Traffic special bottleneck monitoring method based on multi-source road data

Abstract

The invention provides a traffic special bottleneck monitoring method based on multi-source road data, and the method comprises the following steps: S1, collecting road data of a plurality of intelligent network connection data clients, and building a road information database of different road segments; s2, performing precision complementation and error ablation on the road information of the same road section of a plurality of intelligent network connection data clients, outputting multi-source road information, and calculating overall evaluation parameters of multi-dimensional deep road section ablation; s3, calculating a road data fuzzy matrix; s4, setting and calculating the critical entropy of the traffic event; calculating a fuzzy entropy in combination with the multi-source road information and the road data fuzzy matrix; and when the fuzzy entropy is greater than the critical entropy, constructing a special event list, and listing the special event list for continuous tracking. Through the multi-dimensional multi-source deep road data ablation algorithm, the deep road data fuzzy method and the road data fuzzy entropy fluctuation recognition algorithm, the accuracy of monitoring the special traffic events is improved from the way of processing the multi-source road data.

Description

Technical Field

[0001] The present invention relates to the technical field of traffic monitoring, and particularly to a traffic special event monitoring method based on multi-source road data Background Art

[0002] The traffic condition is complex and there are many emergencies, and various traffic special events often occur, such as traffic special events caused by sudden rainfall, traffic special events caused by sudden vehicle failures, and traffic special events caused by sudden accidents, etc. Traffic special events have characteristics such as suddenness and severity, which will cause serious delays and congestion in the road network, greatly affecting the road traffic function and the daily travel of residents

[0003] Traffic monitoring technology can monitor sudden traffic special events in the city, with the advantages of fast discovery time and high monitoring accuracy, and complete the monitoring of traffic special events in a low-manpower and high-efficiency manner, providing technical support for timely traffic management and control. However, most current traffic monitoring technologies identify and monitor based on single-source data. The small data source will cause the problem of low identification efficiency, while the traffic monitoring technology based on multi-source data has problems such as unreasonable data processing and low detection accuracy Summary of the Invention

[0004] In order to solve the problems existing in the above-mentioned prior art, the present invention provides a traffic special bottleneck monitoring method based on multi-source road data, including the following steps

[0005] S1. Multi-dimensional road data collection: Collect road data of several intelligent network-connected data user terminals, and establish a road information database segmented by road sections

[0006] S2. Deep fusion of multi-source road data: Complement the accuracy and eliminate errors of the road information of the same road section of several intelligent network-connected data user terminals, output multi-source road information, and calculate the overall evaluation parameter E of the ablation of the multi-dimensional deep road section i .

[0007] S3. Fuzzification of deep road data: Combine multi-source road information and the overall evaluation parameter E of the ablation of the multi-dimensional deep road section i , and calculate the road data fuzzy matrix that captures the dynamic changes and potential non-linear relationships of the road information

[0008] S4. Traffic special event monitoring: Set the critical entropy FuzzyEn for calculating traffic events c ; Combine multi-source road information and the road data fuzzy matrix to calculate the fuzzy entropy; when the fuzzy entropy is greater than the critical entropy, construct a special event list and include it in the special event list for continuous tracking

[0009] Further, in S4, when the fuzzy entropy is greater than the critical entropy, the special event list also records the road data at this time.

[0010] It also includes S5, three-dimensional virtual synchronous execution of traffic special event scenarios: combining map data, Carla UE4, and virtual reality technology, inputting the road data recorded in the special event list, and constructing and outputting the virtual scenario of traffic events.

[0011] Further, in S1, the road information database includes a road asset database, a traffic flow database, and a road condition database; the road asset database includes the number of lanes L n , lane width L w and the speed limit of the road section L v ; the traffic flow database includes traffic flow Q, traffic flow density K, and average speed The road condition database includes driving visibility α, road surface friction coefficient β, and road straightness parameter γ.

[0012] Further, in S2, the road information of the same road section is fused through an ablation algorithm. The ablation parameters of the multi-source data output by this algorithm include the road asset ablation parameter P i , traffic flow ablation parameter T i and road condition ablation parameter C i .

[0013] The calculation formula of the road asset ablation parameter P i is:

[0014]

[0015] In the formula, i is the road section serial number, t is the multi-source data serial number, S is the total number of data sources, L n,t is the number of lanes under the multi-source data sequence t, L w,t is the lane width under the multi-source data sequence t, L v,t is the speed limit of the road section under the multi-source data sequence t, is the road asset offset parameter under the multi-source data sequence t, and the value range is [0, 1], is the multi-dimensional ablation significant parameter of the road asset data, and the multi-dimensional ablation significant parameter satisfies

[0016] The calculation formula of the traffic flow ablation parameter T i is:

[0017]

[0018] In the formula, Q t is the traffic flow under the multi-source data sequence t; K t is the traffic flow density under the multi-source data sequence t; Specify the speed limit for road segments under the multi-source data sequence t; Is the traffic flow offset parameter under the multi-source data sequence t, with a value range of [0, 1]; Is the road asset offset parameter under the multi-source data sequence t, with a value range of [0, 1]; Is the multi-dimensional ablation significant parameter of traffic flow data.

[0019] Traffic flow data ablation parameter C i The calculation formula is:

[0020]

[0021] In the formula, α t Is the driving visibility under the multi-source data sequence t; β t Is the road surface friction coefficient under the multi-source data sequence t; γ t Is the road straightness parameter under the multi-source data sequence t; Is the driving visibility offset parameter under the multi-source data sequence t, with a value range of [0, 1]; Is the road surface friction coefficient offset parameter under the multi-source data sequence t, with a value range of [0, 1]; Is the road straightness offset parameter under the multi-source data sequence t, with a value range of [0, 1]; Is the multi-dimensional ablation significant parameter of traffic flow data.

[0022] Furthermore, for the overall evaluation parameter E of multi-dimensional deep road segment ablation i The calculation formula is:

[0023]

[0024] In the formula, L e (P i ), L e (T i ), L e (C i ) are the Lebesgue integrable functions of the multi-source road asset data ablation parameter, the multi-source traffic flow data ablation parameter, and the multi-source road condition data ablation parameter respectively.

[0025] Furthermore, in S3, the road data fuzzification method includes the following steps:

[0026] S31. Perform deep data fuzzification on the multi-source road asset data ablation parameter P i , the multi-source traffic flow data ablation parameter T i and the multi-source road condition data ablation parameter C i respectively to obtain the fuzzified multi-source data ablation parameter

[0027]

[0028] S32. Based on the fuzzy multi-source data ablation parameters and the multi-dimensional depth section ablation evaluation parameters, a road data fuzzy matrix R is constructed. P,T,C,E , to achieve parametric monitoring of road sections:

[0029]

[0030] Furthermore, in S4, a fuzzy entropy fluctuation identification algorithm for road data is applied to monitor and identify traffic bottlenecks in real time; including calculating the fuzzy membership function, constructing the fuzzy average value Q(i) of each road section with multi-source ablation parameters, and forming a comprehensive fuzzy representation of the road state, including:

[0031] S41. Calculate the fuzzy membership function of the road data by combining the fuzzy multi-source data ablation parameters:

[0032]

[0033] In the formula, x i It represents the result of weighted combination of multi-source data. The formula is:

[0034]

[0035] S42, calculating the average value of the fuzzy membership function for each road section i:

[0036]

[0037] Where i is the serial number of the road section; N is the total number of road sections; the membership function values ​​of all road sections are summarized to calculate the overall fuzzy average value.

[0038] Further, S43, in the multi-source data ablation parameter P i , multi-source traffic data ablation parameter T i , multi-source traffic data ablation parameter C i and the road data fuzzy matrix R P,T,C,E Based on this, the road data fuzzy function Φ is constructed. m (Q(i)), defined as:

[0039]

[0040] Where m is the length of the sliding window, which is used to perform local aggregation analysis on multi-source road data in the time dimension.

[0041] Furthermore, the fuzzy entropy FuzzyEn(i) of road data monitoring is calculated:

[0042]

[0043] Further, in S44, a critical entropy value FuzzyEn is set c , and the critical entropy value is calculated as the average value of the fuzzy entropy of all road segments:

[0044]

[0045] In the formula: i is the road segment serial number; N is the total number of road segments.

[0046] Through the multi-dimensional multi-source deep road data ablation algorithm, the deep road data fuzzy method, and the road data fuzzy entropy fluctuation recognition algorithm, the present invention improves the accuracy of traffic special event monitoring from the approach of processing multi-source road data. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0048] Figure 1 is a flowchart of the present invention;

[0049] Figure 2 is the traffic flow of the recorded traffic special event of sudden heavy rain;

[0050] Figure 3 is the three-dimensional traffic information picture of the traffic special event of sudden heavy rain;

[0051] Figure 4 is the traffic flow of the recorded traffic special event of sudden vehicle failure;

[0052] Figure 5 is the three-dimensional traffic information picture of the traffic special event of sudden vehicle failure. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0053] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0054] Please refer to Figures 1 - 5 , the technical solutions provided by the present invention include the following steps:

[0055] S1. Multi-dimensional road data collection. Based on multiple intelligent network-connected data collection terminals, various information of the monitored road sections is fed back in real time, and a road asset, traffic flow, and road condition database is established for each road section. The road asset database includes the number of lanes L n 、lane width L w 、section speed limit L v . The traffic flow database includes traffic volume Q, traffic flow density K, and average speed . The road condition database includes driving visibility α, road surface friction coefficient β, and road straightness parameter γ, which are specifically as follows:

[0056] It includes data collection devices with a digital intelligent network-connected information transmission system such as high-precision sensing detectors, radar-vision integrated devices, and autonomous driving vehicles. The high-precision sensing detector collects multi-dimensional data in the environment in real time, including data such as temperature, humidity, light, and obstacle distance. The radar-vision integrated device combines the advantages of radar and vision sensors to form a comprehensive perception system, providing various data including driving visibility. The radar works stably under various climate conditions, providing distance and speed information, while the vision sensor captures high-resolution images. The autonomous driving vehicle integrates multiple sensors and computing platforms and can receive, integrate, and transmit numerous perception data, including the number of lanes L n 、lane width L w 、section speed limit L v 、traffic volume Q, traffic flow density K, average speed driving visibility α, road surface friction coefficient β, and road straightness parameter γ, laying a foundation for the deep fusion of multi-source road data.

[0057] S2. Deep fusion of multi-source road data. A multi-dimensional multi-source deep road data ablation algorithm is proposed to perform accuracy complementation and error ablation on the road asset data, traffic flow data, and road condition data of the same road section from multiple data sources, and output the ablation parameters P of multi-source road asset data i 、ablation parameters T of multi-source traffic flow data i and ablation parameters C of multi-source road condition data i , and on this basis, output the overall evaluation parameter E of multi-dimensional deep section ablation i . This fusion process not only improves the accuracy and consistency of the data but also lays a high-quality data foundation for subsequent fuzzification processing. Specifically as follows:

[0058] A multi-source deep road asset data ablation algorithm is proposed to perform data ablation on the number of lanes L n 、lane width L w 、section speed limit L v in the road asset database and output the ablation parameters P of multi-source road asset data i :

[0059]

[0060] where: i is the section serial number, t is the multi-source data serial number, S is the total number of data sources, L n,t is the number of lanes under the multi-source data sequence t, L w,t is the lane width under the multi-source data sequence t, L v,t is the specified speed limit of the section under the multi-source data sequence t, is the road asset offset parameter under the multi-source data sequence t, with a value range of [0, 1], is the multi-dimensional ablation significant parameter of the road asset data, and the multi-dimensional ablation significant parameter satisfies

[0061] A multi-source deep traffic flow data ablation algorithm is proposed to perform data ablation on the traffic flow database, which includes traffic flow Q, traffic flow density K, and average speed to output the multi-source traffic flow data ablation parameter T i :

[0062]

[0063] where: i is the section serial number; t is the multi-source data serial number; S is the total number of data sources; Q t is the traffic flow under the multi-source data sequence t; K t is the traffic flow density under the multi-source data sequence t; is the specified speed limit of the section under the multi-source data sequence t; is the traffic flow offset parameter under the multi-source data sequence t, with a value range of [0, 1]; is the road asset offset parameter under the multi-source data sequence t, with a value range of [0, 1]; is the multi-dimensional ablation significant parameter of the traffic flow data, and the multi-dimensional ablation significant parameter satisfies

[0064] A multi-source deep traffic flow data ablation algorithm is proposed to perform data ablation on the driving visibility α, road surface friction coefficient β, and road straightness parameter γ in the road condition database, and output the multi-source traffic flow data ablation parameter C i :

[0065]

[0066] where: i is the section serial number; t is the multi-source data serial number; S is the total number of data sources; α t is the driving visibility under the multi-source data sequence t; β t is the road surface friction coefficient under the multi-source data sequence t; γ t is the road straightness parameter under the multi-source data sequence t; is the driving visibility offset parameter under the multi-source data sequence t, and its value range is [0, 1]; is the road surface friction coefficient offset parameter under the multi-source data sequence t, and its value range is [0, 1]; is the road straightness offset parameter under the multi-source data sequence t, and its value range is [0, 1]; is the multi-dimensional ablation significant parameter of traffic flow data, and the multi-dimensional ablation significant parameter satisfies

[0067] After separately performing multi-source deep data ablation on the same road section road asset data, traffic flow data, and road condition data of multiple data sources, on the basis of the output multi-source road asset data ablation parameter P i , multi-source traffic flow data ablation parameter T i and multi-source road condition data ablation parameter C i , a multi-dimensional deep road section ablation evaluation algorithm is proposed to output the multi-dimensional deep road section ablation evaluation parameter E i :

[0068]

[0069] In the formula: i is the road section serial number; t is the multi-source data serial number; S is the total number of data sources; is the road asset offset parameter under the multi-source data sequence t, and its value range is [0, 1]; is the traffic flow offset parameter under the multi-source data sequence t, and its value range is [0, 1]; is the road asset offset parameter under the multi-source data sequence t, and its value range is [0, 1]; is the road surface friction coefficient offset parameter under the multi-source data sequence t, and its value range is [0, 1]; is the road straightness offset parameter under the multi-source data sequence t, and its value range is [0, 1]; L e (P i ), L e (T i ), L e (C i are respectively the Lebesgue integrable functions of the multi-source road asset data ablation parameter, multi-source traffic flow data ablation parameter, and multi-source road condition data ablation parameter; are respectively the multi-dimensional ablation significant parameters of road assets, traffic flow, and traffic flow data, and the multi-dimensional ablation significant parameter satisfies

[0070] S3. Deep road data is fuzzy. When outputting the multi-source road asset data ablation parameter P i , multi-source traffic flow data ablation parameter T i , multi-source road condition data ablation parameter C i , and the overall multi-dimensional deep road section ablation evaluation parameter E iBased on this, a deep road data fuzzification method is proposed to obtain the road data fuzzy matrix R P,T,C,E , realizing full-scenario and all-round parametric monitoring of road sections. This fuzzy matrix performs multi-dimensional parameterization on road section data, capturing dynamic changes and potential non-linear relationships, providing a systematic and highly sensitive representation method for subsequent anomaly detection. Specifically as follows:

[0071] S31. Respectively perform deep data fuzzification on the ablation parameter P of multi-source road assets data i , the ablation parameter T of multi-source traffic flow data i and the ablation parameter C of multi-source road condition data i to obtain the fuzzy multi-source data ablation parameters

[0072]

[0073] In the formula: i is the road section serial number; t is the multi-source data serial number; S is the total number of data sources; is the road asset offset parameter under the multi-source data sequence t, with a value range of [0, 1]; is the traffic flow offset parameter under the multi-source data sequence t, with a value range of [0, 1]; is the road asset offset parameter under the multi-source data sequence t, with a value range of [0, 1]; is the road surface friction coefficient offset parameter under the multi-source data sequence t, with a value range of [0, 1]; is the road straightness offset parameter under the multi-source data sequence t, with a value range of [0, 1]; are respectively the multi-dimensional ablation significant parameters of road assets, traffic flow, and vehicle flow data. The multi-dimensional ablation significant parameters satisfy

[0074] S32. Based on the fuzzy multi-source data ablation parameters and the multi-dimensional depth road section ablation evaluation parameters, a road data fuzzy matrix R P,T,C,E is proposed to realize full-scenario and all-round parametric monitoring of road sections:

[0075]

[0076] In the formula: i is the road section serial number; t is the multi-source data serial number; S is the total number of data sources; is the road asset offset parameter under the multi-source data sequence t, with a value range of [0, 1]; is the traffic flow offset parameter under the multi-source data sequence t, with a value range of [0, 1]; is the road asset offset parameter under the multi-source data sequence t, with a value range of [0, 1]; is the road surface friction coefficient offset parameter under the multi-source data sequence t, with a value range of [0, 1]; It is the road straight offset parameter under the multi-source data sequence t, and its value ranges from [0, 1]; They are respectively the multi-dimensional ablation significant parameters of road assets, traffic flow, and vehicle flow data, and the multi-dimensional ablation significant parameters satisfy

[0077] S4. Traffic special event monitoring. Based on the multi-source road asset ablation parameters P i , T i , C i and the road data fuzzy matrix R on the basis of the road data fuzzy matrix P,T,C,E , a road data fuzzy entropy fluctuation recognition algorithm is proposed, including the road data fuzzy membership function A(x i ), the road data fuzzy function Φ m (Q(i)) and the road data monitoring fuzzy entropy Φ m (Q(i)), and the traffic event critical entropy value FuzzyEn c is defined. When the data exceeds the traffic special event critical entropy value, a special event list is constructed, and this fluctuation is listed as a concern in the special event list and continuously tracked, and the road data at this time is recorded. This step is based on the in-depth analysis of multi-source data, realizing real-time section status monitoring and bottleneck identification, and ensuring that the management department can timely learn about potential traffic anomalies. Specifically as follows:

[0078] S41. Combining the fuzzy multi-source data ablation parameters, a road data fuzzy membership function is proposed:

[0079]

[0080] where x i represents the result of weighted combination of multi-source data. The formula is:

[0081]

[0082] In the formula: i is the section serial number; t is the multi-source data serial number; S is the total number of data sources; is the road asset offset parameter under the multi-source data sequence t, and its value ranges from [0, 1]; is the traffic flow offset parameter under the multi-source data sequence t, and its value ranges from [0, 1]; is the road asset offset parameter under the multi-source data sequence t, and its value ranges from [0, 1]; is the road surface friction coefficient offset parameter under the multi-source data sequence t, and its value ranges from [0, 1]; is the road straight offset parameter under the multi-source data sequence t, and its value ranges from [0, 1]; They are respectively the multi-dimensional ablation significant parameters of road assets, traffic flow, and vehicle flow data, and the multi-dimensional ablation significant parameters satisfy

[0083] S42. Calculate the average value of the fuzzy membership function for each section \(i\):

[0084]

[0085] Where: \(i\) is the section serial number; \(N\) is the total number of sections. This step aggregates the membership function values of all sections to obtain the overall fuzzy average value.

[0086] S43. Based on the multi-source road asset data ablation parameter \(P\) i , the multi-source traffic flow data ablation parameter \(T\) i , the multi-source road condition data ablation parameter \(C\) i and the road data fuzzy matrix \(R\) P,T,C,E , propose the road data fuzzy function \(\varPhi\) m (Q(i)), defined as:

[0087]

[0088] And propose the road data monitoring fuzzy entropy FuzzyEn(i):

[0089]

[0090] Where: \(i\) is the section serial number; \(t\) is the multi-source data serial number; \(S\) is the total number of data sources; is the road asset offset parameter under the multi-source data sequence \(t\), with a value range of \([0, 1]\); is the traffic flow offset parameter under the multi-source data sequence \(t\), with a value range of \([0, 1]\); is the road asset offset parameter under the multi-source data sequence \(t\), with a value range of \([0, 1]\); is the road surface friction coefficient offset parameter under the multi-source data sequence \(t\), with a value range of \([0, 1]\); is the road straightness offset parameter under the multi-source data sequence \(t\), with a value range of \([0, 1]\); are the multi-dimensional ablation significant parameters of road assets, traffic flow, and traffic data respectively. The multi-dimensional ablation significant parameters satisfy

[0091] S44. Set up a critical entropy value FuzzyEn c , used to identify special traffic events. When FuzzyEn(i) exceeds this value, it indicates that a special traffic event may occur. The critical entropy value is calculated as the average value of the fuzzy entropy of all sections:

[0092]

[0093] Where: i is the road segment serial number; N is the total number of road segments. This process gradually constructs the algorithm logic from data ablation to event monitoring, ensuring the identification of potential traffic special events through the precise processing of multi-source data.

[0094] S5. Three-dimensional virtual synchronous execution of traffic special event scenarios. Based on high-definition satellite maps, link Carla UE4 with virtual reality technology, input the road data detected by S1 in the special events recorded in S4, and provide multi-perspective outputs of virtual scenarios of traffic events, providing a comprehensive and multi-dimensional picture of traffic special events for the personnel of the management decision-making department. It helps to deeply understand and evaluate the formation process of traffic bottlenecks and provides important support for subsequent decision-making and management measures.

[0095] Next, the present invention will be further described in conjunction with specific embodiments:

[0096] As Figure 1 , for a traffic special event monitoring method based on multi-source road data described in this embodiment, the flow chart is as Figure 1 shown, and the steps are as follows:

[0097] In step S1, based on multiple intelligent network-connected data acquisition terminals, various information of the monitored road segment is fed back in real time, and a road asset, traffic flow, and road condition database for each road segment is established. The road asset database includes the number of lanes L n , lane width L w , road segment speed limit L v , the traffic flow database includes traffic flow Q, traffic flow density K, and average speed , and the road condition database includes driving visibility α, road surface friction coefficient β, and road straightness parameter γ.

[0098] In step S2, a multi-dimensional and multi-source deep road data ablation algorithm is proposed to perform accuracy complementation and error ablation on the road asset data, traffic flow data, and road condition data of the same road segment from multiple data sources, and output the multi-source road asset data ablation parameter P i , multi-source traffic flow data ablation parameter T i , and multi-source road condition data ablation parameter C i , and on this basis, output the multi-dimensional deep road segment ablation overall evaluation parameter E i :

[0099]

[0100]

[0101] In step S3, when outputting the multi-source road asset data ablation parameter P i , multi-source traffic flow data ablation parameter T i , multi-source road condition data ablation parameter C i, and the overall evaluation parameter E of multi-dimensional depth section ablation i Based on i , a depth road data fuzzification method is proposed to obtain the road data fuzzy matrix R P,T,C,E , realizing full-scenario and all-round parameterized monitoring of sections:

[0102]

[0103] In step S4, based on the multi-source road asset ablation parameters P i , T i , C i and on the basis of the road data fuzzy matrix R P,T,C,E , a road data fuzzy entropy fluctuation recognition algorithm is proposed, including the road data fuzzy membership function A(x i ), the road data fuzzy function Φ m (Q(i)) and the road data monitoring fuzzy entropy Φ m (Q(i)), and the traffic event critical entropy value FuzzyEn c is defined:

[0104]

[0105]

[0106] When special traffic events such as sudden heavy rain and sudden vehicle failure are detected after exceeding the traffic special event critical entropy value, the special traffic events of sudden heavy rain and sudden vehicle failure are included in the special event list, and these two special traffic events are listed as the focus in the event list and continuously tracked, and the road data at this time is recorded and exported to the two-dimensional traffic software to form a traffic flow, such as Figure 2 , 4 .

[0107] In step S5, based on the high-definition satellite map, the special traffic events are linked with Carla UE4 and virtual reality technology, and the road data of the special traffic events of sudden heavy rain and sudden vehicle failure recorded are input to provide a multi-perspective output of the virtual scene of the traffic event, providing a comprehensive and multi-dimensional traffic information picture for the management decision-making department personnel, such as Figure 3 , 5 .

[0108] In summary, through the invented multi-dimensional multi-source depth road data ablation algorithm, depth road data fuzzification method and road data fuzzy entropy fluctuation recognition algorithm, this embodiment provides a reliable and practical means for monitoring special traffic events from the perspective of road data.

[0109] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims

1. A traffic special bottleneck monitoring method based on multi-source road data, characterized in that, The following steps are involved: S1. Multi-dimensional road data collection: collect road data from several types of intelligent network data users and establish a road information database for each section; S2. Deep integration of multi-source road data: Complement the accuracy and eliminate the errors of the road information of the same road section at the client side of several intelligent network connection data, output multi-source road information, and calculate the overall evaluation parameter E of the multi-dimensional deep road section ablation i ; S3. Deep road data fuzzification: Combine multi-source road information and multi-dimensional depth section ablation to obtain the overall evaluation parameter E i and calculate the road data fuzzy matrix that captures the dynamic changes and potential non-linear relationships of road information S4. Traffic special event monitoring: Set the critical entropy FuzzyEn for calculating traffic events c ; Combine multi-source road information and the road data fuzzy matrix to calculate the fuzzy entropy; When the fuzzy entropy is greater than the critical entropy, construct a special event list and include it in the special event list for continuous tracking.

2. The traffic special bottleneck monitoring method based on multi-source road data according to claim 1, wherein, S4 also includes that when the fuzzy entropy is greater than the critical entropy, the special event list also records the road data at this time; It also includes S5, three-dimensional virtual synchronous execution of special traffic event scenes: combining map data, CarlaUE4 and virtual reality technology, inputting road data recorded in the special event list, and constructing a virtual scene output of traffic events.

3. The traffic special bottleneck monitoring method based on multi-source road data according to claim 1, wherein In S1, the road information database includes a road asset database, a traffic flow database, and a road condition database; the road asset database includes the number of lanes L n , the lane width L w , and the speed limit L v for a road section; the traffic flow database includes the traffic volume Q, the traffic flow density K, and the average speed The road condition database includes the driving visibility α, the road surface friction coefficient β, and the road straightness parameter γ.

4. The traffic special bottleneck monitoring method based on multi-source road data according to claim 3, wherein In S2, the road information of the same section is fused through an ablation algorithm, and the ablation parameters of the multi-source data output by this algorithm include the road asset ablation parameter P i , the traffic flow ablation parameter T i , and the road condition ablation parameter C i ; Road asset ablation parameter P i The calculation formula is as follows: Wherein, i is the section serial number, t is the multi-source data serial number, S is the total number of data sources, and L n,t is the number of lanes under the multi-source data sequence t, and L w,t is the lane width under the multi-source data sequence t, and L v,t is the specified speed limit of the section under the multi-source data sequence t, is the road asset offset parameter under the multi-source data sequence t, and its value ranges from [0, 1], is the multi-dimensional ablation significance parameter of the road asset data, and the multi-dimensional ablation significance parameter satisfies Traffic flow ablation parameter T i The calculation formula is as follows: Where Q t is the traffic flow under the multi-source data sequence t; K t is the traffic flow density under the multi-source data sequence t; is the specified speed limit of the road section under the multi-source data sequence t; is the traffic flow offset parameter under the multi-source data sequence t, with a value in [0, 1]; is the road asset offset parameter under the multi-source data sequence t, with a value in [0, 1]; is the multi-dimensional ablation significance parameter of the traffic flow data; Traffic flow data ablation parameter C i The calculation formula is as follows: where α t is the driving visibility under the multi-source data sequence t; β t is the road surface friction coefficient under the multi-source data sequence t; γ t is the road straightness parameter under the multi-source data sequence t; is the driving visibility bias parameter under the multi-source data sequence t, with a value range of [0, 1]; is the road surface friction coefficient bias parameter under the multi-source data sequence t, with a value range of [0, 1]; is the road straightness bias parameter under the multi-source data sequence t, with a value range of [0, 1]; is the multi-dimensional ablation significant parameter of the traffic flow data.

5. The traffic special bottleneck monitoring method based on multi-source road data according to claim 4, characterized in that Multi-dimensional depth section ablation overall evaluation parameter E i The calculation formula is as follows: where L e (P i ), L e (T i ), and L e (C i ) are the Lebesgue integrable functions of the ablation parameters of multi-source road asset data, the ablation parameters of multi-source traffic flow data, and the ablation parameters of multi-source road condition data, respectively.

6. The traffic special bottleneck monitoring method based on multi-source road data according to claim 5, characterized in that In S3, the road data fuzzy method includes the following steps: S31. Respectively perform deep data fuzzing on the ablation parameter P of multi-source road asset data i , the ablation parameter T of multi-source traffic flow data i and the ablation parameter C of multi-source road condition data i to obtain the fuzzy multi-source data ablation parameter S32. Based on the fuzzy multi-source data ablation parameters and the multi-dimensional depth section ablation evaluation parameters, construct the road data fuzzy matrix R P,T,C,E , and realize the parameterized monitoring of the section:

7. The traffic special bottleneck monitoring method based on multi-source road data according to claim 6, wherein In S4, the fuzzy entropy fluctuation recognition algorithm of road data is applied to monitor and identify traffic bottlenecks in real time; It includes calculating the fuzzy membership function, constructing the fuzzy average value Q(i) of each road section with multi-source ablation parameters, and forming a comprehensive fuzzy representation of the road state, including: S41. Calculate the fuzzy membership function of the road data by combining the fuzzy multi-source data ablation parameters: where x i represents the result of weighted combination of multi-source data, and the formula is: S42, calculating the average value of the fuzzy membership function for each road section i: Where i is the serial number of the road section; N is the total number of road sections; the membership function values ​​of all road sections are summarized to calculate the overall fuzzy average value.

8. The traffic special bottleneck monitoring method based on multi-source road data according to claim 7, characterized in that It also includes S43, the ablation parameter P of multi-source road asset data i , the ablation parameter T of multi-source traffic flow data i , the ablation parameter C of multi-source road condition data i and the road data fuzzy matrix R P,T,C,E Based on these, construct the road data fuzzy function Φ m (Q(i)), defined as: Where: m is the length of the sliding window.

9. The traffic special bottleneck monitoring method based on multi-source road data according to claim 8, wherein It also includes the calculation of road data monitoring fuzzy entropy FuzzyEn(i):

10. The traffic special bottleneck monitoring method based on multi-source road data according to claim 9, wherein It also includes S44 and sets the critical entropy value FuzzyEn c , and the critical entropy value is calculated as the average value of the fuzzy entropy of all road sections: Where: i is the road section serial number; N is the total number of road sections.

Images