodiment of the present application can be directly embodied as a hardware processor for execution, or a combination of hardware and software modules in a processor for execution. The software module can be located in a storage medium mature in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in a memory, and the processor reads the information in the memory and completes the steps of the above method in conjunction with its hardware. To avoid repetition, it is not described in detail here.

[0113] It should be noted that the processor in the embodiment of the present application can be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method embodiment can be completed by an integrated logic circuit of hardware in the processor or an instruction in the form of software. The above processor can be a general-purpose processor, a digital signal processor DSP, an application-specific integrated circuit ASIC, a field programmable gate array FPGA or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiment of the present application can be directly embodied as a hardware decoding processor to perform, or the hardware and software modules in the decoding processor can be combined to perform. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in a memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0114] The above is a detailed introduction to the dynamic assessment method for navigation risks of ships in the Arctic route based on a dynamic Bayesian network proposed in the present invention, and the principles and implementation methods of the present invention are explained. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation method and scope of application. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A dynamic assessment method for ship navigation risk in the Arctic route based on a dynamic Bayesian network, characterized by: The method specifically comprises the following steps: Step 1: Identify the risk factors affecting the navigation of ships in the Arctic route from the three aspects of environment, people and ships; Step 2: Establish risk level standards for ships sailing in the Arctic route; Step 3: Construct a static Bayesian network graph model of the risk of ship navigation in the Arctic route; Step 4: Determine the static Bayesian network model parameters; Step 5: Construct a dynamic Bayesian network for ship navigation risks in the Arctic route; Step 6: Dynamically evaluate the navigation risks of ships in the Arctic route based on the dynamic Bayesian network.

2. The evaluation method according to claim 1, characterized in that: In step 1, the risk factors include environmental factors B1, human factors B2, and ship factors B3; The corresponding indicators of environmental factor B1 are: wind N1, temperature N2, visibility N3, waves N4, sea ice density N5 and sea ice thickness N6; The corresponding indicators of human factor B2 are: insufficient experience in ice navigation N7, insufficient professional skills N8, fatigue N9, improper route selection N 10 、Insufficient information 11 , Lookout FailureN 12 and the risk is underestimated 13 ; The corresponding index of ship factor B3 is: speed N 14 、Operation equipment failure N 15 、Irrational loading N 16 、Ship ice class N 17 、Navigation equipment failure N 18 、Digital charts are wrong 19 .

3. The evaluation method according to claim 2, characterized in that: In step 2, the risks of ship navigation in the Arctic route are divided into ice entrapment risk and ship-ice collision risk, and the risk level is determined by the probability of risk occurrence and the severity of the risk consequences.

4. The evaluation method according to claim 3, characterized in that: In step 3, according to the causal relationship of each risk factor, the hierarchical analysis diagram is transformed into a directed acyclic graph, a static Bayesian network model is constructed, and the supplementary node ice condition N is introduced. 20 ; In the Bayesian network model, if an arc points from node A to node B, A is considered to be the parent node and B is the child node. A node without a parent node is called a root node.

5. The evaluation method according to claim 4, characterized in that: In step 4, Step 4.1: Classify the identified risk factors into states; Step 4.2: Determine the prior probability of all root nodes and the conditional probability table of child nodes and input them into the Bayesian network; the data comes from satellite remote sensing monitoring data of environmental factors such as meteorology and hydrology and expert experience knowledge; The prior probability data of the root nodes "wind N1", "temperature N2", "visibility N3", "sea ice density N5" and "sea ice thickness N6" are derived from meteorological and hydrological satellite remote sensing monitoring data, "lack of experience in ice navigation N7", "lack of professional skills N8", "unreasonable loading N9" and "inadequate navigation experience N10" are derived from the prior probability data of the root nodes "wind N1", "temperature N2", "visibility N3", "sea ice density N5" and "sea ice thickness N6" are derived from the meteorological and hydrological satellite remote sensing monitoring data, "lack of experience in ice navigation N7", "inadequate professional skills N8" and "inappropriate navigation experience N11" are derived from the prior probability data of the root nodes "wind N1", "temperature N2", "visibility N3", "sea ice density N5" and "sea ice thickness N6" are derived from the meteorological and hydrological satellite remote sensing monitoring data, 16 " and "The digital chart is wrong 19 The prior probability of ” comes from expert judgment; The conditional probability table data of sub-node wave N4 comes from meteorological and hydrological satellite remote sensing monitoring data, and the conditional probability table data of other sub-nodes comes from expert experience knowledge; Assume that the node variable N i There are n states, and there are m parent nodes associated with them, denoted as parents j (N i ), parents(N i ) represents the vector composed of all parent node variables of the variable, σ represents the state of the variable, and γ represents the vector parents(N i ), the conditional probability of the variable can be calculated as follows:

6. The evaluation method according to claim 5, characterized in that: In step 5, Step 5.1: Construct a static Bayesian network with the navigation risk of ships in the Arctic route as the target node through steps 1 to 3; Step 5.2: Identify the nodes in the constructed static Bayesian network whose states change over time, and record them as dynamic nodes; Step 5.3: Obtain the time interval for the risk assessment of ship navigation in the Arctic route, and divide the time interval into T time slices according to the number of days. Each time slice corresponds to a static Bayesian network with the same structure. Connect the dynamic nodes of the previous time slice with the next time slice to form a dynamic Bayesian network. Step 5.4: Calculate the state transfer matrix of dynamic nodes between different time slices based on Markov chain; Assume that the dynamic node N has three states, namely: State1, State2 and State3; Formula (2) gives the state distribution M of N at time t t , the three states satisfy equation (3); as time goes by, the state of N in each time slice transfers between the three states, forming a Markov chain; η, γ represents the probability of transition between any two states; Equations (4)-(6) represent the calculation process of the probability of occurrence of the three states at time t+1; the state transition matrix A of N is shown in Equation (7); M t =(S1(t),S2(t),S3(t)) (3) S1(t)+S2(t)+S3(t)=1 (4) 7. The evaluation method according to claim 6, characterized in that: In step 6, Step 6.1: Input the calculated state transfer matrix of dynamic nodes into the dynamic Bayesian network, and obtain the navigation risk R of ships in the Arctic route through automatic updating. t The risk probability that changes over time enables preliminary reasoning about navigation risks when there is no evidence input; Step 6.2: When the status of a risk factor node in the past, present and future is known, it is input into the dynamic Bayesian network as evidence to achieve real-time update of navigation risk.

8. An assessment system for the dynamic assessment method for ship navigation risk in the Arctic waterway based on a dynamic Bayesian network according to any one of claims 1 to 7, characterized in that: The evaluation system includes a risk factor identification module, a risk level standard evaluation module, a static-dynamic Bayesian network construction module and a dynamic evaluation module; The risk factor identification module identifies risk factors affecting the navigation of ships in the Arctic route from three aspects: environment, people and ships; The risk level standard assessment module establishes the risk level standard for ships sailing in the Arctic waterway; The static-dynamic Bayesian network construction module first constructs a static Bayesian network graph model of the risk of ship navigation in the Arctic waterway, and then determines the parameters of the static Bayesian network model; finally, a dynamic Bayesian network of the risk of ship navigation in the Arctic waterway is constructed based on the parameters of the static Bayesian network model; The dynamic assessment module dynamically assesses the navigation risks of ships in the Arctic waterway.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium for storing computer instructions, characterized in that: When the computer instructions are executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images