River network robustness evaluation method based on complex network theory

Abstract

The invention discloses a river network robustness evaluation method based on a complex network theory. The method comprises the following steps: acquiring and preprocessing topographic features, river network function indexes and river network vector data; calculating all functional indexes including a flood control capability index, a water supply capability index and an ecological capability index of each river reach in the river network; constructing a river network complex network model G = (V, E) of the drainage basin to be evaluated; describing and quantifying topological structure features of the river network based on the complex network model of the river network; and performing node removal attack simulation on the complex network model of the river network, removing the maximum connected subgraph proportion and the network efficiency of the network after the river reach is removed, and observing the variation trend of the overall robustness level of the complex network model of the river network subjected to the node removal attack so as to determine an evaluation index which has the most significant influence on the global robustness of the network. The method is suitable for evaluating the structural stability and functional evolution of the river network and resisting various risk events.

Description

Technical Field

[0001] The present invention relates to the field of river network evaluation and management, and in particular to a river network robustness evaluation method. Background Art

[0002] Early river network research mainly relied on qualitative description. With the rise of quantitative analysis methods, the Horton-Strahler system, Shreve-Smart random topology model, fractal theory, and self-similarity theory have been proposed one after another, providing quantitative tools for river network research. In recent years, the quantitative description of river network morphological characteristics has made continuous progress, and attempts have been made to link river network characteristic parameters with geological structure, climatic conditions, and watershed factors. However, traditional research focuses on river networks in a specific period of time or digital river networks extracted based on DEM, which are significantly different from the actual "natural + artificial" coupled river networks. Especially in plain areas, the accuracy of river network extraction is difficult to guarantee. The actual river network is affected by many factors and presents a complex network structure rather than an idealized tree structure, which makes it difficult to accurately describe the quantitative relationship between the large basin topography and the actual network structure. In addition, traditional indicators are usually based only on static network topology, which cannot fully capture the dynamic changes and heterogeneity of the network, and the discussion on the dynamic response mechanism of nodes and river sections under external disturbances is still immature.

[0003] In recent years, complex network theory has been widely used in the study of various complex systems, providing a new perspective for understanding and analyzing river network systems. Complex network theory abstracts the river network into a network composed of nodes (river junctions, reservoirs, sluices, etc.) and edges (river sections), and uses network topology indicators to quantify the structural characteristics and functions of the river network. Some studies have applied complex network theory to river network analysis, such as studying the connectivity, hierarchy, and small-world characteristics of river networks. However, most of these studies focus on static structural analysis, which can only reflect the topological characteristics of the river network at a specific moment, and lack in-depth research on the functional evolution and dynamic robustness of the river network. Existing research on river network robustness also has some limitations: traditional robustness indicators are usually based only on static network topology, ignoring the differences in river section functions, unable to fully capture the dynamic changes and heterogeneity of the network, and difficult to reflect the dynamic response mechanism of nodes and river sections under external disturbances; many studies only focus on the impact of a single risk event on the river network, while in actual situations, river networks often face the superposition of multiple risks, leading to cascade effects; in particular, there is a lack of systematic evaluation of changes in river network robustness under artificial modification, and there are few studies that empower river network functions to the edges of complex networks, and then analyze the contribution of river sections with different functions to network robustness. Summary of the invention

[0004] In order to make up for the shortcomings of the prior art, the present invention aims to provide a river network robustness evaluation method and system based on complex network theory, which utilizes the river network complex network model represented by graph theory, and designs a variety of node removal attack simulation strategies that combine static attack and dynamic attack on nodes in the dynamic / static river network complex network model, thereby realizing the evaluation of river network robustness changes under natural and artificial transformation.

[0005] The present invention is achieved by the following technical solutions:

[0006] The present invention proposes a river network robustness evaluation method based on complex network theory, comprising:

[0007] S1, collect and preprocess terrain characteristics, river network function index and river network vector data;

[0008] S2, river network function analysis is performed based on topographic features, river network function index and river network vector data, and all function indexes including flood control capacity index, water supply capacity index and ecological capacity index are calculated for each river section in the river network;

[0009] S3, based on the preprocessed river network vector data, construct a river network complex network model G = (V, E) of the basin to be evaluated, where the node set V represents the source, intersection and key positions of the control engineering in the river section; the edge set E represents the river section as the node edge in the network; the input of the model is the river network vector data, the attribute data of the node and each river section, including the flood control capacity index, water supply capacity index and ecological capacity index; the output of the model is the abstract network topology of the river network represented by graph theory, the network calculation index and the weight attribute of the edge;

[0010] S4, describing and quantifying the topological structural characteristics of the river network based on the river network complex network model, specifically including: calculating the node and edge centrality indicators such as degree centrality, betweenness centrality and closeness centrality of the river network;

[0011] S5. Based on the three indicators of degree centrality, betweenness centrality and closeness centrality, the node removal attack simulation was carried out on the river network complex network model. The maximum connected subgraph ratio and network efficiency of the network after removing the river section were observed, and the changing trend of the overall robustness level of the river network complex network model subjected to the node removal attack was observed to determine the evaluation indicators that have the most significant impact on the global robustness of the network.

[0012] In some embodiments, the S2 further comprises:

[0013] Calculate the flood control capacity index of each river section, the expression is as follows:

[0014]

[0015] FD,i=a1F R,i+a2Fc,i

[0016] Where Fc,i is the flood control capacity improvement index of the i-th river section, F D , i is the flood control capacity index of the i-th river section, step is the path length from the river section to the upstream reservoir in the graph model, As is the storage capacity of the upstream reservoir of the river section, a1 and a2 are the weights of the river section flood discharge capacity and flood control improvement index respectively; among them, the average flood discharge capacity index of each river section is calculated, and the expression is as follows:

[0017] F R,i =Avg(Q1,Q2,Q3)

[0018] Where A is the cross-sectional water flow area, C is the Xie Cai coefficient, R is the hydraulic radius, J is the hydraulic slope, and F is the hydraulic radius. R , i is the average flood discharge capacity index of the i-th river section, i is the river section number, Q1, Q2, and Q3 are the flow rates of the three sections of the upper, middle, and lower reaches of the river section respectively;

[0019] The water supply capacity index of each river section is evaluated as follows:

[0020]

[0021] F S,i =a3E G +a4Z i +a5F c,i

[0022] Among them, E G is the network efficiency index, e st is the efficiency index of node s and node t in the network, d st is the path distance between node s and node t in the network, N is the number of nodes in the graph model, Fc,i is the water supply improvement index of the i-th river section, Zi is the cultivated land area of ​​the river section receiving area of ​​the i-th river section, Fs,i is the water supply capacity index of the i-th river section, that is, the local water supply capacity index, a3 is the influence of the river section network efficiency index on the water supply capacity index, a4 represents the influence of cultivated land distribution on the water supply capacity index, and a5 is the influence of the water supply improvement index on the water supply capacity index;

[0023] Functional connectivity is used to characterize the ecological capacity index of each river section, and the expression is as follows:

[0024]

[0025] In the formula, BC is the network betweenness centrality index, TWI is the terrain wetness index, TCI is the terrain connectivity index, β is the water network connectivity correction coefficient, and F E ,i is the ecological capacity index of the ith river section.

[0026] In some embodiments, the S4 further comprises:

[0027] The expression of degree centrality of each node in the river network complex network model is as follows:

[0028]

[0029] Where: k j is the degree of node j, N is the total number of nodes; node s, node t, node j are any three nodes in the network;

[0030] The expression of the betweenness centrality of each node in the river network complex network model is as follows:

[0031]

[0032] In the formula, g st is the number of shortest paths from node s to node t, is the number of shortest paths from node s to node t through node j;

[0033] The expression of the closeness centrality of each node in the river network complex network model is as follows:

[0034]

[0035] Where: N is the total number of nodes, d sj is the shortest path length between node s and node j.

[0036] In some embodiments, the removal attack in S5 further includes static attack and dynamic attack, and further includes static attack analysis, by removing different nodes one by one and observing the changes in network indicators, identifying the nodes whose removal will cause a significant decrease in the maximum connected subgraph ratio or global efficiency; dynamic attack analysis, sorting the nodes based on different centrality indicators and gradually removing them, observing the changing trend of the overall robustness level of the network, so as to determine the evaluation indicators that have the most significant impact on the global robustness of the network.

[0037] In some embodiments, S5 further includes: weighting based on edge betweenness centrality and the flood control capacity index, water supply capacity index and ecological capacity index calculated in S2, sorting the river sections according to the node centrality index, performing edge removal attack simulation on the sorted river networks, and calculating the maximum connected subgraph ratio and global efficiency of the network after removing the river sections.

[0038] In some implementations, the expression of the maximum connected subgraph ratio σ is as follows:

[0039]

[0040] Where S is the number of nodes contained in the maximum connected subgraph, S=max(Nj'), and Nj' is the set of the number of nodes in the connected subgraph after the network node j' is destroyed.

[0041] In some embodiments, the expression of network efficiency is as follows:

[0042]

[0043] Where: Ej'j = 1 / dj'j, dj'j is the shortest path length between node j' and node j, and N is the total number of nodes in the network.

[0044] Compared with the existing work, the present invention achieves the following positive technical effects:

[0045] 1) The present invention is applicable to assessing the structural stability and functional evolution of river networks and resisting various risk events such as natural disasters and human interference;

[0046] 2) By constructing a river network complex network model that reflects the natural and current river network topology, the topological characteristics of the river network can be described and quantified, which can dynamically and accurately evaluate the robustness of river network nodes and edges;

[0047] 3) Combining various river network node removal attack simulation strategies and centrality indicators, the robustness of the river network is dynamically evaluated to provide a more scientific basis for water network planning and management. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is an overall flow chart of a river network robustness evaluation method based on complex network theory of the present invention;

[0049] Figure 2 It is a flow chart of the implementation of a river network robustness evaluation method based on complex network theory of the present invention;

[0050] Figure 3 Schematic diagram of the construction of a river network complex network model; (a) distribution map of river networks and water conservancy projects, (b) example diagram of the constructed complex network model of an example watershed, including river intersections, reservoirs, and sluices as nodes of the complex network, and river channels as edges of the complex network;

[0051] Figure 4 It is a schematic diagram of the robustness of the river network nodes and edges using a static attack strategy;

[0052] Figure 5 It is a schematic diagram of the robustness of river network nodes and edges using dynamic attack strategy. DETAILED DESCRIPTION

[0053] The technical solution of the river network robustness evaluation method based on complex network theory is described in detail below with reference to the accompanying drawings.

[0054] like Figure 1 As shown, the river network robustness evaluation method based on complex network theory of the present invention comprises the following steps:

[0055] Step 1: Collect and preprocess terrain features, river network function index and river network vector data. Specifically, collect digital elevation model (DEM) data and river network vector data of the study area. DEM data is used to extract terrain features and calculate river network function index, and river network vector data is used to construct the topological structure of the river network. Preprocess the data, including data cleaning, format conversion, projection transformation, etc., to ensure the accuracy and consistency of the data.

[0056] Step 2: Analyze the river network function based on terrain features, river network function index and river network vector data. Specifically, calculate the flood control capacity index, water supply capacity index and ecological capacity index of the river network based on the digital elevation model DEM data and river network vector data, and analyze the spatial distribution characteristics of different functions and the impact of artificial transformation; according to the evaluation results of a certain river basin function, the spatial distribution characteristics include, for example, the statistics of flood control function index, water supply function index and ecological function index; artificial transformation is the result of human intervention in water resources, which directly changes the flow direction and connectivity of natural water systems, and directly affects the spatial distribution of different water network functions. For example, the construction of reservoirs will change the flow process downstream, thereby affecting the flood control and ecological functions downstream; the construction of artificial channels will improve the water supply capacity of a specific area, but may reduce the ecological connectivity of the river; specific treatment includes:

[0057] Step 2-1: The flood discharge capacity of the river and the flood storage capacity of large-scale projects determine the flood control capacity of the local river. Based on these two factors, the average flood discharge capacity index of each river section is calculated. The expression is as follows:

[0058] F R,i =Avg(Q1,Q2,Q3) (1)

[0059] Where A is the cross-sectional water flow area (m 2 ), C is the Xie Cai coefficient, R is the hydraulic radius, J is the hydraulic slope, F R , i is the average flood discharge capacity index of the i-th river section, i is the river section number, Q1, Q2, and Q3 are the flow rates of the three sections of the upper, middle, and lower reaches of the river section (m 3 / s), the flow rate of the cross section follows the formula is (m 3 / s) is calculated;

[0060] Since the reservoir has an enhancing effect on the flood control of the downstream river section, the distance between the river section and the upstream reservoir is calculated, and the flood control enhancement index is estimated according to the reservoir storage capacity. Finally, the flood control capacity index of each river section is weighted and calculated. The expression is as follows:

[0061]

[0062] F D ,i=a1F R ,i+a2Fc,i (3)

[0063] Where Fc,i is the flood control capacity improvement index of the i-th river section, F D ,i is the flood control capacity index of the ith river section, step is the path length from the river section to the upstream reservoir in the graph model, As is the storage capacity of the upstream reservoir of the river section, a1 and a2 are the weights of the river section's flood discharge capacity and flood control improvement index respectively. The flood control function index mainly focuses on the ability of the water network to discharge and resist floods. Reservoirs and other storage projects have an auxiliary role in improving the flood control capacity of their downstream river sections. Therefore, a1 = 0.7, a2 = 0.3 are taken to estimate the flood control function index of each river section in the Haihe River Basin;

[0064] Step 2-2: Use the network efficiency index and the distribution of cultivated land within the water receiving area to evaluate the water supply capacity index of each river section. The expression is as follows:

[0065]

[0066] F S,i =a3E G +a4Z i +a5F c,i (5)

[0067] Among them, E G is the network efficiency index, e st is the efficiency index of node s and node t in the network, d st is the path distance between node s and node t in the network, N is the number of nodes in the graph model, Fc,i is the water supply improvement index of the i-th river section, Zi is the cultivated land area of ​​the river section receiving area of ​​the i-th river section, the water receiving range is approximately determined by the river section buffer zone, and 5 kilometers are selected as the buffer zone demarcation range, Fs,i is the water supply capacity index of the i-th river section, that is, the local water supply capacity index, a3 is the influence of the river section network efficiency index on the water supply capacity index, a4 represents the influence of cultivated land distribution on the water supply capacity index, and a5 is the influence of the water supply improvement index on the water supply capacity index. Based on the actual situation, in the basin water supply network exemplified by the present invention, a3=0.4, a4=0.3, a5=0.3;

[0068] Step 2-3: Use functional connectivity to characterize the ecological capacity index of each river section. The expression is as follows:

[0069]

[0070] In the formula, BC is the network betweenness centrality index, TWI is the terrain wetness index, TCI is the terrain connectivity index, β is the water network connectivity correction coefficient, and F E , i is the ecological capacity index of the i-th river section, that is, the local ecological capacity index. Among them, BC represents the intermediary capacity of the river section, and the water network connectivity correction coefficient β is intended to adjust the difference between the intermediary capacity and the actual flow process. It can be understood as a weight coefficient, which takes into account the impact of topography on the intermediary capacity of the water network;

[0071] Step 3: Based on the preprocessed river network vector data, use Python's NetworkX and igraph modules to construct a river network complex network model G = (V, E) of the basin to be evaluated. The following inputs are required to build this model: 1) River network vector data, including the geometric shape information of the river and the geographical location information of key facilities; 2) Node attribute data, including node type, name and location information, where the node type is river section node and engineering node; 3) All functional indexes of each river section calculated in step 2, including flood control capacity index, water supply capacity index and ecological capacity index; Among them, the node set V represents the key locations such as the source, intersection and control engineering (reservoir, large sluice) of the river section, and these geographical locations are generalized as nodes in the network; the edge set E represents the river sections connecting these nodes, and these river sections are generalized as edges in the network; V = {v j}, j = 1, 2, ..., N, N is the total number of nodes, E = {e i}, i'=1,2,...,M, M is the total number of river sections. After the model is built, it will output: 1) an abstract network topology structure, which is stored in a graph representation, including the connection relationship between nodes and edges; 2) indicators calculated based on the network, such as degree centrality, betweenness centrality and closeness centrality, which are used in subsequent attack simulations; 3) the weight attribute of the edge, that is, the functional index of each river section calculated in step 2, which will be used in the edge removal attack simulation. Figure 2 As shown, a schematic diagram of the constructed river network complex network model is shown;

[0072] Step 4: Describe and quantify the topological structural characteristics of the river network based on the river network complex network model, including: calculating the node and edge centrality indicators such as degree centrality, betweenness centrality and closeness centrality of the river network;

[0073] The degree centrality index evaluates the number of edges connected to a node, highlighting nodes with high connectivity, which may have an important impact on network function. The expression of the degree centrality of each node in the river network complex network is as follows:

[0074]

[0075] Where: k j is the degree of node j, N is the total number of nodes; node s, node t, node j are any three nodes in the network;

[0076] The betweenness centrality index evaluates the number of times a node or connection is located on the shortest path between other pairs of nodes. It is a key element in identifying the flow of information or resources in a network. The expression of the betweenness centrality of each node in a river network complex network is as follows:

[0077]

[0078] In the formula, g st is the number of shortest paths from node s to node t, is the number of shortest paths from node s to node t through node j;

[0079] The closeness centrality index is used to measure the proximity of the node to other nodes in the network. This index reflects the efficiency of a node to reach all other nodes in the network, which may be critical to network communication or efficiency. The expression of the closeness centrality of each node in the river network complex network is as follows:

[0080]

[0081] Where: N is the total number of nodes, d sj is the shortest path length between node s and node j;

[0082] Step 5: Based on the three indicators of degree centrality, betweenness centrality and closeness centrality, the river network complex network model is simulated for node removal attack, and the changing trend of the overall robustness level of the river network complex network model under the node removal attack is observed to determine the evaluation indicator with the most significant impact on the global robustness of the network; the types of attacks involved include:

[0083] (1) Static attack: Sort by node centrality index and remove the highest-ranked nodes in turn. After each node is removed, calculate the maximum connected subgraph ratio and global efficiency of the network, and then restore the network to its initial state for the next node removal. This strategy can help identify key nodes and edges in the network and assess the network’s vulnerability to a single risk event.

[0084] (2) Dynamic attack: Sort by node centrality index and remove the highest-ranked nodes in turn. After each node is removed, the maximum connected subgraph ratio and global efficiency of the network are calculated. The network state is not restored and the next node is removed. This strategy can comprehensively evaluate the overall robustness of the network and identify weak links under the superposition of multiple risks.

[0085] The maximum connected subgraph ratio is σ, and the specific expression is as follows:

[0086]

[0087] Where S is the number of nodes contained in the maximum connected subgraph, S=max(Nj'), and Nj' is the set of the number of nodes in the connected subgraph after the network node j' is destroyed.

[0088] The expression of network efficiency is as follows:

[0089]

[0090] Where: Ej'j = 1 / dj'j, dj'j is the shortest path length between node j' and node j, and N is the total number of nodes in the network. Node j' and node j represent any two different nodes in the network, and j'≠j must be satisfied. They can be river section nodes, engineering nodes, or the source and outlet of the river. The summation symbol in the formula traverses all possible node pairs in the network to ensure that the connection efficiency between all nodes is considered. They can be directly connected, indirectly connected, or in some cases there may be no connectivity.

[0091] In static attack analysis, by removing different nodes one by one and observing the changes in network indicators, key nodes are identified, that is, those nodes whose removal will cause a significant decrease in the maximum connected subgraph ratio or global efficiency. In dynamic attack analysis, nodes are sorted based on different centrality indicators and gradually removed, and the changing trend of the overall robustness level of the network is observed, so as to determine the evaluation indicators that have the most significant impact on the global robustness of the network;

[0092] Specifically, based on the edge betweenness centrality and the weighting of the three capacity indicators (flood control capacity, water supply capacity, and ecological capacity) calculated in step 2, the river sections are sorted, and the edge removal attack simulation is performed on the sorted river network, and the maximum connected subgraph ratio and global efficiency of the network after the river sections are removed are calculated. The attack strategy is the same as the attack strategy based on node removal in step 5, including static attack and dynamic attack. In the static attack analysis, by removing different river sections one by one and observing the changes in network indicators, the key river sections are identified, that is, those river sections whose removal will lead to a significant decrease in the maximum connected subgraph ratio or global efficiency. In the dynamic attack analysis, the river sections are sorted and gradually removed based on different neutrality indicators and functional weightings.

[0093] The above is only a specific implementation method of the present invention. The above implementation steps are only used to help understand the specific methods and core ideas of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of several changes or equivalent substitutions, combinations and modifications made within the technical scope disclosed by the present invention and based on the ideas of the present invention without departing from the principles of the present invention. These changes or equivalent substitutions, combinations and modifications should be deemed to fall within the scope of protection of the present invention and should be covered by the scope of protection of the present invention.

Claims

1. A river network robustness evaluation method based on complex network theory, characterized in that: include: S1, collect and preprocess terrain characteristics, river network function index and river network vector data; S2, river network function analysis is performed based on topographic features, river network function index and river network vector data, and all function indexes including flood control capacity index, water supply capacity index and ecological capacity index are calculated for each river section in the river network; S3, based on the preprocessed river network vector data, construct a river network complex network model G = (V, E) of the basin to be evaluated, where the node set V represents the source, intersection and key positions of the control engineering in the river section; the edge set E represents the river section as the node edge in the network; the input of the model is the river network vector data, the attribute data of the node and each river section, including the flood control capacity index, water supply capacity index and ecological capacity index; the output of the model is the abstract network topology of the river network represented by graph theory, the network calculation index and the weight attribute of the edge; S4, describing and quantifying the topological structural characteristics of the river network based on the river network complex network model, specifically including: calculating the node and edge centrality indicators such as degree centrality, betweenness centrality and closeness centrality of the river network; S5. Based on the three indicators of degree centrality, betweenness centrality and closeness centrality, the node removal attack simulation was carried out on the river network complex network model. The maximum connected subgraph ratio and network efficiency of the network after removing the river section were observed, and the changing trend of the overall robustness level of the river network complex network model subjected to the node removal attack was observed to determine the evaluation indicators that have the most significant impact on the global robustness of the network.

2. The river network robustness evaluation method based on complex network theory according to claim 1 is characterized in that: The S2 further comprises: Calculate the flood control capacity index of each river section, the expression is as follows: F D ,i=a1F R ,i+a2Fc,i Where Fc,i is the flood control capacity improvement index of the i-th river section, F D , i is the flood control capacity index of the i-th river section, step is the path length from the river section to the upstream reservoir in the graph model, As is the storage capacity of the upstream reservoir of the river section, a1 and a2 are the weights of the river section flood discharge capacity and flood control improvement index respectively; among them, the average flood discharge capacity index of each river section is calculated, and the expression is as follows: <h2 style=";text-align:left;direction:ltr">F<h2 style=";text-align:left;direction:ltr"> R,i <h2 style=";text-align:left;direction:ltr"> =Avg(Q1,Q2,Q3) Where A is the cross-sectional water flow area, C is the Xie Cai coefficient, R is the hydraulic radius, J is the hydraulic slope, and F is the hydraulic radius. R , i is the average flood discharge capacity index of the i-th river section, i is the river section number, Q1, Q2, and Q3 are the flow rates of the three sections of the upper, middle, and lower reaches of the river section respectively; The water supply capacity index of each river section is evaluated as follows: <h2 style=";text-align:left;direction:ltr">F<h2 style=";text-align:left;direction:ltr"> S,i <h2 style=";text-align:left;direction:ltr"> =a3E<h2 style=";text-align:left;direction:ltr"> G <h2 style=";text-align:left;direction:ltr"> +a4Z<h2 style=";text-align:left;direction:ltr"> i <h2 style=";text-align:left;direction:ltr"> +a5F<h2 style=";text-align:left;direction:ltr"> c,i Among them, E G is the network efficiency index, e st is the efficiency index of node s and node t in the network, d st is the path distance between node s and node t in the network, N is the number of nodes in the graph model, Fc,i is the water supply improvement index of the i-th river section, Zi is the cultivated land area of ​​the river section receiving area of ​​the i-th river section, Fs,i is the water supply capacity index of the i-th river section, that is, the local water supply capacity index, a3 is the influence of the river section network efficiency index on the water supply capacity index, a4 represents the influence of cultivated land distribution on the water supply capacity index, and a5 is the influence of the water supply improvement index on the water supply capacity index; Functional connectivity is used to characterize the ecological capacity index of each river section, and the expression is as follows: In the formula, BC is the network betweenness centrality index, TWI is the terrain wetness index, TCI is the terrain connectivity index, β is the water network connectivity correction coefficient, and F E ,i is the ecological capacity index of the ith river section.

3. According to a river network robustness evaluation method based on complex network theory according to claim 1, it is characterized in that: The S4 further comprises: The expression of degree centrality of each node in the river network complex network model is as follows: Where: k j is the degree of node j, N is the total number of nodes; node s, node t, node j are any three nodes in the network; The expression of the betweenness centrality of each node in the river network complex network model is as follows: In the formula, g st is the number of shortest paths from node s to node t, is the number of shortest paths from node s to node t through node j; The expression of the closeness centrality of each node in the river network complex network model is as follows: Where N is the total number of nodes, d sj is the shortest path length between node s and node j.

4. The method for evaluating the robustness of a river network based on complex network theory according to claim 1 is characterized in that: The removal attack in S5 further includes static attack and dynamic attack, and further includes static attack analysis, by removing different nodes one by one and observing the changes in network indicators, identifying the nodes that will cause a significant decrease in the maximum connected subgraph ratio or global efficiency after removal; dynamic attack analysis, sorting the nodes based on different centrality indicators and gradually removing them, observing the changing trend of the overall robustness level of the network, so as to determine the evaluation indicators that have the most significant impact on the global robustness of the network.

5. The method for evaluating river network robustness based on complex network theory according to claim 1 is characterized in that: The S5 further includes: assigning weights based on the edge betweenness centrality and the flood control capacity index, water supply capacity index and ecological capacity index calculated in S2, sorting the river sections according to the node centrality index, performing edge removal attack simulation on the sorted river networks, and calculating the maximum connected subgraph ratio and global efficiency of the network after removing the river sections.

6. The method for evaluating the robustness of a river network based on complex network theory according to claim 1 is characterized in that: in, The expression of the maximum connected subgraph ratio σ is as follows: Where S is the number of nodes contained in the maximum connected subgraph, S=max(Nj'), and Nj' is the set of the number of nodes in the connected subgraph after the network node j' is destroyed.

7. The method for evaluating the robustness of a river network based on complex network theory according to claim 1 is characterized in that: in, The expression of network efficiency is as follows: Where: E j'j =1 / d j'j , dj'j is the shortest path length between node j' and node j, and N is the total number of nodes in the network.

Images