River-sea combined transportation network reliability detection method and device based on important node identification, and storage medium

Abstract

The invention relates to a river-sea combined transportation network reliability detection method and device based on important node identification and a storage medium, and the method comprises the steps: S1, importing a river-sea combined transportation network, and calculating the feature indexes of each port node according to the imported river-sea combined transportation network, the feature indexes at least comprising a PR value; s2, respectively calculating the failure probability of each port node; s3, multiplying the failure probability of each port by each characteristic index to obtain a failure probability risk corresponding to each index; and S4, comparing the failure probability risk of each index with a pre-configured risk threshold corresponding to the index, and outputting a reliability detection result based on a comparison result. Compared with the prior art, the method has the advantages that the weak points of the river-sea combined transportation network can be accurately positioned, and the like.

Description

Technical Field

[0001] The present invention relates to the field of transportation, and in particular to a method, device and storage medium for detecting the reliability of a river-sea combined transport network based on important node identification. Background Art

[0002] In the actual operation of the river-sea transport network, different port nodes have different levels of importance. For example, seaports with superior geographical locations and large throughput have higher status and functions than river ports. These ports with high importance play an important role in ship transshipment, berthing, and distribution in the network. These important port nodes have different levels of importance in the entire network due to their different scales and geographical locations. These important port nodes are usually key points in the river-sea transport network, which affect the reliability of the entire network. When these ports fail due to accidents or attacks, the entire river-sea transport network may fail. Therefore, it is necessary to sort the importance of port nodes first and filter out such nodes.

[0003] At present, there are mainly the following methods for evaluating the importance of complex network nodes:

[0004] (1) Directly measure the importance of nodes through characteristic parameters of network nodes, such as degree and degree centrality;

[0005] (2) Through the spatial relationship between nodes and connected nodes in the network, the length and number of shortest paths are used to calculate such as betweenness centrality and closeness centrality;

[0006] (3) Through the interaction between nodes in the network, the importance of a node depends on the importance of the nodes connected to it, such as eigenvector centrality;

[0007] (4) It comprehensively considers the node's own attributes and the spatial relationship between connected nodes, such as Katz centrality, which not only considers the impact of neighboring nodes, but also adds a constant to consider the central node itself.

[0008] The above methods are relatively simple in evaluating nodes and have poor evaluation accuracy, especially the evaluation of the river-sea transport network model is inconsistent with the actual situation of the port.

[0009] Since the river-sea transport network is subject to uncertainties such as seasonal fluctuations and climate change during its actual operation, system reliability is an indicator of the possibility of the system performing its intended functions within a predetermined time under specified conditions. In the context of the river-sea transport network, system reliability analysis is essential to ensure smooth operation under various uncertainties.

[0010] At the same time, in the context of river-sea transport networks, the evaluation of network reliability should focus more on the interdependence between nodes and the sensitivity of the entire system to uncertain factors. The possibility of node failure is not only affected by its own properties, but also by its position in the entire network. In addition, from the perspective of risk management, the reliability risk of the network involves not only the probability of node failure, but also the potential losses caused by node failure.

[0011] Therefore, it is actually necessary to analyze the reliability of the river-sea transport network and the probability of node failure, and to strengthen and optimize the river-sea transport network based on the analysis results. Summary of the invention

[0012] The purpose of the present invention is to provide a method, device and storage medium for detecting the reliability of a river-sea transport network based on important node identification.

[0013] The purpose of the present invention can be achieved by the following technical solutions:

[0014] A reliability detection method for river-sea transport network based on important node identification, comprising:

[0015] Step S1: importing a river-sea transport network, and calculating characteristic indicators of each port node according to the imported river-sea transport network, wherein the characteristic indicators at least include a PR value;

[0016] Step S2: Calculate the failure probability of each port node respectively;

[0017] Step S3: multiply the failure probability of each port by each characteristic index to obtain the failure probability risk corresponding to each index;

[0018] Step S4: Compare the failure probability risk of each indicator with the pre-configured risk threshold corresponding to the indicator, and output the reliability detection result based on the comparison result.

[0019] The calculation process of the PR value of the port node includes:

[0020] Step S1-1: Read the port nodes and edges of the river-sea transport network;

[0021] Step S1-2: constructing a weighted directed graph of the river-sea transport network according to the read port nodes and edges, wherein the weighted directed graph of the river-sea transport network includes a port node and a directed edge connecting two port nodes;

[0022] Step S1-3: Initialize the PR value of each port node;

[0023] Step S1-4: Iteratively update the PR value of each port node based on the weighted adjacency matrix until the iteration termination condition is met;

[0024] Step S1-5: Output the PR value of each port node when the iteration termination condition is met as the final PR value of each port node.

[0025] In step S1-3, the PR value of each port node is initialized to:

[0026]

[0027] Where: PR(p i ) is the PR value of port node pi, N is the number of port nodes, d is the damping factor, M(p i ) is composed of all nodes pointing to port p i The set of starting points of the directed edges of L(p j ) points to the port node p i The number of directed edges.

[0028] The step S1-4 comprises:

[0029] Step S1-4-1: Generate the weight of each directed edge based on the ship's stay time, use the obtained weight as an element in the weighted adjacency matrix, and construct an initial weighted adjacency matrix.

[0030] Step S1-4-2: Update the PR value of each port node based on the initial weighted adjacency matrix:

[0031]

[0032] Where: w ji is the weight of the directed edge from port node pj to port node pi, S(p j ) is from port node p j The sum of the weights of the outgoing directed edges;

[0033] Step S1-4-3: perform allocation transfer and update the weighted adjacency matrix;

[0034] Step S1-4-4: recalculate the PR value of each port node based on the updated weighted adjacency matrix, and determine whether the change of the PR value of all port nodes is less than the pre-configured convergence threshold. If yes, the iteration termination condition is met, otherwise, execute step S1-4-5;

[0035] Step S1-4-5: Determine whether the number of allocations reaches the target number. If so, the iteration termination condition is met. Otherwise, return to step S1-4-3.

[0036] The step S2 comprises:

[0037] Step S2-1: Calculate the reliability index of each port node;

[0038] Step S2-2: Based on the reliability index of each port node, calculate the failure probability of each port node.

[0039] The reliability index of the port node is:

[0040]

[0041] in: is the port node p i The reliability index, is the average actual throughput, is the average value of the designed throughput, is the standard deviation of the actual throughput, is the standard deviation of the designed throughput.

[0042] The failure probability of the port node is:

[0043]

[0044] in: is the port node p i is the failure probability of the standard normal distribution.

[0045] The characteristic indicators also include degree centrality, betweenness centrality, closeness centrality, eigenvector centrality and node strength.

[0046] A device for detecting the reliability of a river-sea transport network based on identifying important nodes, comprising a memory, a processor, and a program stored in the memory, wherein the processor implements the above-mentioned method when executing the program.

[0047] A storage medium stores a program, which implements the above method when executed.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] First, the improved Pagerank algorithm is used as the basis for reliability detection to identify important port nodes in the network. Secondly, the reliability of the river-sea transport network nodes and the entire network is obtained through the reliability index and failure probability. Finally, the failure probability of the node is calculated through the node reliability index, and the node failure risk is calculated in combination with the node importance index. Through the above method, the importance of each port node in the river-sea transport network, the reliability of each node and the entire network, and the impact of network node failure on the network can be obtained. This provides a feasible technical solution for the reliability analysis of each port and provides guidance for port operations and river-sea shipping travel. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 The river-sea transport network model constructed using the data of major domestic ports selected for this application;

[0051] Figure 2 This is the relationship between the overall reliability coefficient and failure probability of the river-sea transport network in this application;

[0052] Figure 3 This is a schematic diagram of the main steps of the invention of this application. DETAILED DESCRIPTION

[0053] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0054] A reliability detection method for river-sea transport network based on important node identification, such as Figure 3 As shown, including:

[0055] Step S1: importing a river-sea transport network, and calculating characteristic indicators of each port node according to the imported river-sea transport network, wherein the characteristic indicators at least include a PR value;

[0056] In this embodiment, a river-sea transport network model is constructed using data from major domestic ports. In addition, in this embodiment, characteristic indicators also include degree centrality, betweenness centrality, closeness centrality, eigenvector centrality and node strength.

[0057] The calculation process of the PR value of the port node includes:

[0058] Step S1-1: Read the port nodes and edges of the river-sea transport network;

[0059] Step S1-2: constructing a weighted directed graph of the river-sea transport network according to the read port nodes and edges, wherein the weighted directed graph of the river-sea transport network includes a port node and a directed edge connecting two port nodes;

[0060] Step S1-3: Initialize the PR value of each port node, wherein the PR value of each port node is initialized to:

[0061]

[0062] Where: PR(p i ) is the PR value of port node pi, N is the number of port nodes, d is the damping factor, M(p i ) is composed of all nodes pointing to port p i The set of starting points of the directed edges of L(pj ) points to the port node p i The number of directed edges of . In this embodiment, the damping factor is 0.85.

[0063] Step S1-4: Iteratively update the PR value of each port node based on the weighted adjacency matrix until the iteration termination condition is met, including:

[0064] Step S1-4-1: Generate the weight of each directed edge based on the ship's stay time, use the obtained weight as an element in the weighted adjacency matrix, and construct an initial weighted adjacency matrix.

[0065] Step S1-4-2: Update the PR value of each port node based on the initial weighted adjacency matrix:

[0066]

[0067] Where: w ji is the weight of the directed edge from port node pj to port node pi, S(p j ) is from port node p j The sum of the weights of the outgoing directed edges;

[0068] Step S1-4-3: perform allocation transfer and update the weighted adjacency matrix;

[0069] The allocation transfer process is similar to the existing PageRank algorithm. First, we need to calculate the j The sum of the weights of the outgoing edges:

[0070]

[0071] Then calculate the matrix R:

[0072]

[0073] Afterwards, the transfer matrix is ​​calculated based on the calculated matrix R:

[0074]

[0075] Among them, P is the transfer matrix and E is the identity matrix of order n.

[0076] The weights are updated according to the transfer matrix to achieve the update of the weighted adjacency matrix. In the specific process, each element in the weighted adjacency matrix is ​​multiplied by the corresponding element in the transfer matrix.

[0077] Step S1-4-4: recalculate the PR value of each port node based on the updated weighted adjacency matrix, and determine whether the change of the PR value of all port nodes is less than the pre-configured convergence threshold. If yes, the iteration termination condition is met, otherwise, execute step S1-4-5;

[0078] Step S1-4-5: Determine whether the number of allocations reaches the target number. If so, the iteration termination condition is met. Otherwise, return to step S1-4-3.

[0079] Step S1-5: Output the PR value of each port node when the iteration termination condition is met as the final PR value of each port node.

[0080] In this embodiment, the PR value of each port node is obtained as shown in Table 1:

[0081] Table 1

[0082] Node Name PR value Node Name PR value Shanghai Port 0.05393 Zhenjiang Port 0.00718 Suzhou Port 0.05037 Huangshi Port 0.00714 Nantong Port 0.03569 Ma On Shan 0.00714 Yueyang Port 0.01486 Jingzhou Port 0.00714 Shenzhen Port 0.01173 Chongqing Port 0.00714 Taizhou Port 0.01123 Hangzhou Port 0.00714 Yichang Port 0.00794 Guangzhou Port 0.00714 Yangzhou Port 0.00787 Jiangyin Port 0.00714 Wuhan Port 0.00737 Nanjing Port 0.00714 Wuhu Port 0.00732 Jiujiang Port 0.00714

[0083] As can be seen from Table 1, the important port nodes of the Yangtze River Economic Belt regional river-sea transport network are Shanghai Port, Suzhou Port, and Nantong Port. Among them, Shanghai Port ranks first with a PR value of 0.05393, reflecting the importance of Shanghai as an international trade and logistics center. It has a huge shipping network and is a key port node connecting multiple domestic ports. Next are Nantong Port and Suzhou Port, whose PR values ​​are similar, reflecting their important roles in river-sea transport and regional logistics. In particular, Wuhan Port has a superior geographical location. As an important port in the middle reaches of the Yangtze River, it gathers east-west and north-south ships.

[0084] There are also some port nodes with medium PR values, such as Shenzhen, Yueyang, and Taizhou. The PR values ​​of these ports are between 0.014 and 0.017, indicating that they also have a place in the river-sea transport network and have some special routes in regional logistics. For example, Yueyang Port and Taizhou Port have two routes with high edge betweenness in Table 3-7: Shanghai Port-Yueyang Port and Nantong Port-Taizhou Port. Although their degrees in the network are not high, the importance of the ports they connect is relatively high, and the weight of the edges they connect is also relatively high. The final calculation results also reflect the core idea of ​​the PR algorithm, that is, the importance of a node is determined by the importance of other nodes pointing to it.

[0085] In order to further verify the importance of each port node in river-sea transport, the network moderateness, betweenness centrality, closeness centrality and port nodes with higher node weight rankings are compared and analyzed. The importance of port nodes in the river-sea transport network under different ranking methods is shown in Table 2:

[0086] Table 2

[0087]

[0088] From Table 2, we can see that the importance ranking of port nodes varies under various sorting methods. In the four sorting methods except eigenvector centrality and close centrality, Shanghai Port, Nantong Port and Suzhou Port are ranked in the top three, which further proves that these three ports are the core hub ports of the river-sea transport network. They have superior geographical locations. Among them, Nantong Port and Shanghai Port are both located at the outlet of the lower reaches of the Yangtze River. In addition, the ports have large areas, complete infrastructure and high transshipment capacity. It can be seen that the top ten port nodes under degree sorting appear in the other four sorting methods, such as Jiangyin Port and Nanjing Port. This is mainly because the degree distribution of the nodes in the constructed river-sea transport network is uneven, which also reflects the insufficiency of measuring the importance of nodes by relying solely on the single attribute of port nodes. On the contrary, the ranking of port node importance using the PR value (using the information of port nodes and edges) sorting method is more in line with the actual situation.

[0089] Step S2: Calculate the failure probability of each port node separately, including:

[0090] Step S2-1: Calculate the reliability index of each port node:

[0091]

[0092] in: is the port node p i The reliability index, is the average actual throughput, is the average value of the designed throughput, is the standard deviation of the actual throughput, is the standard deviation of the designed throughput.

[0093] Step S2-2: Based on the reliability index of each port node, calculate the failure probability of each port node:

[0094]

[0095] in: is the port node p i is the failure probability of the standard normal distribution.

[0096] The connotation of network reliability refers to calculating the failure probability of port nodes under the influence of various uncertain factors, and judging the reliability of port nodes by the size of the failure probability. The greater the failure probability, the lower the reliability and the smaller the reliability index, and vice versa.

[0097] In order to calculate the reliability of port nodes, this application collects the cargo throughput data of major ports in the Yangtze River Economic Belt from January to December 2019. The cargo throughput data of some cities are shown in Table 3:

[0098] Table 3

[0099]

[0100] Combining the above information, Table 4 shows the reliability index of the top nine port nodes ranked by PR value under different port capacities.

[0101] Table 4

[0102]

[0103]

[0104] Overall, the reliability index of all port nodes increases with the increase of port capacity; at the original capacity, although Shanghai Port is the port with the highest PR value, its reliability index is not the highest. This shows that the port throughput of Shanghai Port fluctuated greatly in 2019, which may be caused by uncertain factors such as weather, operational efficiency, and logistics management capabilities, resulting in a decrease in the reliability index of port nodes.

[0105] In addition, the reliability modeling of the river-sea transport network as a whole can be carried out by using the Ditlevsen upper and lower bound method.

[0106] The Ditlevsen upper and lower bounds method is used to model the reliability of the river-sea transport network as a whole. The upper bound in the Ditlevsen upper and lower bounds method is the probability that all components fail at the same time, and the lower bound refers to the probability that the system will fail as long as one component fails. The calculation formula is as follows:

[0107]

[0108] Where: P f is the overall failure probability of river-sea transport, P upper is the upper bound failure probability, P lower is the lower bound failure probability.

[0109] Then we can get the overall reliability index of Jianghai network:

[0110] β=-φ -1 (P f )

[0111] Step S3: multiply the failure probability of each port by each characteristic index to obtain the failure probability risk corresponding to each index;

[0112] In the context of the river-sea transport network, the evaluation of network reliability should focus more on the interdependence between nodes and the sensitivity of the entire system to uncertain factors. The possibility of node failure is not only affected by its own properties, but also by its position in the entire network. In addition, from the perspective of risk management, the reliability risk of the network involves not only the probability of node failure, but also the potential losses caused by node failure. Therefore, the evaluation of the reliability of the river-sea transport network should include two aspects: one is the probability of node failure itself; the other is the impact of the node failure on the overall function of the network.

[0113] In the river-sea transport network, the failure probability of the port node is calculated based on the initial capacity of the port node. The risk value of the node calculated by multiplying the node's degree, betweenness centrality, closeness centrality, eigenvector centrality and PR value is shown in Table 5.

[0114] Table 5

[0115]

[0116]

[0117] Step S4: Compare the failure probability risk of each indicator with the pre-configured risk threshold corresponding to the indicator, and output the reliability detection result based on the comparison result.

[0118] It can be seen from Table 5 that the three most critical nodes in the river-sea transport network are Shanghai Port, Suzhou Port and Nantong Port.

[0119] As the only port with the maximum value (1.0) in PR value, node strength, degree, and close centrality risk, Shanghai Port proves that its failure will have a serious impact on the network, and also proves that Shanghai Port is the most important node in the river-sea transport network. However, its risk value under eigenvector centrality is almost 0. This is because although Shanghai Port connects many nodes in the constructed network, the eigenvector centrality of these nodes is not high, so the eigenvector centrality of Shanghai Port will also be relatively low.

[0120] Suzhou Port has the highest value (1.0) in eigenvector centrality risk, which indicates that although Suzhou Port does not have as many direct connections as Shanghai Port, the nodes it connects to are extremely important. However, its betweenness centrality risk value is 0, indicating that it does not appear on the shortest path between other nodes in the network.

[0121] Nantong Port has the largest value (1.0) in the betweenness centrality risk, indicating that it plays an important role as a bridge in the network, connecting multiple other important nodes. Its degree risk is also high (0.532), but its PR value and node strength risk are relatively low, which means that although it has a large number of direct connections, the overall strength of the connection is not high.

[0122] Most port nodes have low risk, indicating that they play a limited role in the river-sea transport network.

[0123] If the above functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program code.

Claims

1. A reliability detection method for river-sea transport network based on important node identification, characterized in that: include: Step S1: importing a river-sea transport network, and calculating characteristic indicators of each port node according to the imported river-sea transport network, wherein the characteristic indicators at least include a PR value; Step S2: Calculate the failure probability of each port node respectively; Step S3: multiply the failure probability of each port by each characteristic index to obtain the failure probability risk corresponding to each index; Step S4: Compare the failure probability risk of each indicator with the pre-configured risk threshold corresponding to the indicator, and output the reliability detection result based on the comparison result.

2. According to claim 1, a method for detecting the reliability of a river-sea transport network based on important node identification is characterized in that: The calculation process of the PR value of the port node includes: Step S1-1: Read the port nodes and edges of the river-sea transport network; Step S1-2: constructing a weighted directed graph of the river-sea transport network according to the read port nodes and edges, wherein the weighted directed graph of the river-sea transport network includes a port node and a directed edge connecting two port nodes; Step S1-3: Initialize the PR value of each port node; Step S1-4: Iteratively update the PR value of each port node based on the weighted adjacency matrix until the iteration termination condition is met; Step S1-5: Output the PR value of each port node when the iteration termination condition is met as the final PR value of each port node.

3. The method for detecting the reliability of river-sea transport network based on important node identification according to claim 2 is characterized in that: In step S1-3, the PR value of each port node is initialized to: Where: PR(p i ) is the PR value of port node pi, N is the number of port nodes, d is the damping factor, M(p i ) is composed of all nodes pointing to port p i The set of starting points of the directed edges of L(p j ) points to the port node p i The number of directed edges.

4. The method for detecting the reliability of river-sea transport network based on important node identification according to claim 2 is characterized in that: The step S1-4 comprises: Step S1-4-1: Generate the weight of each directed edge based on the ship's stay time, use the obtained weight as an element in the weighted adjacency matrix, and construct an initial weighted adjacency matrix. Step S1-4-2: Update the PR value of each port node based on the initial weighted adjacency matrix: Where: w ji is the port node p j The weight of the directed edge of port node pi, S(p j ) is from port node p j The sum of the weights of the outgoing directed edges; Step S1-4-3: perform allocation transfer and update the weighted adjacency matrix; Step S1-4-4: recalculate the PR value of each port node based on the updated weighted adjacency matrix, and determine whether the change of the PR value of all port nodes is less than the pre-configured convergence threshold. If yes, the iteration termination condition is met, otherwise, execute step S1-4-5; Step S1-4-5: Determine whether the number of allocations reaches the target number. If so, the iteration termination condition is met. Otherwise, return to step S1-4-3.

5. The method for detecting the reliability of river-sea transport network based on important node identification according to claim 1 is characterized in that: The step S2 comprises: Step S2-1: Calculate the reliability index of each port node; Step S2-2: Based on the reliability index of each port node, calculate the failure probability of each port node.

6. A method for detecting the reliability of river-sea transport network based on important node identification according to claim 5, characterized in that: The reliability index of the port node is: in: is the port node p i The reliability index, is the average actual throughput, is the average value of the designed throughput, is the standard deviation of the actual throughput, is the standard deviation of the designed throughput.

7. A method for detecting the reliability of river-sea transport network based on important node identification according to claim 6, characterized in that: The failure probability of the port node is: in: is the port node p i is the failure probability of the standard normal distribution.

8. The method for detecting the reliability of river-sea transport network based on important node identification according to claim 1 is characterized in that: The characteristic indicators also include degree centrality, betweenness centrality, closeness centrality, eigenvector centrality and node strength.

9. A device for detecting the reliability of a river-sea transport network based on identifying important nodes, comprising a memory, a processor, and a program stored in the memory, characterized in that: When the processor executes the program, the method according to any one of claims 1 to 8 is implemented.

10. A storage medium having a program stored thereon, characterized in that: When the program is executed, the method according to any one of claims 1 to 8 is implemented.

Images