A congestion pre-control range division method based on local node importance

Abstract

The application discloses a congestion pre-control range division method based on local node importance, and comprises the following steps: firstly, determining a congestion early warning starting threshold value through congestion sample analysis; secondly, determining a congestion pre-control node set based on an adjacency matrix for a road section reaching the early warning threshold value; then, defining a local node importance evaluation index capable of representing the congestion road section in combination with a correlation coefficient of a traffic state associated road section, and performing node importance evaluation by using an analytic hierarchy process; and finally, dividing the local node importance level by using a K-Means++ Medoids clustering algorithm, and determining a graded pre-control range. The application has higher application value by identifying potential congestion points, finding a graded congestion pre-control range, improving road network operation efficiency, and reducing congestion control cost.

Description

Technical Field

[0001] This invention belongs to the field of traffic control, and specifically relates to a method for dividing congestion pre-control range based on the importance of local nodes. Background Technology

[0002] With the continuous improvement of urbanization and the expansion of urban scale, the travel demands of urban residents are rising, leading to frequent traffic congestion. Congestion not only causes traffic delays but also exacerbates environmental pollution, increases the risk of traffic accidents, and poses a significant challenge to urban traffic management. Once traffic congestion occurs, even with congestion mitigation measures, it is difficult to completely avoid the negative impacts. Against this backdrop, this invention aims to identify potential congestion points and determine the scope of congestion prevention and control based on road traffic condition prediction data. This breaks through the limitations of traditional congestion mitigation measures, which rely on post-event control, and becomes a core support for proactively controlling congestion. Regarding the current state of research on congestion control scope, existing methods for determining congestion control scope mainly suffer from the following problems:

[0003] (1) The research on congestion warning is not in-depth enough, and it is easy to mistakenly activate congestion warning due to occasional fluctuations in traffic flow. It fails to distinguish between congestion that can dissipate on its own in a short time and congestion events that require pre-control measures, resulting in waste of resources and possible traffic disruption.

[0004] (2) Lack of differentiation in congestion control scope. Most studies use clustering methods to divide areas into zones, so that regional control can be carried out on a zone-by-zone basis after congestion is identified. This division method generally results in a large coverage area for the zones and ignores the differences in the specific causes and characteristics of congestion within different zones, which does not meet the needs of refined management.

[0005] (3) The importance of nodes in congested road segments in congestion propagation was not evaluated. Some studies have identified important nodes in the entire road network through node importance evaluation, but the evaluation indicators selected usually consider the global level of the road network and cannot characterize the importance of each node to the congested road segment. Therefore, correctly identifying and distinguishing which nodes are key nodes related to congested road segments and taking different control measures for nodes of different levels is an important prerequisite for achieving precise and efficient congestion control. Summary of the Invention

[0006] Objective: To address the technical problems mentioned in the background section, this invention constructs a method for delineating congestion pre-control ranges based on the importance of local nodes, using road traffic state prediction data. Through congestion sample analysis, a congestion warning activation threshold is determined. For road segments reaching the warning threshold, a set of congestion pre-control nodes is determined based on an adjacency matrix. An evaluation index characterizing the importance of local nodes for congested road segments is defined, and the analytic hierarchy process (AHP) is used to evaluate node importance. The K-Means++Medoids clustering algorithm is used to classify the importance levels of local nodes. Compared to traditional methods, the congestion pre-control range determination and classification method designed in this invention is superior primarily in that: through continuous observation of traffic state labels, to avoid erroneous congestion warning activation due to occasional traffic flow fluctuations, a clear trigger condition is defined. When moderate or severe congestion is detected for more than 10 minutes, an warning is activated, avoiding unnecessary intervention that wastes resources and may cause traffic disruption. Compared to traditional methods that use clustering to identify traffic zones and determine control areas, this invention defines local node density, local node betweenness, and local node strength for congested road sections. Combining node degree and clustering coefficients, it quantifies the importance of pre-control nodes to congestion propagation. Furthermore, it designs a K-Means++Medoids algorithm to determine a three-tiered pre-control range, transforming the evaluation of node local importance from continuous quantitative assessment to non-continuous qualitative grading. This method provides a scientific basis and methodological support for developing differentiated congestion pre-control measures, making congestion pre-control more precise and targeted.

[0007] Technical Solution: To achieve the above technical objectives, this invention proposes a method for dividing congestion pre-control range based on the importance of local nodes. This method includes the following steps:

[0008] (1) Identify road segments that meet the congestion warning threshold based on the predicted traffic conditions;

[0009] (2) Delineate the initial control range based on the adjacency matrix and obtain the set of pre-controlled nodes;

[0010] (3) Considering the impact of each node on the speed of congestion propagation, define a local node importance evaluation index for congested road sections;

[0011] (4) Calculate the importance evaluation value of local nodes using the analytic hierarchy process;

[0012] (5) Design the K-Means++Medoid clustering algorithm and classify the nodes according to their importance values ​​based on the pre-obtained control range.

[0013] Furthermore, step (1) specifically includes:

[0014] (101) Observe the traffic status data of all road segments, plot the duration distribution of traffic congestion, take 10 minutes as the congestion warning threshold, and immediately activate the congestion warning for congestion events that last for 10 minutes or more.

[0015] (102) Identify road segments where the traffic condition is congested and the congestion lasts for more than 10 minutes.

[0016] Furthermore, the specific steps of step (2) are as follows:

[0017] (201) Define the road network as a directed graph G = (V, E), where V = {v1, v2, v3, ... v}. k} represents the set of all nodes in the road network, including the starting node and intermediate nodes, where k is the number of nodes, and E = {(v i ,v j Let )|i,j=0,1,2,……k} represent the set of road segments consisting of all nodes, where (v i ,v j Let G represent a road segment connected by nodes i and j. Each edge in G is assigned a weight, which represents the distance between road segments between each road node. The adjacency matrix A[i,j] of the connection relationship between each node and the distance matrix D[i,j] of the road segments between nodes are established, as follows:

[0018]

[0019] D[i,j]=d ij

[0020] Among them, a ij d represents the elements corresponding to the i-th row and j-th column of matrix A[i,j]; ij This represents the distance between the i-th road segment and the j-th road segment;

[0021] (202) After determining the road segment where the congestion detector is located, extract the congested road segment (v) based on the adjacency matrix. m ,v n The first-level adjacent nodes of ) that satisfy a mo ∪a om =1 and a np ∪a pn =1 for nodes, where o, p∈V; extract the second-level adjacent nodes using the same method based on the first-level adjacent nodes, i.e., a oq ∪a oq =1 and a pr ∪a rp For nodes with a value of 1, where r, q ∈ V, and so on, extract the three layers of adjacent nodes, i.e., a rs ∪a sr =1 and aqt ∪a tq Nodes with a value of 1, where s,t∈V. The extracted three-layer adjacent nodes are used as the pre-control node set.

[0022] Furthermore, the specific steps of step (3) are as follows:

[0023] (301) Define local node importance identification parameters for congested road sections, and combine node degree and node density to construct local node importance evaluation indicators. The definitions and calculations of various indicators are as follows:

[0024] (1) Local node density: In the traffic network structure, node density is usually represented by the reciprocal of its connectivity, which reflects the density of each node itself. In the local node density of congested road sections, the reciprocal of the average path length from each node to the congested node is used to represent the local density of each node to the congested node.

[0025]

[0026] Where, d im B represents the average distance of all paths between node i and congested node m; i This represents the local node density of the i-th node.

[0027] (2) Local node strength: Node strength reflects the strength of the traffic flow carried by a node. The connection weight between nodes uses the average traffic flow of the road between the two nodes. In the local node strength of congested road segments, in order to reflect the strength of the traffic flow carried by a node and the traffic flow related to the congested road segment, the influence of the congested road segment on the node is defined by using the congestion correlation between the traffic state related road segments based on the original node strength definition. That is, when the node and the congested node are not geographically adjacent, the product of the traffic state correlation coefficient and the average traffic flow between the two points is used as the local node strength.

[0028] S i1 =∑ i∈V a im w im / 2

[0029] S i2 =∑r im w im / 2

[0030] S i =S i1 +S i2

[0031] Among them, w im The weights between nodes are represented by the average lane flow at nodes i and m; r imLet i be the traffic state correlation coefficient between nodes i and m;

[0032] The traffic state correlation coefficient between road segments is calculated by analyzing the cross-correlation of traffic states between road segments and using the Pearson correlation coefficient formula. This method includes three steps: constructing a traffic state time series considering time lags; calculating the degree of cross-correlation between these time series; and determining the road segments associated with traffic states based on the Pearson correlation coefficient. The relevant definitions of the road segment traffic state cross-correlation analysis method are as follows:

[0033] (a) Traffic state time series: assuming road segment lane i The sequence of traffic conditions changing over time is labeled as Seq. i The sequence consists of a series of elements, each with a value of 0 or 1, where 0 represents smooth flow and 1 represents congestion.

[0034] (b) Time-lag traffic status time series: when road segment lane i The congestion situation after time t k Later, it affected the lane section. j At that time, this time interval t k Defined as the time lag between two lanes; for each lane j Its time-lag traffic state time series KSeq j The construction process is as follows: from the original congestion state time series Seq j Delete the first t k From the data points, we obtain the truncated sequence Seq. j_truncated In the truncated sequence Seq j_truncated Add k "-1" values ​​to the end to form the complete time-lag traffic state time series KSeq. j Due to the differences in the speed of traffic flow diffusion between different road segments, the time lag between each pair of road segments is also different. With the maximum correlation coefficient as the target, the optimal time lag is found for each pair of road segments by setting a loop function.

[0035] (c) Inter-segment correlation: Assume there are two road segments... i and lane j Among them, lane i The traffic state time series is Seq i And lane j Considering the time lag in traffic state time series, KSeq j When the time lag cross-correlation between two segments, i.e. TLCC(Seq) i ,KSeq j When the threshold r is exceeded, lane is considered to be...i and lane j There is a causal relationship between them, forming a set of road segments with related traffic conditions;

[0036] (d) Traffic state correlation coefficient: The traffic state correlation coefficient between two road segments is calculated using the Pearson correlation coefficient formula, as follows:

[0037]

[0038] Where, x i and y i They are lane i and lane j The congestion status value at the corresponding time; n1 is the total number of time series; Seq i for lane i Traffic state time series; KSeq j for lane j Consider time series of traffic conditions with time lag;

[0039] Assuming lane i The sequence of traffic conditions changing over time is labeled as Seq. i lane i The congestion situation after time t k Later, it affected the lane section. j A traffic state time series KSeq considering time lag was constructed. j According to TLCC(Seq) i ,KSeq j Calculate the traffic state correlation coefficient between road segments; when the time lag cross-correlation degree between two road segments, i.e., TLCC(Seq1,NSeq2), exceeds the preset threshold r, it is determined that there is a causal relationship between lane1 and lane2, which is a group of traffic state associated road segments. Select the corresponding road segments with Pearson correlation coefficient > 0.6 and construct the associated road segment matrix.

[0040] (3) Local node betweenness refers to the frequency of a node appearing on the shortest path between all node pairs. In the local node betweenness of congested road segments, the local node betweenness is defined as the frequency of a node appearing on the shortest path between all node pairs that pass through the congested nodes.

[0041]

[0042] Where, σ stm σ is the number of shortest paths from node s to node t via congested node m; stm (i) represents the number of nodes that these paths pass through;

[0043] (4) Node degree and node clustering coefficient: the degree k of the i-th node i This represents the total number of edges connected to the node; for the i-th node, the clustering coefficient C is... i The number of relations E between the neighboring nodes of the i-th node i degree k with the i-th node i The relationship between them is determined:

[0044]

[0045] Among them, E i Let k be the relation count of the neighboring nodes of the i-th node, obtained by counting the triangles of the i-th node; i Let be the degree of the i-th node.

[0046] Furthermore, the specific steps of step (4) are as follows:

[0047] (401) Due to the local node density B i Local node betweenness J i Local node strength S i The impact on the spread rate of congested nodes was considered, compared to node degree k. i and node clustering coefficient C i It can more accurately reflect the impact on the spread speed of congestion nodes. Node degree has a greater impact on the spread speed of congestion than node clustering coefficient. Construct a judgment matrix B, where the values ​​in the matrix represent the relative importance of each local node and the node evaluation index.

[0048]

[0049] (402) The judgment matrix B in the analytic hierarchy process is calculated using normalization, and the sum of its j-th column is expressed as sum. j Then the element b in the i-th row and j-th column of the matrix ij The normalized value n ij The calculation is as follows:

[0050] n ij =b ij / sum j

[0051] The normalized matrix is ​​represented as N, and its elements n ij That is, element b of the original judgment matrix. ij The results of column-normalization are as follows:

[0052]

[0053] (403) For a normalized matrix N, the summation by row yields a vector. Let vector S represent the result of the summation by row of the normalized matrix N. Each component in S is the sum of the elements of the corresponding row of the matrix.

[0054] (404) Normalize S to obtain the eigenvector ω = (0.16, 0.05, 0.26, 0.26, 0.26). T Let ω be an eigenvector of matrix B. Find the largest eigenvalue λ of the eigenvectors. max The calculation formula is as follows:

[0055]

[0056] Where: λ max ω is the largest eigenvalue of the matrix being judged; n is the order of the matrix being judged; ω is the largest eigenvalue of the matrix being judged. i These are elements in the eigenvector ω;

[0057] (405) Perform a consistency check to evaluate the consistency of the judgment matrix. If the consistency of the judgment matrix is ​​not acceptable, re-evaluate the elements of the matrix using the following formula:

[0058]

[0059] Where CI < 0.1, the judgment matrix is ​​considered to have acceptable consistency; CI ≥ 0.1, the consistency is considered insufficient, and the judgment matrix needs to be re-examined and adjusted.

[0060] (406) The weight coefficients of each eigenvalue index obtained from the analytic hierarchy process are ω = (0.16, 0.05, 0.26, 0.26, 0.26). T Calculate the local node importance evaluation matrix R according to the following formula:

[0061]

[0062] By combining the characteristic index values ​​of each node and their corresponding weights, the importance of each node is quantitatively scored to obtain the local node importance evaluation value dataset X.

[0063] Furthermore, the specific steps of step (5) are as follows:

[0064] (501) Randomly select a sample point from the local node importance evaluation value dataset X as the first center point;

[0065] (502) Calculate the shortest distance D(x) from each point to the current center point, and determine the probability P(x) of selecting the next center point based on this distance. The farther the point is, the higher the probability of it being selected as the new center.

[0066] (503) Iteratively update the center points until k initial cluster centers are selected;

[0067]

[0068] (504) Within each cluster, calculate an actual sample point x for that cluster. s The cluster E is determined by selecting the point with the smallest sum of squared distances to all points within the cluster. l New cluster center u l ;

[0069]

[0070] Where, x s Let x be the candidate new cluster centers, and x be the sample points of the cluster;

[0071] (505) Based on these selected cluster centers, calculate the distance between each sample and all cluster centers, and assign it to the nearest cluster such that each sample point x is within a cluster center μ. l The minimum SSE is found in the formula below:

[0072]

[0073] (406) Repeat steps (404) and (405) until the cluster centers are stable or the maximum number of iterations is reached;

[0074] (407) Based on the silhouette coefficient and the minimum sum of squared errors, the optimal number of clusters is determined to be 3. Based on the node importance value, the pre-control range is classified into levels.

[0075] Beneficial effects: Compared with the prior art, the technical solution of the present invention has the following beneficial technical effects:

[0076] The congestion pre-control range determination and classification method designed in this invention, compared with traditional methods, is superior in that: by continuously observing traffic status labels, a clear trigger condition is defined to avoid erroneous congestion warnings due to occasional traffic flow fluctuations. When moderate or severe congestion is detected for more than 10 minutes, an warning is activated, avoiding unnecessary intervention that wastes resources and may disrupt traffic. Compared to the traditional method of determining the control range by clustering to find traffic zones, this invention defines local node density, local node betweenness, and local node strength for congested road segments. Combining node degree and node clustering coefficient, it quantifies the importance of pre-control nodes to congestion propagation and designs a K-Means++Medoids algorithm to determine the three-level pre-control range, realizing the transformation of the local importance of nodes from continuous quantitative evaluation to non-continuous qualitative classification. This method provides a scientific basis and methodological support for formulating differentiated congestion pre-control measures, making congestion pre-control more accurate and targeted. Attached Figure Description

[0077] Figure 1 This is a flowchart of the method of the present invention;

[0078] Figure 2 This describes the distribution of the duration of congestion events in this invention;

[0079] Figure 3 This is a comparison chart of lane flow and traffic status for the detection station number "G103L221110112" in this invention;

[0080] Figure 4 This is a schematic diagram of the congestion pre-control node set in this invention;

[0081] Figure 5 This is the correlation coefficient matrix of the road sections associated with the detection points in Tongzhou District in this invention;

[0082] Figure 6 This is a graph showing the variation of the silhouette coefficient (SC) with the number of clusters in this invention;

[0083] Figure 7 This is a graph showing the variation of the sum of squared errors (SSE) with the number of clusters in this invention;

[0084] Figure 8 This is a schematic diagram illustrating the division of the congestion three-level pre-control range based on node importance in this invention. Detailed Implementation

[0085] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings.

[0086] The present invention proposes a method for dividing the congestion pre-control range based on the importance of local nodes, as follows: Figure 1 As shown, it includes the following steps:

[0087] (1) Based on the predicted road traffic status data of roads at all levels in Tongzhou District, Beijing, the distribution of the duration of congestion events is plotted to determine a reasonable threshold for congestion warnings, such as... Figure 2 As shown in Table 1, the statistical values ​​of congestion frequency distribution indicators are as follows:

[0088] Table 1. Statistics on Congestion Frequency Distribution Indicators

[0089]

[0090]

[0091] Combination Figure 2 As shown in Table 1, since the detector's statistical data interval is 5 minutes, the minimum congestion duration in the sample is 5 minutes, and this type of congestion sample accounts for 69% of the total congestion samples. This type of congestion can dissipate on its own within a short time (5 minutes) and will not cause serious impact, so no congestion pre-control measures are required. Therefore, when designing the warning activation threshold, 10 minutes is used as a key time node. For congestion events that reach or exceed this threshold, a congestion warning is immediately activated.

[0092] (2) Define the road network as a directed graph G = (V, E), where V = {v1, v2, v3, ... v}. n} represents the set of all nodes in the road network, including the starting node and intermediate nodes, E = {(v i ,v j Let )|i,j=0,1,2,……n} represent the set of road segments consisting of all nodes, where (v i ,v j Let G represent a road segment connected by nodes i and j. Each edge in G is assigned a weight, which represents the distance between road segments between each road node. Establish the adjacency matrix A[i,j] for the connection relationship between each node and the distance matrix D[i,j] for the road segments between nodes.

[0093]

[0094] D[i,j]=d ij

[0095] Among them, a ij d represents the elements corresponding to the i-th row and j-th column of matrix A[i,j]; ij This represents the distance between road segment i and road segment j. A traffic status distribution map is drawn using the monitoring station number "G103L221110112" which has a high frequency of congestion. Its traffic status labels 0 to 3 represent the range from smooth traffic to severe congestion, respectively. Figure 3As shown in the diagram, a congestion event was detected at the black dashed line, triggering a congestion warning. After determining that the congested road segment where detector "G103L221110112" is located is node 12-23, the first-level adjacent nodes of the congested road segment are extracted based on the adjacency matrix. Then, the second-level adjacent nodes are extracted using the same method, and so on, until the third-level adjacent nodes are extracted. These adjacent road segments are used as the initial control range to obtain the set of pre-control nodes to be calculated. The initially determined pre-control range is shown below. Figure 4 As shown.

[0096] (3) Calculate the local node importance evaluation value based on the definition of the local node importance evaluation index for congested road sections. Obtain the basic road network parameter node distance, as shown in Table 2:

[0097] Table 2 shows the distances between some nodes in the road network.

[0098]

[0099]

[0100] For each lane i The construction process of the time-lag congestion state time series NSeqi is as follows: The first k data points are deleted from the original congestion state time series Seqi to obtain the truncated sequence Seqi_truncated, ensuring that the time-lag sequence accurately reflects the change in road segment congestion status after a time lag of k units. K "-1" values ​​are added to the end of the truncated sequence Seqi_truncated to form the complete time-lag congestion state time series NSeqi. Adding "-1" ensures that the length of the time series is consistent with the original series, and also serves as a marker value to distinguish between the original data and the padding data added due to the time lag.

[0101] After constructing the traffic state time series Seq1 for lane 1 and the time-lag-considered traffic state time series NSeq2 for lane 2, the cross-correlation between these time series is calculated based on the time-lag cross-correlation coefficient TLCC(Seq1,NSeq2). When TLCC(Seq1,NSeq2) exceeds a preset threshold r, a causal relationship can be identified between lane 1 and lane 2, constituting a set of traffic state-related road segments. The correlation coefficient matrix of traffic state-related road segments at the Tongzhou District detection points is plotted as follows: Figure 5 As shown, road segments with a Pearson correlation coefficient > 0.6 were selected as traffic state-related road segments.

[0102] The relevant eigenvalues ​​of the importance evaluation index for each local node are calculated, and then normalized and scaled to between 0 and 1 to eliminate the influence of different eigenvalues ​​on the same scale, ensuring that they are comparable on the same measure. The normalized eigenvalues ​​of the nodes are shown in Table 3.

[0103] Table 3 shows the normalized eigenvalues ​​of the nodes.

[0104]

[0105]

[0106] Based on the pairwise comparison method and introducing a nine-point proportional scale, Table 4 shows the quantified values ​​of the proportional scale and the meaning of the weight ratios they represent:

[0107] Table 4. Standardized values ​​of the proportion

[0108]

[0109] When the relative importance of evaluation indicators is 1, it indicates that the two indicators are equally important; a value of 3 indicates that one indicator is slightly more important than the other; and a value of 9 indicates that one indicator is far more important than the other. For cases where indicators are between adjacent evaluations, four intermediate values ​​of 2, 4, 6, and 8 are assigned. Based on this rule, the elements b in the judgment matrix B are... ij The importance of the i-th indicator relative to the j-th indicator is defined. Since local node strength, local betweenness, and local density consider the impact on the congestion spread rate, they are more important than node degree and clustering coefficient. Node degree, compared to density, has a greater impact on the congestion spread rate. Based on the analytic hierarchy process, the judgment matrix B is constructed as shown in Table 5.

[0110] Table 5 shows the judgment matrix.

[0111]

[0112] The judgment matrix in the analytic hierarchy process is calculated using the summation method and normalization. Specifically, if the judgment matrix is ​​B, the sum of its j-th column is represented as sum. j Then the element b in the i-th row and j-th column of the matrix ij The normalized value ni j The calculation is as follows:

[0113] n ij =b ij / sum j

[0114] The normalized matrix, denoted as N, has n elements. ij That is, element b of the original judgment matrix. ij The results of column-normalization are as follows:

[0115]

[0116] For a normalized matrix N, summing by rows yields a vector, where each component is the sum of the elements in the corresponding row of the matrix. Let vector S represent the result of summing by rows of the normalized matrix N, then each component S... i The calculation formula is:

[0117]

[0118] Performing the above calculations will yield the overall weight S for each node.

[0119] Normalizing S yields the eigenvectors:

[0120] ω=(ω1,ω2,…,ω n ) T

[0121] The calculated eigenvector is ω = (0.16, 0.05, 0.26, 0.26, 0.26). T Let ω be an eigenvector of matrix A, and then find the largest eigenvalue λ of the eigenvectors. max The calculation formula is as follows:

[0122]

[0123] The largest eigenvalue λ is calculated. max It is 5.01.

[0124] Consistency testing is an important step in the analytic hierarchy process (AHP) used to assess the consistency of the decision matrix. If the consistency of the decision matrix is ​​unacceptable, the elements of the matrix must be re-evaluated, as shown in the following formula:

[0125]

[0126] Where: λ max is the largest eigenvalue of the judgment matrix; n is the order of the judgment matrix.

[0127] If CI < 0.1, the judgment matrix is ​​considered to have acceptable consistency; if CI ≥ 0.1, the consistency is considered insufficient, and the judgment matrix needs to be re-examined and adjusted. ω = (0.16, 0.05, 0.26, 0.26, 0.26) T The consistency test results are as follows:

[0128]

[0129] CI < 0.1 indicates that the set judgment matrix is ​​reasonable and passes the consistency test. Based on the results calculated using the analytic hierarchy process (AHP), the corresponding weight coefficients for each eigenvalue index are shown below:

[0130]

[0131] Calculate the local node importance evaluation matrix R according to the following formula:

[0132]

[0133] Based on the graph structure drawn above and the obtained feature values ​​of each node, substituting them into the importance evaluation matrix R, we can obtain the importance evaluation values ​​of each node as shown in Table 6:

[0134] Table 6 shows the evaluation values ​​for the importance of local nodes.

[0135]

[0136]

[0137] (4) After obtaining the local node importance values ​​calculated for congested nodes, to facilitate hierarchical control, the node importance needs to be converted from quantitative values ​​to hierarchical qualitative ratings. Different control levels are applied based on varying degrees of congestion. Based on the local node importance data of the road network, the K-Means++Medoid algorithm is used to determine the node importance classification levels. Clustering numbers ranging from 2 to 6 are used, and the SC and SSE values ​​are employed to analyze the clustering effect. A comparison of the evaluation indicators for different clustering numbers is shown in Table 7.

[0138] Table 7 shows the performance evaluation indicators for different numbers of clusters.

[0139]

[0140] Plot the changes in the silhouette coefficient index for different numbers of clusters, such as... Figure 6 As shown in Table 7. The silhouette coefficient ranges from -1 to 1, with higher values ​​indicating better clustering results. Figure 6 The data shows that the SC value is the highest (0.62) when the number of clusters is 3, which indicates that the clustering effect is best when the number of clusters is 3.

[0141] Plot the squared error and its variation for different numbers of clusters, such as... Figure 7 As shown in Table 8 and Figure 7Based on the data, SSE decreases with increasing cluster size because more cluster centers result in shorter distances between samples and their nearest cluster centers. The rate of decrease slows down when SSE values ​​change from 3 to 4 compared to the change from 2 to 3, indicating that a cluster size of 3 represents the "elbow" of SSE. Combining the analysis results of SC and SSE, the optimal number of clusters is 3. After determining the number of clusters, the K-Means++Medoids clustering algorithm is used to classify node importance into three levels, dividing the pre-control nodes into three tiers. The congestion pre-control range based on local node importance is plotted, as shown below. Figure 8 As shown in Table 8, the importance levels of nodes are classified.

[0142] Table 8 Results of the classification of importance levels of different nodes

[0143]

[0144] (5) After predicting that severe congestion will occur on the road section where nodes 12-23 are located at around 6:30 am, the corresponding level of control area will be activated based on the congestion level assessment results (moderate or severe congestion). The first-level control area (nodes 29, 8, 9, 22, 11, 10, 41) will be designated as a high-priority control area, the second-level control area (nodes 15, 14, 20, 6, 36, 31, 27, 21, 3, 24, 34, 25, 32, 13, 37, 43, 53, 38, 46, 45) as a medium-level control area, and the third-level control area (nodes 5, 4, 1, 2, 19, 26, 35, 33, 54, 39, 30, 17, 42, 18) as an auxiliary control area. Temporary control measures will be taken in advance to avoid congestion.

[0145] The embodiments are merely illustrative of the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made to the technical solution based on the technical concept proposed in this invention shall fall within the scope of protection of this invention.

Claims

1. A method for dividing congestion pre-control range based on the importance of local nodes, characterized in that, The method includes the following steps: (1) Identify road segments that meet the congestion warning threshold based on the predicted traffic conditions; (2) Delineate the initial control range based on the adjacency matrix and obtain the set of pre-controlled nodes; (3) Considering the impact of each node on the speed of congestion propagation, define a local node importance evaluation index for congested road sections; (4) Calculate the importance evaluation value of local nodes using the analytic hierarchy process; (5) Design the K-Means++Medoid clustering algorithm and classify the nodes according to their importance values ​​based on the preliminary control range; The specific steps of step (3) are as follows: (301) Define local node importance identification parameters for congested road sections, and combine node degree and node density to construct local node importance evaluation indicators. The definitions and calculations of various indicators are as follows: (1) Local node density In the structure of a transportation network, node density is usually represented by the reciprocal of its connectivity, which reflects the density of each node itself. In the local node density of a congested road segment, the reciprocal of the average path length from each node to the congested node is used to represent the local density of each node relative to the congested node. Where, d im B represents the average distance of all paths between node i and congested node m; i Represents the local node density of the i-th node; (2) Local node strength Node strength reflects the strength of the traffic flow a node can handle. The connection weight between nodes uses the average traffic flow of the road between the two nodes. In the local node strength for congested road segments, in order to reflect the strength of the traffic flow a node can handle in relation to the congested road segment, the influence of the congested road segment on the node is defined based on the original node strength definition and the congestion correlation between the traffic state-related road segments. That is, when a node and a congested node are not geographically adjacent, the product of the traffic state correlation coefficient and the average traffic flow between the two points is used as the local node strength. S i1 =∑ i∈V a im w im / 2 S i2 =∑r im w im / 2 S i =S i1 +S i2 Among them, w im It represents the weight between nodes, expressed as the average lane flow between node i and congested node m; r im Let be the traffic state correlation coefficient between node i and congested node m; The traffic state correlation coefficient between road segments is calculated by analyzing the cross-correlation of traffic states between road segments and using the Pearson correlation coefficient formula. This method includes three steps: constructing a traffic state time series considering time lags; calculating the degree of cross-correlation between these time series; and determining the road segments associated with traffic states based on the Pearson correlation coefficient. The relevant definitions of the road segment traffic state cross-correlation analysis method are as follows: (a) Traffic state time series: assuming road segment lane i The sequence of traffic conditions changing over time is labeled as Seq. i The sequence consists of a series of elements, each with a value of 0 or 1, where 0 represents smooth flow and 1 represents congestion. (b) Time-lag traffic status time series: when road segment lane i The congestion situation after time t k Later, it affected the lane section. j At that time, this time interval t k Defined as the time lag between two lanes; for each lane j Its time-lag traffic state time series KSeq j The construction process is as follows: from the original congestion state time series Seq j Delete the first t k From the data points, we obtain the truncated sequence Seq. j_truncated In the truncated sequence Seq j_truncated Add k "-1" values ​​to the end to form the complete time-lag traffic state time series KSeq. j The speed of traffic flow diffusion varies between different road segments, and the time lag between each road segment pair is different. With the maximum correlation coefficient as the target, the optimal time lag is found for each road segment pair by setting a loop function. (c) Inter-segment correlation: Assume there are two road segments... i and lane j Among them, lane i The traffic state time series is Seq i And lane j Considering the time lag in traffic state time series, KSeq j When the time lag cross-correlation between two segments, i.e. TLCC(Seq) i ,KSeq j When the threshold r is exceeded, lane is considered to be... i and lane j There is a causal relationship between them, forming a set of road segments with related traffic conditions; (d) Traffic state correlation coefficient: The traffic state correlation coefficient between two road segments is calculated using the Pearson correlation coefficient formula, as follows: Where, x i and y i They are lane i and lane j The congestion status value at the corresponding time; n1 is the total number of time series; Seq i for lane i Traffic state time series; KSeq j for lane j Consider time series of traffic conditions with time lag; Assuming lane i The sequence of traffic conditions changing over time is labeled as Seq. i lane i The congestion situation after time t k Later, it affected the lane section. j A traffic state time series KSeq considering time lag was constructed. j According to TLCC(Seq) i ,KSeq j ) Calculate the traffic state correlation coefficient between road segments; when the time lag cross-correlation degree between two road segments, i.e. TLCC(Seq1,NSeq2), exceeds the preset threshold r, it is determined that there is a causal relationship between lane1 and lane2, which is a group of traffic state associated road segments. Select the corresponding road segments with Pearson correlation coefficient > 0.6 and construct the associated road segment matrix. (3) Local node betweenness refers to the frequency of a node appearing on the shortest path between all node pairs. In the local node betweenness of congested road segments, the local node betweenness is defined as the frequency of a node appearing on the shortest path between all node pairs that pass through the congested nodes. Where, σ stm σ is the number of shortest paths from node s to node t via congested node m; stm (i) represents the number of nodes that the path passes through; (4) Node degree and node clustering coefficient: the degree k of the i-th node i This represents the total number of edges connected to the node; for the i-th node, the clustering coefficient C is... i The number of relations E between the neighboring nodes of the i-th node i degree k with the i-th node i The relationship between them is determined: Among them, E i Let k be the relation count of the neighboring nodes of the i-th node, obtained by counting the triangles of the i-th node; i Let be the degree of the i-th node.

2. The method for dividing congestion pre-control range based on local node importance according to claim 1, characterized in that, Step (1) specifically includes: (101) Observe the traffic status data of all road segments, plot the duration distribution of traffic congestion, take 10 minutes as the congestion warning threshold, and immediately activate the congestion warning for congestion events that last for 10 minutes or more. (102) Identify road segments where the traffic condition is congested and the congestion lasts for more than 10 minutes.

3. The method for dividing congestion pre-control range based on local node importance according to claim 2, characterized in that, The specific steps of step (2) are as follows: (201) Define the road network as a directed graph G = (V, E), where V = {v1, v2, v3, ... v}. k } represents the set of all nodes in the road network, including the starting node and intermediate nodes, where k is the number of nodes, and E = {(v i ,v j Let )|i,j=0,1,2,……k} represent the set of road segments consisting of all nodes, where (v i ,v j Let G represent a road segment connected by nodes i and j. Each edge in G is assigned a weight, which represents the distance between road segments between each road node. The adjacency matrix A[i,j] of the connection relationship between each node and the distance matrix D[i,j] of the road segments between nodes are established, as follows: D[i,j]=d ij Among them, a ij d represents the elements corresponding to the i-th row and j-th column of matrix A[i,j]; ij This represents the distance between the i-th road segment and the j-th road segment; (202) After determining the road segment where the congestion detector is located, extract the congested road segment (v) based on the adjacency matrix. m ,v n The first-level adjacent nodes of ) that satisfy a mo ∪a om =1 and a np ∪a pn =1 for nodes, where o, p∈V; extract the second-level adjacent nodes using the same method based on the first-level adjacent nodes, i.e., a oq ∪a oq =1 and a pr ∪a rp For nodes with a value of 1, where r, q ∈ V, and so on, extract the three layers of adjacent nodes, i.e., a rs ∪a sr =1 and a qt ∪a tq =1 for nodes, where s,t∈V, and the extracted three-layer adjacent nodes are used as the pre-control node set.

4. The method for dividing congestion pre-control range based on local node importance according to claim 1, characterized in that, The specific steps of step (4) are as follows: (401) Construct a judgment matrix B, where the values ​​in the matrix represent the relative importance of the evaluation indicators of each local node importance node; (402) The judgment matrix B in the analytic hierarchy process is calculated using normalization, and the sum of its j-th column is expressed as sum. j The element b in the i-th row and j-th column of the matrix ij The normalized value n ij The calculation is as follows: n ij =b ij / sum j The normalized matrix is ​​represented as N, and its elements n ij That is, element b of the original judgment matrix. ij The results of column-normalization are as follows: (403) For a normalized matrix N, the summation by row yields a vector. Let vector S represent the result of the summation by row of the normalized matrix N. Each component in S is the sum of the elements of the corresponding row of the matrix. (404) Normalize S to obtain the eigenvector ω = (0.16, 0.05, 0.26, 0.26, 0.26). T Let ω be an eigenvector of matrix B. Find the largest eigenvalue λ of the eigenvectors. max The calculation formula is as follows: Where, λ max ω is the largest eigenvalue of the matrix being judged; n is the order of the matrix being judged; ω is the largest eigenvalue of the matrix being judged. i These are elements in the eigenvector ω; (405) Perform a consistency check to evaluate the consistency of the judgment matrix. If the consistency of the judgment matrix is ​​not acceptable, re-evaluate the elements of the matrix using the following formula: Where: if CI < 0.1, the judgment matrix is ​​considered to have acceptable consistency; if CI ≥ 0.1, the consistency is considered insufficient, and the judgment matrix is ​​readjusted. (406) The weight coefficients of each eigenvalue index obtained from the analytic hierarchy process are ω = (0.16, 0.05, 0.26, 0.26, 0.26). T Calculate the local node importance evaluation matrix R according to the following formula: By combining the characteristic index values ​​of each node and their corresponding weights, the importance of each node is quantitatively scored to obtain the local node importance evaluation value dataset X.

5. The method for dividing congestion pre-control range based on local node importance according to claim 4, characterized in that, The specific steps of step (5) are as follows: (501) Randomly select a sample point from the local node importance evaluation value dataset X as the first center point; (502) Calculate the shortest distance D(x) from each point to the current center point, and determine the probability P(x) of selecting the next center point based on this distance. The farther the point is, the higher the probability of it being selected as the new center. (503) Iteratively update the center points until k initial cluster centers are selected; (504) Within each cluster, calculate an actual sample point x for that cluster. s The cluster E is determined by selecting the point with the smallest sum of squared distances to all points within the cluster. l New cluster center u l ; Where, x s Let x be the sample point of the cluster, and dist(x) be the candidate new cluster center. s x) is x s The distance between x and x; (505) Based on these selected cluster centers, calculate the distance between each sample and all cluster centers, and assign it to the nearest cluster such that each sample point x is within a cluster center μ. l The minimum SSE is found in the formula below: (506) Repeat steps (504) and (505) until the cluster centers are stable or the maximum number of iterations is reached; (507) Based on the silhouette coefficient and the minimum sum of squared errors, the optimal number of clusters is determined to be 3. Based on the node importance value, the pre-control range is classified into levels.

Images