Real-time monitoring method for traffic state of expressway reconstruction and expansion construction area
By integrating time series prediction with a multi-segment spatiotemporal correlation correction mechanism, and using the Hungarian algorithm to match the optimal time series, the system automatically matches the most similar historical pattern clusters by combining the historical traffic flow similarity and time delay between the target segment and adjacent segments. This solves the reliability problem of real-time traffic condition monitoring in highway reconstruction and expansion construction areas, and achieves efficient real-time traffic condition monitoring and decision support.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- WUHAN ZHONGJIAO TRAFFIC ENG CO LTD
- Filing Date
- 2026-05-09
- Publication Date
- 2026-06-05
AI Technical Summary
Traditional static or low-frequency monitoring methods are unable to accurately capture the rapid evolution of traffic conditions in highway reconstruction and expansion construction areas, leading to decreased traffic efficiency and increased safety risks. The prediction results of existing prediction algorithms have low reliability.
By integrating time series prediction with multi-segment spatiotemporal correlation correction mechanism, the Hungarian algorithm is used to match the optimal time series. Combining the historical traffic flow similarity and time delay between the target road segment and adjacent road segments, the most similar historical pattern cluster is automatically matched, the reliability of the time delay is evaluated, and traffic flow prediction is corrected.
It significantly improves the reliability of traffic flow prediction results, can promptly identify changes in traffic conditions in construction areas, provide effective decision support, and reduce safety risks.
Smart Images
Figure CN122157495A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of condition monitoring technology, specifically to a method for real-time traffic condition monitoring in highway reconstruction and expansion construction areas. Background Technology
[0002] Expansion and reconstruction projects are typically carried out while maintaining existing traffic operations. Construction areas often face complex situations such as road segment compression, alignment adjustments, temporary speed limits, and traffic flow reorganization. These changes significantly alter the original road characteristics, easily leading to decreased traffic efficiency, abnormal vehicle behavior, and even increased safety risks. In such dynamic and constrained traffic environments, traditional static or low-frequency monitoring methods struggle to accurately capture the rapid evolution of traffic conditions, failing to provide timely and effective decision support for on-site management. However, the development of modern intelligent transportation technologies has made it possible to construct a highly responsive and adaptable real-time monitoring system. By integrating multiple sensing methods and intelligent analysis techniques, it is hoped that continuous and accurate perception of traffic flow conditions in construction areas (such as capacity, speed, congestion levels, and abnormal events) can be achieved. Therefore, there is an urgent need to research a real-time traffic condition monitoring method suitable for highway expansion and reconstruction construction areas, capable of adapting to constantly changing road conditions and traffic environments during construction, effectively identifying operational trends and potential risks, and providing reliable technical support for traffic organization optimization, safety early warning, and collaborative management during the construction period.
[0003] Traffic flow is highly variable in real time. Predictions based solely on existing algorithms have low reliability. Therefore, it is necessary to analyze other factors and correct the prediction results to improve the reliability of road traffic flow predictions obtained using algorithms. Summary of the Invention
[0004] To address the technical problem of low accuracy in traffic flow prediction, this application provides a real-time traffic condition monitoring method for highway reconstruction and expansion construction areas. The specific technical solution adopted is as follows:
[0005] This application proposes a method for real-time traffic monitoring in highway reconstruction and expansion construction areas, which includes the following steps:
[0006] The collected vehicle traffic flow constitutes a traffic flow time series;
[0007] The traffic flow time series of the target road segment is predicted by the prediction algorithm to obtain the corrected traffic flow for the next time step from the current time; the correlation between the target road segment and its neighboring road segments is determined based on the traffic flow differences at the same time on multiple historical days.
[0008] Based on historical traffic flow and time, the target road segment and each neighboring road segment are matched to obtain the time delay at each moment; for the target road segment at the same time on different days, the traffic flow of neighboring road segments is clustered; the reliability of the time delay is determined by the inter-class and intra-class differences of the clusters.
[0009] The current time delay of the target road segment and its neighboring road segments is filtered by the real-time traffic flow difference between the target road segment and the cluster. The predicted correction reference value is determined based on historical traffic flow changes and the current time delay. The affected road segments are filtered based on the positive and negative relationship of the current time delay. The corrected traffic flow is obtained by weighting the traffic flow before correction and the predicted correction reference value of all affected road segments with the correlation between the target road segment and the affected road segments.
[0010] Traffic conditions are monitored in real time by adjusting traffic flow.
[0011] In the aforementioned scheme, this application effectively overcomes the problem of insufficient response of traditional single models to complex dynamic factors such as sudden traffic events and upstream and downstream congestion transmission by integrating time series prediction (exponential smoothing) with a multi-segment spatiotemporal correlation correction mechanism. Furthermore, by calculating the historical traffic flow similarity between the target road segment and multiple neighboring road segments, lanes with genuine traffic linkages are identified, avoiding the blind use of geographically proximate but functionally unrelated road segment data. Considering that the propagation of congestion or traffic flow changes in the road network takes time, and that this delay is affected by vehicle speed and traffic density, the optimal time series alignment is matched using the Hungarian algorithm, ensuring that the corrected reference value comes from historical moments that "truly affect the current state," significantly improving physical plausibility. The most similar historical pattern cluster is automatically matched based on the current traffic flow state, and the credibility of the time delay under this pattern is evaluated. Ultimately, this improves the reliability of the corrected prediction results.
[0012] In one embodiment, the correlation between the target road segment and adjacent road segments is negatively correlated with the difference in all traffic flows.
[0013] In one embodiment, the method for obtaining the time delay at each moment by matching the target road segment and each neighboring road segment based on historical traffic flow and time is as follows:
[0014] When matching the traffic flow of the target road segment with each neighboring road segment at all times, the Euclidean distance between the traffic flow and the normalized result at each time is used as the matching distance. The target road segment and the neighboring road segments are constructed into corresponding distance matrices. The Hungarian algorithm is used to match the distance matrices to obtain the matching time of the target road segment in the neighboring road segments at each time. The time difference between the target road segment and the matching time is recorded as the time delay. The mode of the time delays of all days is taken as the time delay of the road segment at each time.
[0015] In one embodiment, the method for clustering traffic flow based on neighboring road segments at the same time on different days for a target road segment is as follows:
[0016] For the target road segment and each neighboring road segment, the distance between the traffic flow of the neighboring road segments at different times corresponding to different days at each time of the target road segment is used as the clustering distance, and the historical days are clustered to obtain the clusters.
[0017] In one embodiment, the method for obtaining the inter-class differences is as follows:
[0018] For each cluster corresponding to the target road segment at each time point, the average traffic flow of the target road segment on different days within the cluster is taken as the average traffic flow of the cluster; the average absolute value of the difference between the average traffic flow of each cluster and the average traffic flow of the other clusters at each time point is taken as the inter-cluster difference of each cluster.
[0019] In one embodiment, the method for obtaining the intra-class differences is as follows:
[0020] For each cluster, calculate the absolute value of the difference between the daily traffic flow and the traffic flow on other days at the same time for the target road segment, and sum all the absolute values of the differences as the intra-cluster difference.
[0021] In one embodiment, the time delay reliability is positively correlated with inter-class differences and negatively correlated with intra-class differences.
[0022] In one embodiment, the method for filtering the current time delay of the target road segment and neighboring road segments based on the real-time traffic flow difference between the target road segment and the cluster is as follows:
[0023] The absolute value of the difference between the real-time traffic flow of the target road segment and the average traffic flow of all target road segments in the same cluster at the same time is recorded as the real-time difference. The cluster with the smallest real-time difference is taken as the reference cluster. The average time delay of all elements in the reference cluster is taken as the current time delay of the target road segment and the adjacent road segments at the current time.
[0024] In one embodiment, the method for determining the prediction correction reference value based on historical traffic flow changes and current time delay is as follows:
[0025] Based on the current time delay, find the traffic flow of the neighboring road segments corresponding to the target road segment at the same time in the reference cluster, and calculate the difference between the traffic flow of the target road segment at the same time in history and the neighboring road segments corresponding to the current time delay, which is recorded as the traffic flow change value.
[0026] The time obtained by subtracting the current time from the current time delay is taken as the corresponding time of the neighboring road segment; the sum of the traffic flow and traffic flow change value of the neighboring road segment at the next time of the corresponding time is taken as the prediction correction reference value of the neighboring road segment for the target road segment at the next time of the current time.
[0027] In one embodiment, the larger the target weight, the smaller the impact of the traffic flow before correction; the larger the target weight, the larger the correction parameter. The target weight is obtained by averaging the correlation between the target road segment and all affected road segments. The ratio of the time delay reliability of the target road segment and each affected road segment to the sum of the time delay reliability of the target road segment and all affected road segments is used as the weight of each affected road segment. The correction parameter is obtained by weighted summing of the weight and the predicted correction reference value obtained for the affected road segment.
[0028] The beneficial effects of this application are as follows:
[0029] This application effectively overcomes the shortcomings of traditional single models in responding to complex dynamic factors such as sudden traffic events and upstream / downstream congestion transmission by integrating time series prediction (exponential smoothing) with a multi-segment spatiotemporal correlation correction mechanism. Furthermore, by calculating the historical traffic flow similarity between the target road segment and multiple neighboring road segments, it identifies lanes with genuine traffic linkages, avoiding the blind use of geographically proximate but functionally unrelated road segment data. Considering the time required for congestion or traffic flow changes to propagate within the road network, and the impact of vehicle speed and traffic density, the application uses the Hungarian algorithm to match optimal time series alignment, ensuring that the corrected reference values come from historical moments that "truly affect the current state," significantly improving physical plausibility. Based on the current traffic flow state, the application automatically matches the most similar historical pattern clusters and evaluates the reliability of time delays within these patterns. Ultimately, this improves the reliability of the corrected prediction results. Attached Figure Description
[0030] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0031] Figure 1 This is a flowchart of a method for real-time traffic monitoring in a highway reconstruction and expansion construction area, provided as an embodiment of this application. Detailed Implementation
[0032] To further illustrate the technical means and effects adopted by this application to achieve the intended purpose of the invention, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the real-time traffic status monitoring method for highway reconstruction and expansion construction areas proposed in this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.
[0033] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.
[0034] Example of a real-time traffic monitoring method for highway reconstruction and expansion construction areas:
[0035] The following, in conjunction with the accompanying drawings, details the specific scheme of the real-time traffic status monitoring method for highway reconstruction and expansion construction areas provided in this application.
[0036] Please see Figure 1 The document illustrates a flowchart of a real-time traffic status monitoring method for highway reconstruction and expansion construction areas, according to an embodiment of this application. The method includes the following steps:
[0037] Step S001: Collect traffic flow data to form a traffic flow time series.
[0038] Traffic flow is monitored on various road sections using existing traffic flow monitoring equipment. Traffic flow monitoring results are obtained for each road section at different times. Traffic flow refers to the number of motor vehicles passing through a given time period. For each time period, the number of motor vehicles passing through in the preceding time period is taken as the traffic flow for that time. The daily traffic flow is then compiled into a traffic flow time series. In this embodiment, the number of motor vehicles passing through within 10 minutes is taken as the traffic flow for that monitoring time. The number of motor vehicles is collected every 10 minutes as the traffic flow for that time. The traffic flow collected the previous day at the current time is used to construct a time series as the traffic flow time series.
[0039] At this point, the traffic time series has been obtained.
[0040] Step S002: Predict traffic flow using existing prediction algorithms and determine the correlation between road segments and adjacent road segments using historical data.
[0041] Using existing time series prediction algorithms, traffic flow on a target road segment is predicted. In this embodiment, the existing exponential smoothing algorithm is used to predict the obtained traffic flow time series. The traffic flow time series of the road segment on the current day is input, and the traffic flow of the road segment at the next predicted moment is recorded as the traffic flow before correction.
[0042] Considering the continuity between different road segments and the practical factors such as traffic merging and diverging, if one adjacent road experiences congestion, the congestion may spread to the other road in the next moment. Therefore, by analyzing the historical traffic flow data of each road in real time, the relationship between the traffic flow of n adjacent road segments centered on the target road segment can be obtained. This allows for the prediction and correction of the target road segment's traffic flow based on changes in the traffic flow of other road segments. The correlation between traffic flow of different road segments is calculated accordingly. For the target road segment, road segments adjacent to the target road segment are denoted as neighboring road segments.
[0043] Therefore, the traffic flow of the n road segments closest to the target road segment is analyzed to obtain the traffic flow time series of the target road segment and its neighboring road segments over historical days T. The difference in traffic flow between road segments at the same time on the same day is calculated. The smaller the difference in traffic flow between two road segments, the stronger the correlation between the two road segments. In this embodiment, n is set to 30, and T is set to 30.
[0044] The correlation between road segments is negatively correlated with the differences in traffic flow across all road segments.
[0045] It should be noted that negative correlation means that when one variable increases, the other variable decreases accordingly, and the two variables change in opposite directions. When one variable changes from large to small or from small to large, the other variable also changes from small to large or from large to small. The specific relationship is determined by practical application, and this application does not impose any special restrictions.
[0046] Preferably, in this embodiment, the expression for the correlation between road segments is:
[0047] , This represents the difference in traffic flow between road segment A and road segment B at time j on day i. Indicates the number of historical days. Indicates the number of hours per day. Represents the normalization function. This represents the correlation between road segment A and road segment B. The difference in traffic flow is the absolute value of the difference in traffic flow, and the normalization function is the maximum-minimum normalization algorithm.
[0048] At this point, the correlation between the target road segment and all its neighboring road segments has been obtained.
[0049] Step S003: Based on historical traffic flow and time, match the target road segment and each neighboring road segment to obtain the time delay at each moment; cluster the time of different days matched by the target road segment based on traffic flow; determine the reliability of the time delay by the inter-cluster differences and intra-cluster differences of the clusters.
[0050] Considering that traffic flow varies at different monitoring times—for example, during the morning rush hour, traffic may be more concentrated on the right side of the road (e.g., if road segment A is ahead of road segment B, a larger volume of traffic on the right side of A may cause congestion on the right side of road segment B, leading to increased traffic congestion there as well. Conversely, during the evening rush hour, traffic flow may be concentrated on the left side of the road. Therefore, if there is heavy traffic on the left side of road segment B, causing congestion, it may slow down traffic on the left side of road segment A, leading to increased traffic. Thus, analyzing only the historical traffic flow correlation between two road segments may not be applicable across different time periods.
[0051] Meanwhile, considering the time-sensitive nature of the interaction between traffic flows on different road segments, and its potential correlation with vehicle speed, for example, if road segment A has low traffic flow while road segment B has high traffic flow, the time it takes for road segment B to influence the increase in traffic flow on road segment A will be longer. Conversely, if road segment A has high traffic flow while road segment B's traffic flow remains consistent, the time it takes for road segment A's traffic flow to increase its own traffic flow may be shorter than when road segment traffic flow is low. Therefore, time delay needs to be considered when making corrections. That is, when correcting the prediction result for road segment A at the next moment, its traffic flow may be affected by the detection result for road segment B at the previous moment. Thus, it is necessary to analyze the time delay of the interaction between road segment traffic flows under different conditions.
[0052] For historical data, the traffic flow of the target road segment is matched with that of any neighboring road segment at all times. The Euclidean distance between the traffic flow and the normalized result at each time moment is used as the matching distance. A distance matrix is constructed between the target road segment and the neighboring road segments, and the Hungarian algorithm is used to match the distance matrix to obtain the matching time of the target road segment in the neighboring road segments at each time moment. The time difference between this time moment and the matching time moment is recorded as the time delay. For all historical days, the mode of the time delay between this time moment and the matching time moment is used as the time delay between the target road segment and the neighboring road segments at that time moment.
[0053] Furthermore, considering that time delay is not only related to time but may also be related to the magnitude of traffic flow, for the target road segment, a road segment vector Lt(V, Ts) is constructed by matching the target road segment with neighboring road segments at each time step; V represents the traffic flow of neighboring road segments, and Ts represents the time delay between the target road segment and neighboring road segments at each time step. Over a historical period of T days, the target road segment generates T road segment vectors with neighboring road segments at the same time step.
[0054] By clustering the number of days corresponding to the target road segment at the same time using the distance between the traffic flow of neighboring road segments at each matched time, the time delay of matching road segment traffic flow caused by different traffic flow can be obtained. Then, based on the real-time monitoring of traffic flow of the target road segment and neighboring road segments, the corresponding time delay can be obtained, allowing for correction of the target road segment prediction results based on the corresponding time delay. In this embodiment, the DBSCAN clustering algorithm with minpts=3 and r=3 is used to cluster the road segment vectors of the target road segment at the same time. The clustering distance is the distance between the traffic flow of neighboring road segments, resulting in K clusters, meaning that the historical T days are divided into K clusters. For example, for the target road segment at 6 o'clock, it matches the traffic flow of neighboring road segments at multiple times within the corresponding T days, and clustering is performed based on the distance between these traffic flow values.
[0055] To improve the reliability of the time delay correction results for the next time step of the target road segment based on this method, this application analyzes the differences in traffic flow of the target road segment within the same cluster, as well as the differences in traffic flow of the target road segment between different clusters, thereby obtaining the reliability of the time delay correction. Specifically, it analyzes the inter-cluster and intra-cluster differences among clusters.
[0056] Specifically, for each cluster at each time point, there are several days. The average traffic flow of the target road segment on these days at that time point is taken as the average traffic flow of that cluster. The average absolute value of the difference between the average traffic flow of each cluster and the average traffic flow of the other clusters at each time point is taken as the inter-cluster difference of each cluster.
[0057] For each cluster, calculate the absolute value of the difference between the daily traffic flow of the target road segment and the daily traffic flow of the other segments. Sum all the absolute values of these differences as the intra-cluster variance. Calculate the time delay reliability based on the inter-cluster and intra-cluster variances.
[0058] The reliability of the time delay is positively correlated with inter-class differences and negatively correlated with intra-class differences.
[0059] It should be noted that positive correlation means that when one variable increases, the other variable also increases, and the two variables change in the same direction. When one variable changes from large to small or from small to large, the other variable also changes from large to small or from small to large. The specific relationship is determined by the actual application, and this application does not impose any special restrictions.
[0060] Preferably, in this embodiment, the expression for time delay reliability is:
[0061] , This represents the inter-class differences of the h-th cluster. This represents the difference in traffic flow between the e-th element and the other elements in the h-th cluster, corresponding to the target road segment. This represents the number of elements in the h-th cluster, i.e., the number of days. This represents the intra-cluster variance of the h-th cluster. Indicates the number of clusters. This represents the reliability of the time delay in correcting the prediction results of road segment A based on the traffic flow of road segment B. This value is defined as 1 when the intra-class difference is 0.
[0062] The greater the difference between each cluster and the smaller the difference between elements within a cluster, the better it is at inferring the traffic flow of neighboring road segments at each moment of the target road segment based on the time delay. This information is then used to make prediction corrections.
[0063] Thus, the time delay reliability of the target road segment and its neighboring road segments at each moment was obtained.
[0064] Step S004: Determine the traffic flow relationship between the current road segment and adjacent road segments by analyzing the historical traffic flow relationship between the road segment and adjacent road segments; then calculate the corrected traffic flow by combining the influence relationship of the road segments with the traffic flow before correction.
[0065] Real-time traffic flow data for the target road segment is collected, and the collection time is obtained. The traffic flow data at this time is compared with the traffic flow data of each cluster of the target road segment at the same time. The absolute value of the difference between the real-time traffic flow data of the target road segment and the average traffic flow data of all elements in the cluster is calculated to obtain the real-time difference. The cluster with the smallest real-time difference is selected as the reference cluster. The average time delay of all elements in the reference cluster is used as the current time delay of the target road segment and its neighboring road segments.
[0066] Based on the current time delay, the traffic flow of the target road segment at the same time in the reference cluster is found, and the difference between the traffic flow of the target road segment at the same time in the past and the traffic flow of the adjacent road segment at the time corresponding to the current time delay is calculated and recorded as the traffic flow change value.
[0067] Based on the current time delay, the corresponding time of the neighboring road segment is found for the target road segment at the current time. That is, the time obtained by subtracting the current time from the current time delay is taken as the corresponding time of the neighboring road segment. The sum of the traffic flow and traffic flow change value of the neighboring road segment at the next time is used as the prediction correction reference value of the neighboring road segment for the target road segment at the next time.
[0068] In this application, when the current time delay is positive (i.e., the time difference is positive), it indicates that the traffic flow data of the target road segment can be predicted based on the traffic flow data of neighboring road segments; conversely, when the current time delay is negative (i.e., the time difference is negative), it indicates that the traffic flow data of neighboring road segments can be predicted based on the traffic flow data of the target road segment. Therefore, when making prediction corrections for the target road segment, it is necessary to select the road segments with positive time delays at that moment from among the n road segments for analysis.
[0069] Therefore, the current time delay between the target road segment and all neighboring road segments is calculated at the current moment. Neighboring road segments with positive current time delays are selected for prediction correction, and these road segments are recorded as affected road segments.
[0070] The corrected traffic flow is obtained by weighting the traffic flow before correction and the predicted correction reference values of all affected road segments, using the correlation between the target road segment and the affected road segments as the weight.
[0071] Specifically, the target weight is obtained by averaging the correlation between the target road segment and all affected road segments. The weight of each affected road segment is the ratio of the time delay reliability of the target road segment and each affected road segment to the sum of the time delay reliability of the target road segment and all affected road segments. The correction parameter is obtained by weighted summing of the weight and the corresponding prediction correction reference value of the affected road segment.
[0072] The larger the target weight, the smaller the impact of the traffic flow before correction; the larger the target weight, the larger the correction parameter.
[0073] Preferably, in this embodiment, the expression for the corrected traffic flow is:
[0074] , Indicates the target weight. This represents the corrected traffic flow at time r+1. This represents the reliability of the time delay in correcting the prediction result of road segment A based on the traffic flow of road segment C at time r. Indicates the number of road sections affected. This represents the prediction correction reference value of road segment C for road segment A at time r+1. This represents the corrected traffic flow for road segment A at time r+1. Road segment A is the target road segment, and road segment C is the affected road segment. If there are no affected road segments, the traffic flow remains unchanged.
[0075] At this point, the corrected traffic flow for the target road segment at the current time and the next time is obtained.
[0076] Step S005: Real-time monitoring of traffic conditions by adjusting traffic flow.
[0077] Using the above method, the traffic flow of the target road segment at each moment is obtained, and the traffic status of each highway reconstruction and expansion construction area is monitored in real time.
[0078] In highway reconstruction and expansion construction areas, the system directly connects the corrected traffic flow data generated in the above steps to the cloud-based traffic management platform. This data, combined with road capacity, yields a congestion index, which classifies road segments according to congestion levels based on international standards.
[0079] The system generates traffic diversion plans or control instructions and disseminates them to drivers through roadside variable message signs (information boards), intelligent broadcasting, and other equipment. For example, when the system predicts severe congestion upstream of the construction zone at time t+1, it will display "Construction ahead, it is recommended to slow down" or suggest detour routes on the information boards in the direction of oncoming traffic in advance, realizing closed-loop active management from perception, decision-making to dissemination.
[0080] This completes the real-time traffic monitoring.
[0081] It should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.
[0082] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.
Claims
1. A method for real-time traffic monitoring in highway reconstruction and expansion construction areas, characterized in that, The method includes the following steps: The collected vehicle traffic flow constitutes a traffic flow time series; The traffic flow time series of the target road segment is predicted by the prediction algorithm to obtain the corrected traffic flow for the next time step from the current time; the correlation between the target road segment and its neighboring road segments is determined based on the traffic flow differences at the same time on multiple historical days. Based on historical traffic flow and time, the target road segment and each neighboring road segment are matched to obtain the time delay at each moment; for the target road segment at the same time on different days, the traffic flow of neighboring road segments is clustered; the reliability of the time delay is determined by the inter-class and intra-class differences of the clusters. The current time delay of the target road segment and its neighboring road segments is filtered by the real-time traffic flow difference between the target road segment and the cluster. The predicted correction reference value is determined based on historical traffic flow changes and the current time delay. The affected road segments are filtered based on the positive and negative relationship of the current time delay. The corrected traffic flow is obtained by weighting the traffic flow before correction and the predicted correction reference value of all affected road segments with the correlation between the target road segment and the affected road segments as the weight and the reliability of the time delay as the second weight. Traffic conditions are monitored in real time by adjusting traffic flow.
2. The method for real-time traffic status monitoring in highway reconstruction and expansion construction areas as described in claim 1, characterized in that, The correlation between the target road segment and adjacent road segments is negatively correlated with the differences in all traffic flows.
3. The method for real-time traffic status monitoring in highway reconstruction and expansion construction areas as described in claim 1, characterized in that, The method for obtaining the time delay at each moment by matching the target road segment and each adjacent road segment based on historical traffic flow and time is as follows: When matching the traffic flow of the target road segment with each neighboring road segment at all times, the Euclidean distance between the traffic flow and the normalized result at each time is used as the matching distance. The target road segment and the neighboring road segments are constructed into corresponding distance matrices. The Hungarian algorithm is used to match the distance matrices to obtain the matching time of the target road segment in the neighboring road segments at each time. The time difference between the target road segment and the matching time is recorded as the time delay. The mode of the time delays of all days is taken as the time delay of the road segment at each time.
4. The method for real-time traffic status monitoring in highway reconstruction and expansion construction areas as described in claim 3, characterized in that, The method for clustering traffic flow of a target road segment at the same time on different days based on neighboring road segments is as follows: For the target road segment and each neighboring road segment, the distance between the traffic flow of the neighboring road segments at different times corresponding to different days at each time of the target road segment is used as the clustering distance, and the historical days are clustered to obtain the clusters.
5. The method for real-time traffic status monitoring in highway reconstruction and expansion construction areas as described in claim 1, characterized in that, The method for obtaining the inter-class differences is as follows: For each cluster corresponding to the target road segment at each time point, the average traffic flow of the target road segment on different days within the cluster is taken as the average traffic flow of the cluster; the average absolute value of the difference between the average traffic flow of each cluster and the average traffic flow of the other clusters at each time point is taken as the inter-cluster difference of each cluster.
6. The method for real-time traffic status monitoring in highway reconstruction and expansion construction areas as described in claim 1, characterized in that, The method for obtaining the intra-class differences is as follows: For each cluster, calculate the absolute value of the difference between the daily traffic flow and the traffic flow on other days at the same time for the target road segment, and sum all the absolute values of the differences as the intra-cluster difference.
7. The method for real-time traffic status monitoring in highway reconstruction and expansion construction areas as described in claim 1, characterized in that, The reliability of the time delay is positively correlated with inter-class differences and negatively correlated with intra-class differences.
8. The method for real-time traffic status monitoring in highway reconstruction and expansion construction areas as described in claim 1, characterized in that, The method for filtering the current time delay of the target road segment and neighboring road segments based on the real-time traffic flow difference between the target road segment and the cluster is as follows: The absolute value of the difference between the real-time traffic flow of the target road segment and the average traffic flow of all target road segments in the cluster at the same time is recorded as the real-time difference. The cluster with the smallest real-time difference is used as the reference cluster. The average time delay of all elements in the reference cluster is used as the current time delay of the target road segment and neighboring road segments at the current moment.
9. The method for real-time traffic status monitoring in highway reconstruction and expansion construction areas as described in claim 1, characterized in that, The method for determining the prediction correction reference value based on historical traffic flow changes and current time delay is as follows: Based on the current time delay, find the traffic flow of the neighboring road segments corresponding to the target road segment at the same time in the reference cluster, and calculate the difference between the traffic flow of the target road segment at the same time in history and the neighboring road segments corresponding to the current time delay, which is recorded as the traffic flow change value. The time obtained by subtracting the current time from the current time delay is taken as the corresponding time of the neighboring road segment; the sum of the traffic flow and traffic flow change value of the neighboring road segment at the next time of the corresponding time is taken as the prediction correction reference value of the neighboring road segment for the target road segment at the next time of the current time.
10. The method for real-time traffic status monitoring in highway reconstruction and expansion construction areas as described in claim 1, characterized in that, The larger the target weight, the smaller the impact of the traffic flow before correction; the larger the target weight, the larger the correction parameter. The target weight is obtained by averaging the correlation between the target road segment and all affected road segments. The ratio of the time delay reliability of the target road segment and each affected road segment to the sum of the time delay reliability of the target road segment and all affected road segments is used as the weight of each affected road segment. The correction parameter is obtained by weighted summing the weight and the predicted correction reference value obtained for the affected road segment.