Traffic jam analysis method based on intelligent traffic platform

Abstract

The invention belongs to the technical field of intelligent traffic, and particularly discloses a traffic jam analysis method based on an intelligent traffic platform, which comprises the following steps: collecting multi-source traffic flow data of a target road section in real time, judging whether the target road section is jammed in combination with historical data, verifying the traffic flow in a continuous monitoring period to improve the accuracy, and if the target road section is jammed, judging whether the target road section is jammed or not. If yes, the congestion influence range is dynamically determined by identifying a core area and analyzing the upstream and downstream speed propagation trend, then frequent or accidental congestion types are accurately distinguished and the congestion level is evaluated by calculating the deviation degree of current data and a historical traffic mode in multiple dimensions, and finally the congestion influence range is determined according to the congestion types, the congestion level and the increase trend. A differentiated traffic dispersion scheme is generated and executed; according to the invention, through deep fusion of type identification, degree evaluation, range determination and scheme generation links, an integrated decision link is formed, and the intelligent level and response efficiency of the system for coping with a complex congestion scene are significantly improved.

Description

Technical Field

[0001] This invention belongs to the field of intelligent transportation technology and relates to a traffic congestion analysis method based on an intelligent transportation platform. Background Technology

[0002] With the continued acceleration of urbanization and the rapid growth of motor vehicle ownership, the traffic pressure on urban road networks is intensifying, and traffic congestion has become a key issue significantly restricting urban operational efficiency and reducing the quality of public travel. Faced with this challenge, effective congestion analysis and management technologies are particularly important. Early traffic congestion analysis methods mostly relied on threshold judgments based on single-dimensional traffic parameters. Due to the limitations of data dimensions, such methods struggle to comprehensively and accurately depict the actual operating status of road segments, thus failing to meet the precision and intelligence requirements of modern intelligent traffic management.

[0003] For example, Chinese invention patent CN119252035A discloses an intelligent traffic information processing platform. This platform typically includes modules for data collection, preprocessing, pattern recognition, traffic flow prediction, and information dissemination. By integrating real-time and historical traffic data for trend prediction, it achieves early warning and macro-level guidance for congestion to a certain extent. This type of technical solution represents the general development level in this field.

[0004] The existing technologies mentioned above have the following shortcomings: 1. Existing technologies only build predictive models by using real-time and historical data, but their analysis is mostly focused on traffic flow prediction and macro-pattern recognition. They lack a fine distinction and judgment of the congestion types to which frequent congestion and occasional congestion belong. As a result, they cannot effectively identify frequent congestion caused by regular daily commuting needs and occasional congestion caused by random events such as traffic accidents and temporary construction. Therefore, they cannot guarantee the pertinence of the generated traffic management plan and cannot fundamentally and efficiently alleviate congestion of different causes.

[0005] 2. The response mechanism of existing technology platforms focuses on advance guidance based on prediction results, lacking real-time assessment of the degree of congestion after it occurs, as well as quantification and response strategies for dynamic changes in the congestion range. Their diversion plans are static or predefined and cannot be dynamically adjusted and optimized according to the aggravation or relief of congestion or the expansion or contraction of the impact range. As a result, the allocation of diversion resources is not accurate enough and the response efficiency needs to be improved. Summary of the Invention

[0006] In view of this, in order to solve the problems mentioned in the background technology, a traffic congestion analysis method based on an intelligent transportation platform is proposed.

[0007] The objective of this invention can be achieved through the following technical solution: This invention provides a traffic congestion analysis method based on an intelligent transportation platform, including: S1, real-time collection of multi-source traffic flow data of the target road segment in each monitoring period, wherein the multi-source traffic flow data includes average traffic speed, overall lane occupancy rate, total traffic volume and real-time video.

[0008] S2. Based on the multi-source traffic flow data of the target road segment during each monitoring period, combined with the historical traffic data of the target road segment, determine whether the target road segment is in a congested state.

[0009] S3. If the target road segment is congested, the congestion impact range of the target road segment is determined based on the multi-source traffic flow data.

[0010] S4. Based on the multi-source traffic flow data and historical traffic data within the congestion impact range, calculate the deviation of the target road segment in each dimension, analyze the congestion type based on the deviation of each dimension, and perform weighted fusion of the deviation of each dimension to obtain a comprehensive congestion index, thereby determining the congestion level.

[0011] S5. Based on the congestion type and congestion level, generate a traffic diversion plan corresponding to the congestion type, and send the diversion plan to the intelligent transportation platform for execution.

[0012] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) By constructing historical traffic patterns and calculating the deviation of real-time data in multiple dimensions to determine the threshold, the present invention can accurately distinguish between regular and occasional congestion and sudden and occasional congestion, overcoming the current deficiency of unclear identification of the internal mechanism of congestion, and providing a reliable basis for implementing cause-oriented differentiated diversion strategies.

[0013] (2) This invention generates a comprehensive congestion index by weighting and integrating the deviation of each dimension, and accurately matches it with the preset level range. This achieves an objective and quantitative classification of the severity of congestion, overcomes the drawbacks of the crude response of traditional methods, and lays a technical foundation for precise resource allocation and scheme control based on congestion intensity.

[0014] (3) By identifying the core area, analyzing the spread trend, and determining the start and end points, this invention can dynamically and accurately depict the scope of congestion and its evolution, overcoming the limitations of static scope determination and providing key spatial decision support for implementing forward-looking spatial diversion measures.

[0015] (4) By deeply integrating type identification, degree assessment, scope determination and scheme generation, this invention forms an integrated decision-making link, which significantly improves the intelligence level and response efficiency of the system in dealing with complex congestion scenarios, avoids the situation of relatively fragmented links and reliance on manual intervention, and improves the efficiency and intelligence level of traffic congestion management.

[0016] (5) By constructing a unified deviation calculation framework, this invention realizes data sharing and logical coordination between congestion type identification and degree assessment, overcoming the drawbacks of system processing redundancy and inconsistent decision results caused by the independent calculation logic of congestion type analysis and degree assessment in traditional methods, and significantly improving system processing efficiency and decision consistency. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram showing the connections between the steps of the method of the present invention.

[0019] Figure 2 This is a schematic diagram showing the connection steps of the congestion verification and analysis process in this invention.

[0020] Figure 3 This is a schematic diagram showing the steps for determining the congestion impact range of the target road segment in this invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Please see Figure 1 As shown, the present invention provides a traffic congestion analysis method based on an intelligent transportation platform. The method includes: S1, real-time collection of multi-source traffic flow data of the target road segment in each monitoring period, wherein the multi-source traffic flow data includes average traffic speed, overall lane occupancy rate, total traffic volume and real-time video.

[0023] S2. Based on the multi-source traffic flow data of the target road segment during each monitoring period, combined with the historical traffic data of the target road segment, determine whether the target road segment is in a congested state.

[0024] For example, determining whether the target road segment is in a congested state includes: obtaining the average speed of the target road segment under the same time period and the same week type of smooth traffic from historical traffic data and the overall average lane occupancy rate, and using them as the reference speed for smooth traffic and the critical occupancy rate for congestion, respectively.

[0025] Specifically, by extracting the average speed, congestion threshold, and baseline total traffic volume of the target road segment under the same time period and week type from historical traffic data, the judgment benchmark is deeply linked to historical traffic patterns. This approach enables customized judgment based on a specific benchmark for each scenario.

[0026] The average traffic speed and overall lane occupancy rate were compared with the corresponding smooth traffic reference speed and congestion threshold occupancy rate for each road segment.

[0027] If the average traffic speed is lower than the reference speed for smooth traffic on the road segment, and the overall lane occupancy rate is higher than the critical occupancy rate for congestion, then the traffic efficiency of the target road segment is preliminarily determined to be abnormal.

[0028] For target road segments initially identified as having abnormal traffic efficiency, a congestion verification analysis is conducted based on their total traffic volume to obtain verification results on whether the target road segment is in a congested state.

[0029] It should be added that, by first using the combination of average traffic speed being lower than the historical smooth traffic benchmark and overall lane occupancy rate being higher than the historical congestion threshold benchmark, potential congested road segments are initially screened out. This avoids invalid verification of road segments with normal speed and occupancy rate, thus improving the efficiency of judgment. At the same time, compared with single data screening, the combination of speed and occupancy rate can exclude non-congested scenarios such as low speed but normal occupancy rate and high occupancy rate but normal speed, reducing the workload of subsequent verification.

[0030] Please see Figure 2 As shown, further, the congestion verification analysis includes: obtaining the total traffic flow data sequence of the target road segment over K consecutive monitoring periods from multi-source traffic flow data, where K is the number of consecutive monitoring periods used to collect and analyze traffic flow data, and is a core parameter for quantifying traffic flow change trends. K is an integer greater than 1, such as K=3, K=5, K=7. Essentially, by using traffic flow data sequences from multiple consecutive monitoring periods, the interference of accidental fluctuations in single-period data on the judgment results is avoided, ensuring the statistical significance of conclusions such as congestion status judgment and congestion propagation trend analysis.

[0031] The total traffic flow of the target road segment under the same time period and week type is obtained from the historical traffic data, and then the average of the total traffic flow is calculated to obtain the baseline total traffic flow.

[0032] Calculate the reasonable fluctuation ratio of the total traffic flow of the target road segment under the same time period and the same week type. The product of this ratio and the baseline total traffic flow is used as the amplitude threshold. The reasonable fluctuation ratio is determined based on the ratio of the standard deviation of the total traffic flow of the target road segment under the same time period and the same week type in the past three months to the baseline total traffic flow. It is used to distinguish between occasional slight fluctuations and significant traffic flow exceeding the standard.

[0033] The total traffic flow in each period of the total traffic flow data sequence is compared with the baseline total traffic flow. If the total traffic flow is less than the baseline total traffic flow, it is marked as a potential congestion period.

[0034] The relative deviation between the total traffic flow during the potential congestion period and the baseline total traffic flow is calculated to obtain the congestion magnitude.

[0035] If the congestion magnitude of a potential congestion cycle is greater than or equal to the magnitude threshold, it is determined to be a valid congestion cycle, and the number of valid congestion cycles is counted.

[0036] If the number of effective congestion cycles is greater than half the number of monitoring cycles, the target road segment is determined to be congested; otherwise, it is determined not to be congested. Where K is an odd number, it is the rounded-up value of half the number of monitoring cycles. When the number of effective congestion cycles exceeds this threshold, it is determined to be congested. This is essentially based on a dual consideration of traffic flow continuity and statistical significance. From the perspective of traffic flow characteristics, the formation and dissipation of congestion is a continuous process rather than an instantaneous state. This standard requires congestion to cover at least half of the monitoring cycles, which can exclude interference from momentary anomalies such as temporary cutting in or pedestrians suddenly crossing the road, accurately matching the essence of congestion—a continuous decline in traffic efficiency—and improving the reliability of the judgment and the universality of the solution.

[0037] It should be added that a secondary verification of the initial judgment result is achieved by comparing the traffic flow data sequence of K consecutive monitoring periods with the baseline total traffic flow and the magnitude threshold. If the total traffic flow is lower than the baseline but the deviation does not reach the magnitude threshold, it is judged as temporary congestion, and no diversion plan needs to be initiated to avoid wasting resources. If the total traffic flow is lower than the baseline and the deviation is significant, and the number of effective congestion periods exceeds half of the monitoring period, it is judged as real congestion, triggering the subsequent process of determining the congestion range and generating a diversion plan. This verification logic improves the accuracy of distinguishing between real congestion and false congestion, and reduces the problem of ineffective or missed diversion due to misjudgment.

[0038] S3. If the target road segment is congested, the congestion impact range of the target road segment is determined based on the multi-source traffic flow data.

[0039] It's important to add that accurately determining the extent of congestion is a prerequisite for effective traffic management. If the defined area is too small, the management measures may not cover the entire congested area, resulting in a superficial solution. If the defined area is too large, it may lead to unnecessary traffic control, wasting resources and even interfering with normal traffic flow. Therefore, scientifically and dynamically defining the spatial boundaries of congestion is crucial for developing precise and efficient traffic management plans.

[0040] Please see Figure 3 As shown, for example, determining the congestion impact range of the target road segment includes: determining the core congestion area based on the average traffic speed and overall lane occupancy rate of the target road segment in the multi-source traffic flow data.

[0041] It's important to note that in the overall process of determining the congestion impact range of a target road segment, identifying the core congestion area pinpoints the source of congestion, providing a baseline coordinate for analyzing congestion propagation trends. The slow-moving traffic in the core congestion area will gradually spread upstream or downstream. If the core congestion area is not identified, it's impossible to determine which upstream or downstream location to use as a baseline for collecting speed data and analyzing propagation trends, leading to misjudgments of the propagation direction and subsequent deviations in the delineation of congestion boundaries.

[0042] Furthermore, determining the core congestion area includes dividing the target road segment into several continuous equidistant sub-regions at preset fixed intervals.

[0043] It should be added that dividing the sub-regions into equidistant sub-regions according to a preset fixed interval ensures that the spatial range of each analysis unit is consistent, avoiding misjudgment of the core area due to uneven distribution of detection sections of the target road segment. At the same time, the equidistant sub-regions can integrate all detection data within the range to ensure that the average traffic speed and overall lane occupancy rate data of each sub-region are representative, laying an accurate data foundation for the location of the core area.

[0044] The average traffic speed and overall lane occupancy rate of each equidistant sub-region in the latest monitoring period are compared with their preset thresholds.

[0045] Adjacent equidistant sub-regions with average traffic speeds lower than the smooth traffic reference speed and overall lane occupancy rates higher than the congestion threshold are clustered and merged to form candidate congestion areas.

[0046] The candidate congestion area with the lowest average traffic speed is selected from all candidate congestion areas and designated as the core congestion area.

[0047] Based on the core congestion area, the average traffic speed data sequence of the upstream and downstream adjacent road segments is collected over K consecutive monitoring periods.

[0048] Linear regression analysis was performed on the average traffic speed data sequence to obtain the speed change slope of each adjacent road segment.

[0049] It should be added that by performing linear regression on the average traffic speed data sequence over K consecutive monitoring periods, the slope of speed change is obtained. The larger the absolute value of the negative slope, the more significant the speed decrease and the stronger the congestion propagation trend. The slope transforms the speed change trend into a specific value. Combined with a preset speed change slope threshold, the existence and intensity of propagation can be objectively determined, avoiding the subjective bias of manual observation.

[0050] If the slope of the speed change of the upstream adjacent road segment is less than the preset speed change slope threshold, it is determined that the congestion has a tendency to spread upstream. If the slope of the speed change of the downstream adjacent road segment is less than the preset speed change slope threshold, it is determined that the congestion has a tendency to spread downstream.

[0051] It should be added that the preset speed change slope threshold can be obtained through historical data statistical analysis. Specifically, historical traffic flow data that meets the current monitoring scenario within the past 6 months is first collected from the target road segment and its upstream and downstream adjacent road segments and pre-processed by cleaning. Then, real congestion propagation samples are extracted, and the speed change slope of the upstream and downstream road segments in each sample is calculated. The slope samples are then statistically analyzed, and the threshold is obtained by combining the traffic flow stability requirements. At the same time, a dynamic update mechanism can be established to update the threshold periodically or in a timely manner according to the changes in the road segment. In this way, the preset speed change slope threshold is accurately quantified, and its numerical attribute is usually negative.

[0052] In directions where there is a tendency for congestion to spread, the point where the average traffic speed first drops to the reference speed for smooth traffic flow is marked as the starting point of congestion.

[0053] The point where the average traffic speed first recovers to the reference speed for smooth traffic flow is marked as the end of congestion.

[0054] The contiguous road segment covered by the congestion start point, the core congestion area, and the congestion end point is defined as the congestion impact range of the target road segment.

[0055] S4. Based on the multi-source traffic flow data and historical traffic data within the congestion impact range, calculate the deviation of the target road segment in each dimension, analyze the congestion type based on the deviation of each dimension, and perform weighted fusion of the deviation of each dimension to obtain a comprehensive congestion index, thereby determining the congestion level.

[0056] For example, the calculation of the deviation of the target road segment in each dimension includes: calculating the ratio of the difference between the average traffic speed and the historical average traffic speed to obtain the average vehicle speed deviation.

[0057] The ratio of the difference between the overall lane occupancy rate and the historical overall lane occupancy rate is used as the deviation of the overall lane occupancy rate.

[0058] The absolute value of the relative deviation between the current real-time traffic flow and the historical traffic flow is calculated to obtain the traffic flow deviation. Then, the deviation of each dimension is obtained. Among them, the historical average traffic speed, historical overall lane occupancy rate and historical traffic flow are the statistical average values ​​of the average traffic speed, historical overall lane occupancy rate and historical traffic flow obtained by multiple monitoring in the same period and the same week type of frequent congestion scenario of the target road segment.

[0059] For example, the analysis of congestion types includes: if the average vehicle speed deviation is lower than its vehicle speed deviation threshold, the overall lane occupancy deviation is lower than its occupancy deviation threshold, and the traffic flow deviation is lower than its traffic flow deviation threshold, then the congestion type is initially determined to be frequent congestion; otherwise, the congestion type is initially determined to be occasional congestion.

[0060] It's important to note that for cases initially identified as recurring congestion, the determination is based on a high degree of consistency between the current traffic flow and historical patterns. Recurring congestion is caused by regular and predictable factors, and its traffic flow parameters have formed stable and reproducible patterns in historical data. Therefore, when deviations across multiple dimensions indicate that the current state does not exceed the historical normal fluctuation range, there is a high degree of confidence in confirming it as recurring congestion, without needing to initiate additional confirmation procedures. This approach aligns with the causal patterns of recurring congestion while simultaneously improving system processing efficiency.

[0061] Among them, the speed deviation threshold is a critical value for determining whether the current average speed deviation meets the characteristics of frequent congestion; essentially, it is the statistical upper limit of the average speed deviation of the target road segment during frequent congestion in the same historical period and week type. The lane occupancy deviation threshold is a critical value for determining whether the current overall lane occupancy deviation meets the characteristics of frequent congestion; it corresponds to the statistical upper limit of the overall lane occupancy deviation in historical frequent congestion scenarios for the target road segment. The traffic flow deviation threshold is a critical value for determining whether the current traffic flow deviation meets the characteristics of frequent congestion; that is, the statistical upper limit of the traffic flow deviation in historical frequent congestion scenarios for the target road segment.

[0062] For events initially determined to be sporadic congestion, the system detects whether there are traffic accidents, road construction, or temporary traffic control events within the congestion area based on the real-time video. If such events are detected, the congestion is determined to be sporadic; otherwise, enhanced analysis is triggered for a final determination.

[0063] It should be added that for cases initially judged as sporadic congestion, the basis for the judgment is that the current state significantly deviates from historical patterns, but the specific reasons for this deviation are varied. Besides traffic accidents, construction, and other targeted events, it could also stem from large-scale events, extreme weather, or instantaneous noise during data collection. Therefore, the preliminary judgment only indicates an anomaly, but cannot specifically pinpoint the sporadic event that this invention aims to precisely address. The purpose of introducing the cause analysis process is to confirm the cause of the preliminary judgment. By proactively detecting traffic accidents, road construction, or temporary traffic control events using sensor data such as video, other interfering factors can be eliminated, ensuring that the final judgment of sporadic congestion is caused by a clear event requiring a specific traffic management strategy. This greatly improves the accuracy of the type determination and the targeted nature of subsequent solution generation.

[0064] The enhanced analysis includes: re-verifying whether there are any anomalies in the multi-source traffic flow data used to initially determine occasional congestion, such as whether the data acquisition equipment is malfunctioning and causing data distortion. If there are any data anomalies, the data is corrected and the initial determination of the congestion type is performed again.

[0065] If the multi-source traffic flow data shows no abnormalities, then the traffic operation status of the congested area is analyzed again by combining other sensing data from the intelligent transportation platform, such as millimeter-wave radar data from traffic incident detection equipment and vehicle flow time-series data from traffic flow monitoring equipment.

[0066] If, after integrating various types of sensor data, no obvious special event trigger is found, and a deep comparison with traffic flow data of the same time period and week type reveals that the current traffic flow characteristics of congestion differ significantly from those of previous sporadic congestion, and also do not conform to historical traffic patterns of frequent congestion, then it is determined to be a new type of sporadic congestion. This new type of sporadic congestion may be caused by factors that have not yet been recorded or identified by the system, such as sudden large-scale gatherings of people or clusters of vehicle breakdowns.

[0067] Based on the preliminary judgment results and the results of event detection and enhancement analysis for sporadic congestion, the congestion type is determined.

[0068] For example, determining the congestion level includes: performing a weighted fusion calculation on the average vehicle speed deviation, the overall lane occupancy deviation, and the traffic flow deviation to obtain a comprehensive congestion index.

[0069] It should be added that the formula for calculating the comprehensive congestion index is as follows: In the formula To calculate the overall congestion index, , and These are the weights for average vehicle speed deviation, overall lane occupancy deviation, and traffic flow deviation, used to quantify the impact of deviations in different dimensions on the overall congestion level. , , . , and These are the deviations from average vehicle speed, overall lane occupancy rate, and traffic flow.

[0070] By using weighted fusion to calculate the comprehensive congestion index, on the one hand, the weight allocation can reflect the actual weight of the impact of average vehicle speed deviation, overall lane occupancy deviation, and traffic flow deviation on different dimensions of comprehensive congestion, reflecting the different contributions of deviations in different dimensions to comprehensive congestion. On the other hand, it can directly integrate the information of deviations in the three dimensions of average vehicle speed, overall lane occupancy, and traffic flow, and comprehensively consider the impact of the three on comprehensive congestion.

[0071] The weights can be set based on traffic management needs and actual traffic operation experience, or they can be obtained through experimental data. For example, historical data on average vehicle speed deviation, overall lane occupancy deviation, traffic flow deviation, and corresponding overall congestion perception under different congestion scenarios can be collected first. The correlation coefficients between average vehicle speed deviation, overall lane occupancy deviation, traffic flow deviation, and overall congestion level can be calculated. The contribution of the three factors to the overall congestion level can be determined through regression analysis. After normalization, the contribution is converted into the weights of average vehicle speed deviation, overall lane occupancy deviation, and traffic flow deviation, and the sum of the weights is 1. In this way, the overall congestion index can be accurately quantified.

[0072] The overall congestion index is matched with the overall congestion index range corresponding to each congestion level to obtain the congestion level.

[0073] It should be added that the comprehensive congestion index range corresponding to each congestion level is a numerical range used to divide different levels of congestion severity. Different ranges correspond to different levels of congestion, from smooth traffic to severe congestion, reflecting the correspondence between the comprehensive congestion index and the actual severity of congestion.

[0074] The comprehensive congestion index range corresponding to each congestion level can be obtained through historical data statistical analysis. Specifically, historical traffic flow data that meets the current monitoring scenario is first collected from the target road segment and surrounding related road segments within the past 6 months. Then, typical samples of different congestion levels are extracted, and the comprehensive congestion index corresponding to each sample is calculated. Statistical analysis is then performed on the comprehensive congestion index samples under each congestion level to determine the index distribution range corresponding to each congestion level. At the same time, a dynamic update mechanism can be established to update the index range regularly or in a timely manner based on changes in road segment traffic conditions, thereby accurately quantifying the comprehensive congestion index range corresponding to each congestion level.

[0075] S5. Based on the congestion type and congestion level, generate a traffic diversion plan corresponding to the congestion type, and send the diversion plan to the intelligent transportation platform for execution.

[0076] For example, generating a traffic management plan corresponding to frequent congestion includes: if the congestion type is frequent congestion, obtaining the historical comprehensive congestion index of the target road segment in the same time period and the same week type from historical traffic data.

[0077] Using date as the horizontal axis and the comprehensive congestion index of the same date within the same time period as the vertical axis, a multi-period congestion trend baseline curve is constructed, and the slope is extracted as the congestion growth rate of the target road segment.

[0078] If the congestion level is high, an emergency congestion mitigation plan will be triggered directly.

[0079] If the congestion level is medium or low, and the congestion growth rate exceeds the preset congestion growth rate threshold, an emergency congestion mitigation plan will be triggered.

[0080] It should be added that the preset congestion growth threshold is a critical indicator used to quantitatively determine whether the growth rate of recurring congestion exceeds the normal range. Essentially, it is a specific value of the slope of the multi-cycle congestion trend baseline curve. This threshold reflects the critical rate of change when recurring congestion transitions from gradual growth to rapid deterioration. When the congestion growth rate of the target road segment exceeds this threshold, it indicates that the congestion level is worsening at a rate exceeding historical norms, requiring the activation of an emergency traffic management plan. If it is less than or equal to this threshold, the congestion growth conforms to normal patterns, and a conventional traffic management plan can be matched according to the congestion level.

[0081] The preset congestion growth threshold is obtained based on statistical analysis of historical frequent congestion data for the target road segment, ensuring deep compatibility with the road segment's congestion growth pattern and the calculation logic of the comprehensive congestion index. The specific steps are as follows: First, historical data screening and preprocessing are performed to define the data scope: Historical data on frequent congestion within the past 12 months is collected for the target road segment. The screening criteria must strictly match the current frequent congestion scenario. The screened historical data are grouped by continuous period, with date as the horizontal axis and the comprehensive congestion index as the vertical axis. A historical congestion trend curve is constructed for each group. The historical congestion growth rate is extracted. Linear regression using the least squares method is performed on each historical trend curve to calculate the curve slope. This slope represents the congestion growth rate of a particular historical frequent congestion event.

[0082] Statistical analysis is performed on all historical congestion growth samples, and their 90th percentile is calculated and used as the preset congestion growth threshold. Using the 90th percentile can filter out the critical value that exceeds 90% of the normal growth rate, which not only avoids the interference of extremely rapid growth samples on the threshold, but also accurately covers rapid growth scenarios that require emergency intervention.

[0083] If the congestion level is medium or low, and the congestion growth rate is less than or equal to the preset congestion growth rate threshold, then the regular congestion mitigation plan will be triggered.

[0084] For example, the generation of traffic management solutions corresponding to occasional congestion includes: if the real-time video identifies that there are stationary vehicles, crowds, or scattered vehicle parts, then the triggering event type is determined to be a traffic accident.

[0085] If construction barriers and engineering vehicles are identified from real-time video data, the triggering event type is determined to be road construction.

[0086] If a temporary traffic restriction or abnormal traffic light status is identified from the real-time video, the triggering event type is determined to be temporary control.

[0087] If no matching event type is found, the event type is determined to be an unknown cause.

[0088] Based on the matching of the trigger event type and congestion level with the traffic management schemes corresponding to each trigger type and congestion level in historical traffic data, a traffic management scheme for occasional congestion is obtained.

[0089] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0090] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0091] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0092] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0093] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0094] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A traffic congestion analysis method based on an intelligent transportation platform, characterized by: The method includes: S1. Real-time collection of multi-source traffic flow data of the target road segment during each monitoring period, including average traffic speed, overall lane occupancy rate, total traffic volume, and real-time video. S2. Based on the multi-source traffic flow data of the target road segment during each monitoring period, combined with the historical traffic data of the target road segment, determine whether the target road segment is in a congested state. S3. If the target road segment is congested, the congestion impact range of the target road segment is determined based on the multi-source traffic flow data. S4. Based on the multi-source traffic flow data and historical traffic data within the congestion impact range, calculate the deviation of the target road segment in each dimension, analyze the congestion type based on the deviation of each dimension, and perform weighted fusion of the deviation of each dimension to obtain a comprehensive congestion index, thereby determining the congestion level. S5. Based on the congestion type and congestion level, generate a traffic diversion plan corresponding to the congestion type, and send the diversion plan to the intelligent transportation platform for execution.

2. The traffic congestion analysis method based on an intelligent transportation platform according to claim 1, characterized in that: The determination of whether the target road segment is congested includes: The average speed and overall lane occupancy rate of the target road segment under the same time period and week type are obtained from historical traffic data and used as the reference speed for smooth traffic and the critical occupancy rate for congestion, respectively. The average traffic speed and overall lane occupancy rate are compared with the corresponding smooth traffic reference speed and congestion critical occupancy rate for each road segment. If the average traffic speed is lower than the reference speed for smooth traffic on the road segment, and the overall lane occupancy rate is higher than the critical occupancy rate for congestion, then it is preliminarily determined that the traffic efficiency of the target road segment is abnormal. For target road segments initially identified as having abnormal traffic efficiency, a congestion verification analysis is conducted based on their total traffic volume to obtain verification results on whether the target road segment is in a congested state.

3. The traffic congestion analysis method based on an intelligent transportation platform according to claim 2, characterized in that: The congestion verification analysis includes: Obtain the total traffic flow data sequence of the target road segment over K consecutive monitoring periods from multi-source traffic flow data; The total traffic flow of the target road segment under the same time period and the same week type is obtained from the historical traffic data, and then the average of the total traffic flow is calculated to obtain the baseline total traffic flow. Calculate the reasonable fluctuation ratio of the total traffic flow of the target road segment under the same time period and the same week type, and use the product of this ratio and the baseline total traffic flow as the amplitude threshold. The total traffic flow in each monitoring period of the total traffic flow data sequence is compared with the baseline total traffic flow. If the total traffic flow is less than the baseline total traffic flow, it is marked as a potential congestion period. The relative deviation between the total traffic flow during the potential congestion period and the baseline total traffic flow is calculated to obtain the congestion magnitude; If the congestion magnitude of a potential congestion cycle is greater than or equal to the magnitude threshold, it is determined to be a valid congestion cycle, and the number of valid congestion cycles is counted. If the number of effective congestion cycles is greater than half of the number of monitoring cycles, the target road segment is determined to be in a congested state; otherwise, the target road segment is determined not to be in a congested state.

4. The traffic congestion analysis method based on an intelligent transportation platform according to claim 1, characterized in that: The congestion impact range of the target road segment includes: Based on the average speed and overall lane occupancy rate of the target road segment in the multi-source traffic flow data, the core congestion area is determined. Based on the core congestion area, the average traffic speed data sequence of the upstream and downstream adjacent road segments is collected over K consecutive monitoring periods; Linear regression analysis was performed on the average traffic speed data sequence to obtain the speed change slope of each adjacent road segment; If the slope of the speed change of the upstream adjacent road segment is less than the preset speed change slope threshold, it is determined that the congestion has a tendency to spread upstream. If the slope of the speed change of the downstream adjacent road segment is less than the preset speed change slope threshold, it is determined that the congestion has a tendency to spread downstream. In directions where there is a tendency for congestion to spread, the point where the average traffic speed first drops to the reference speed for smooth traffic flow is marked as the starting point of congestion; The point where the average traffic speed first returns to the reference speed for smooth traffic flow is marked as the end of congestion. The contiguous road segment covered by the congestion start point, the core congestion area, and the congestion end point is defined as the congestion impact range of the target road segment.

5. The traffic congestion analysis method based on an intelligent transportation platform according to claim 4, characterized in that: The identified core congestion areas include: The target road segment is divided into several continuous equidistant sub-regions at preset fixed intervals; The average traffic speed and overall lane occupancy rate of each equidistant sub-region in the latest monitoring period are compared with their preset thresholds. Adjacent equidistant sub-regions with average traffic speeds lower than the smooth traffic reference speed and overall lane occupancy rates higher than the congestion threshold occupancy rate are clustered and merged to form candidate congestion areas. The candidate congestion area with the lowest average traffic speed is selected from all candidate congestion areas and designated as the core congestion area.

6. The traffic congestion analysis method based on an intelligent transportation platform according to claim 1, characterized in that: The deviation of the target road segment in each dimension is calculated as follows: The average speed deviation is obtained by calculating the ratio of the difference between the average speed and the historical average speed. The ratio of the overall lane occupancy rate to the historical overall lane occupancy rate is used as the deviation of the overall lane occupancy rate. Calculate the absolute value of the relative deviation between the current real-time traffic flow and the historical traffic flow to obtain the traffic flow deviation, and then obtain the deviation of each dimension.

7. The traffic congestion analysis method based on an intelligent transportation platform according to claim 1, characterized in that: The types of congestion analyzed include: If the average vehicle speed deviation is lower than its speed deviation threshold, the overall lane occupancy deviation is lower than its occupancy deviation threshold, and the traffic flow deviation is lower than its traffic flow deviation threshold, then the congestion type is initially determined to be frequent congestion; otherwise, the congestion type is initially determined to be occasional congestion. For events initially determined to be occasional congestion, the system detects whether there are traffic accidents, road construction, or temporary traffic control events within the congestion area based on the real-time video. If such events are detected, the event is determined to be occasional congestion; otherwise, enhanced analysis is triggered for a final determination. Based on the preliminary judgment results and the results of event detection and enhanced analysis for sporadic congestion, the congestion type is determined.

8. The traffic congestion analysis method based on an intelligent transportation platform according to claim 1, characterized in that: The determination of the congestion level includes: The average vehicle speed deviation, overall lane occupancy deviation, and traffic flow deviation are weighted and fused to obtain the comprehensive congestion index; The overall congestion index is matched with the overall congestion index range corresponding to each congestion level to obtain the congestion level.

9. The traffic congestion analysis method based on an intelligent transportation platform according to claim 1, characterized in that: The generation of traffic management plans corresponding to frequent congestion includes: If the congestion type is frequent congestion, obtain the historical comprehensive congestion index of the target road segment for the same time period and the same week from historical traffic data; Using the date as the horizontal axis and the comprehensive congestion index of the same date within the same time period as the vertical axis, a multi-cycle congestion trend baseline curve is constructed, and the slope is extracted as the congestion growth rate of the target road segment. If the congestion level is high, the emergency congestion relief plan will be triggered directly. If the congestion level is medium or low, and the congestion growth rate exceeds the preset congestion growth rate threshold, an emergency congestion relief plan will be triggered. If the congestion level is medium or low, and the congestion growth rate is less than or equal to the preset congestion growth rate threshold, then the regular congestion mitigation plan will be triggered.

10. The traffic congestion analysis method based on an intelligent transportation platform according to claim 7, characterized in that: The traffic management solutions generated for occasional congestion include: If the real-time video identifies a stationary vehicle, a group of people, or scattered vehicle parts, the triggering event type is determined to be a traffic accident. If construction barriers and engineering vehicles are identified from real-time video data, the triggering event type is determined to be road construction. If a temporary traffic restriction or abnormal traffic light status is identified from the real-time video, the triggering event type is determined to be temporary control. If no matching event type is found, the event type is determined to be an unknown cause. Based on the matching of the trigger event type and congestion level with the traffic management schemes corresponding to each trigger type and congestion level in historical traffic data, a traffic management scheme for occasional congestion is obtained.

Images