Multi-source data fusion traffic hidden danger identification method and system

Abstract

The invention relates to the technical field of intelligent traffic and traffic safety, in particular to a multi-source data fusion traffic hidden danger recognition method, which comprises the following steps: S100, acquiring a traffic influence factor in a first specified time period corresponding to any traffic unit in an urban traffic network; wherein the traffic unit is a road intersection in an urban traffic network; s200, according to the traffic influence factors in the first specified time period corresponding to the traffic unit, obtaining a hidden danger score of the first specified time period corresponding to the traffic unit; s300, comparing the hidden danger score of the first specified time period corresponding to the traffic unit with a plurality of hidden danger score intervals obtained in advance, and determining a hidden danger score interval to which the hidden danger score of the first specified time period corresponding to the traffic unit belongs; and S400, determining a preset traffic processing plan corresponding to the traffic unit based on the hidden danger score interval to which the hidden danger score of the first specified time period corresponding to the traffic unit belongs.

Description

Technical Field

[0001] The present invention relates to the field of intelligent transportation and traffic safety technology, and in particular to a multi-source data fusion traffic hazard identification method and system. Background Art

[0002] With the acceleration of urbanization, urban traffic networks are becoming more complex, traffic volume is increasing, and road traffic hazards have become an important factor affecting traffic safety and urban operation efficiency. Traditional methods for identifying and managing traffic hazards rely on manual inspections, single data source analysis, or post-event statistics. These methods have the following defects:

[0003] Traditional methods are often unable to collect and analyze complex traffic data in real time, making it difficult to detect potential hazards in a timely manner, resulting in delayed treatment measures. Traditional hazard identification methods usually rely on a single data source (such as accident statistics), ignoring the combined effects of multiple traffic influencing factors such as weather, traffic flow, and road design, resulting in inaccurate hazard identification results. In the existing technology, even if some traffic hazards can be identified, there is a lack of specific traffic handling plans based on the characteristics of the hazards, which cannot effectively alleviate the hazards or avoid accidents. Summary of the invention

[0004] In view of the above-mentioned shortcomings and deficiencies of the prior art, the present invention provides a traffic hazard identification method and system based on multi-source data fusion.

[0005] In order to achieve the above object, the main technical solutions adopted by the present invention include:

[0006] In a first aspect, an embodiment of the present invention provides a method for identifying traffic hazards by fusion of multi-source data, characterized by comprising:

[0007] S100, obtaining traffic influencing factors within a first specified time period corresponding to any traffic unit in the urban traffic network;

[0008] Wherein, the traffic unit is a road crossroad in the urban traffic network;

[0009] S200, obtaining a hidden danger score for a first specified time period corresponding to the traffic unit according to traffic influencing factors in a first specified time period corresponding to the traffic unit;

[0010] S300, comparing the hidden danger score of the first specified time period corresponding to the traffic unit with a plurality of hidden danger score intervals obtained in advance, and determining the hidden danger score interval to which the hidden danger score of the first specified time period corresponding to the traffic unit belongs;

[0011] S400: Determine a preset traffic handling plan corresponding to the traffic unit based on the hidden danger score interval to which the hidden danger score of the first specified time period corresponding to the traffic unit belongs.

[0012] Preferably,

[0013] Traffic influencing factors in the first specified time period corresponding to the traffic unit include: pedestrian saturation X 1 , Non-motor vehicle saturation X 2 , Vehicle saturation X 3 、Number of pedestrian violations X 4 、Number of non-motor vehicle violation incidents X 5 、Number of motor vehicle violations X 6 , signal equipment impact value X 7 , Monitoring equipment impact value X 8 、Sign and marking influence value X 9 , Road construction impact value X 10 、Rain and snow weather impact value X 11 , morning and evening peak impact value X 12 、Holiday impact value X 13 , Building environmental impact value X 14 .

[0014] Preferably,

[0015] Pedestrian Saturation X 1 It is calculated by formula (1);

[0016] The formula (1) is:

[0017] N p is the volume of people in the traffic unit during the first specified time period;

[0018] C p =M p *B p ;

[0019] M p The maximum number of pedestrians allowed to pass through the crosswalk section of the traffic unit;

[0020] B p The number of pedestrian crossings in the traffic unit;

[0021] The non-motor vehicle saturation is calculated by formula (2);

[0022] The formula (2) is:

[0023]

[0024] N mis the number of non-motor vehicles that pass through the traffic unit during the first specified time period;

[0025] C m =M m *B m ;

[0026] M m is the traffic capacity of the lane section at the intersection in the traffic unit;

[0027] B m The number of non-motorized vehicle lanes at the intersection of the traffic unit;

[0028] The motor vehicle saturation is calculated by formula (3);

[0029] The formula (3) is:

[0030]

[0031] N v is the number of motor vehicles in the first specified time period for the traffic unit;

[0032] C v =M v , *B v ;

[0033] M v is the traffic capacity of the lane section at the traffic unit intersection;

[0034] B v is the number of lanes at the traffic unit intersection;

[0035] Number of pedestrian violations X 4 is: the number of pedestrian violation events that occurred in the traffic unit within the first specified time period; the pedestrian violation events include: pedestrians waiting to cross illegally, pedestrians not walking on the sidewalk, and pedestrians running red lights;

[0036] Number of non-motor vehicle violations X 5 is: the number of non-motor vehicle violation events that occurred in the traffic unit within the first specified time period; the non-motor vehicle violation events include: riding an electric scooter without a helmet, riding an electric scooter with a passenger, non-motor vehicle running a red light, and non-motor vehicle driving in the wrong direction;

[0037] Number of motor vehicle violations X 6 is: the number of motor vehicle violations that occurred in the traffic unit during the first specified time period;

[0038] The motor vehicle violations include: motor vehicles not driving according to guidance, motor vehicles changing lanes illegally, motor vehicles not giving way to pedestrians, and motor vehicles running red lights.

[0039] Preferably, the signal device influence value X7 is the ratio of the number of first traffic accidents counted by the traffic unit in the first time period before the first specified time period to the total number of all traffic accidents counted by the traffic unit in the first time period before the first specified time period; the first time period is 7 days; the first traffic accident is a traffic accident unrelated to signal equipment;

[0040] Monitoring equipment impact value X 8 is the ratio of the number of second traffic accidents counted by the traffic unit in the first time period before the first specified time period to the total number of all traffic accidents counted by the traffic unit in the first time period before the first specified time period; the second traffic accident is a traffic accident unrelated to the monitoring device;

[0041] Sign and marking influence value X 9 is the ratio of the number of third traffic accidents counted by the traffic unit in the first time period before the corresponding designated time period to the total number of all traffic accidents counted by the traffic unit in the first time period before the corresponding designated time period; the third traffic accident is a traffic accident unrelated to signs and markings;

[0042] Road construction impact value X 10 is the ratio of the number of fourth traffic accidents counted by the traffic unit in the first time period before the corresponding designated time period to the total number of all traffic accidents counted by the traffic unit in the first time period before the corresponding designated time period; the fourth traffic accident is a traffic accident unrelated to road construction;

[0043] Rain and snow weather impact value X 11 is the ratio of the number of the fifth traffic accidents counted by the traffic unit in the first time period before the corresponding designated time period to the total number of all traffic accidents counted by the traffic unit in the first time period before the corresponding designated time period; the fifth traffic accident is a traffic accident not related to rain or snow;

[0044] Morning and evening peak impact value X 12 The ratio of the number of the sixth traffic accidents counted by the traffic unit in the first time period before the corresponding designated time period to the total number of all traffic accidents counted by the traffic unit in the first time period before the corresponding designated time period; the sixth traffic accident is a traffic accident unrelated to the morning and evening rush hours;

[0045] Holiday impact value X 13is the ratio of the number of the seventh traffic accidents counted in the first time period before the corresponding designated time period of the traffic unit to the total number of all traffic accidents counted in the first time period before the corresponding designated time period of the traffic unit; the seventh traffic accident is a traffic accident not related to holidays;

[0046] Building environmental impact value X 14 It is the ratio of the number of the seventh traffic accidents counted in the first time period before the corresponding designated time period of the traffic unit to the total number of all traffic accidents counted in the first time period before the corresponding designated time period of the traffic unit; the seventh traffic accident is a traffic accident in which the location of occurrence is more than a preset distance from a designated building; the designated buildings include schools, hospitals, and supermarkets.

[0047] Preferably, the S200 specifically includes:

[0048] According to the traffic influencing factors in the first specified time period corresponding to the traffic unit, the hidden danger score in the first specified time period of the traffic unit is obtained using formula (4);

[0049] The formula (4) is:

[0050] Da=w a ×(q 1 +...q i ...+q 14 );

[0051] Among them, Da is the hidden danger score of the ath traffic unit in the urban traffic network in the first specified time period;

[0052] w a is the weight value of the importance corresponding to the ath traffic unit in the urban traffic network;

[0053] q i X is the factor affecting traffic i Corresponding normalized value; i∈(1, 2, ..., 14).

[0054] Preferably, the traffic influencing factor X i The corresponding normalized value q i It is calculated by formula (5);

[0055] The formula (5) is:

[0056]

[0057] X imin It is the hidden danger influencing factor X in the second time period before the first specified time period of the traffic unit i The minimum value of

[0058] X imax It is the hidden danger influencing factor X in the second time period before the first specified time period of the traffic unit i The maximum value of .

[0059] Preferably, the weight value w corresponding to the importance of the ath traffic unit in the urban traffic network is a It is calculated by formula (6);

[0060] The formula (6) is:

[0061]

[0062] Where n is the number of transportation units in the urban transportation network;

[0063] e a =b 1 F a +b 2 G a +b 3 A a ;

[0064] e a is the first parameter value corresponding to the ath traffic unit in the urban traffic network;

[0065] F a is the total number of vehicles passing through the ath traffic unit in the urban traffic network in a preset historical time period;

[0066] G a is the road congestion index corresponding to the ath traffic unit in the urban traffic network counted in advance within a preset historical time period;

[0067] G a =1-(average speed of vehicles passing through the traffic unit in the historical time period / maximum allowed speed corresponding to the traffic unit);

[0068] A a It is the average waiting time of vehicles in a pre-set historical time period corresponding to the ath traffic unit in the urban traffic network when passing through the traffic unit under red light or congestion conditions.

[0069] Preferably, before S100, the method further includes:

[0070] S01. Obtaining traffic influencing factors within a second specified time period corresponding to any traffic unit in the urban traffic network;

[0071] The second specified time period is before the first specified time period;

[0072] S02. Obtaining a hidden danger score for the second specified time period of the traffic unit according to the traffic influencing factors for the second specified time period corresponding to the traffic unit;

[0073] S03. Obtain multiple hidden danger score intervals according to the hidden danger score of the second specified time period corresponding to each traffic unit in the urban traffic network.

[0074] Preferably, the S03 specifically includes:

[0075] S03-1. Obtain an initial sample set D according to the hidden danger score of each traffic unit in the urban traffic network in the second specified time period;

[0076] D={D1, D2, D3,..., Dn};

[0077] Where Dn is the hidden danger score of the nth traffic unit in the urban traffic network in the second specified time period;

[0078] S03-2, randomly selecting t traffic units' hidden danger scores within the second specified time period from the initial sample set D and duplicating them respectively, and putting the duplicated hidden danger scores into the initial sample set D to obtain a new training set;

[0079] The new training set includes the hidden danger scores of n+t traffic units in the second specified time period;

[0080] S03-3, performing step S03-2 r times, and obtaining r new training sets respectively;

[0081] Where r = h - 1;

[0082] S03-4, combining the new training set obtained in step S03-2 and the r new training sets obtained in step S03-3 to obtain a sample set group;

[0083] S03-5. Based on the sample set, multiple hidden danger score intervals are obtained;

[0084] Wherein, the S03-5 specifically includes:

[0085] Using a clustering algorithm to divide all hidden danger scores in the sample set into k groups;

[0086] For each group's hidden danger score, obtain the hidden danger score interval corresponding to the group;

[0087] Among them, the hidden danger score interval corresponding to each group is from the minimum value of the hidden danger score of the group to the maximum value of the hidden danger score of the group.

[0088] On the other hand, this embodiment also provides a traffic hazard identification system based on multi-source data fusion, comprising: at least one processor; and at least one memory communicatively connected to the processor, wherein the memory stores program instructions executable by the processor, and the processor calls the program instructions to execute the traffic hazard identification method based on multi-source data fusion as in the first aspect.

[0089] The beneficial effects of the present invention are:

[0090] The traffic hazard identification method of the present invention using multi-source data fusion adopts a hazard scoring and zoning method based on multi-source data fusion technology. By obtaining the traffic influencing factors of the traffic unit within a specified time period and performing scoring and zoning, compared with the existing technology, it can comprehensively consider multiple traffic influencing factors, improve the accuracy and real-time performance of hazard identification, and achieve the effect of comprehensively evaluating traffic hazards and providing targeted treatment plans.

[0091] The traffic hazard identification method of the present invention using multi-source data fusion can automatically trigger traffic handling plans through hazard scoring and interval matching. Compared with the prior art, it can reduce manual intervention, improve the degree of automation of traffic management, and achieve the effect of quickly responding to traffic hazards.

[0092] The traffic hazard identification method of the present invention by multi-source data fusion adopts the calculation model of pedestrian saturation, non-motor vehicle saturation and motor vehicle saturation. Compared with the prior art, it can achieve the effect of comprehensively reflecting the traffic capacity of road intersections by accurately quantifying the dynamic characteristics of traffic flow. In addition, since the violation situation of the traffic unit is refined and counted by the number of pedestrian violation events, the number of non-motor vehicle violation events and the number of motor vehicle violation events, the violation behavior characteristics of different traffic subjects can be accurately identified, and the effect of improving the efficiency of accurate identification of traffic hazards is achieved. At the same time, the calculation of the influence value of various facilities such as signal equipment, monitoring equipment, signs and markings is introduced, and the quantification of environmental factors such as road construction, rainy and snowy weather, morning and evening rush hours and holidays is adopted, and the environmental impact value of buildings is targetedly quantified, so as to achieve the effect of improving the accuracy of traffic hazard prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0093] Figure 1 A flow chart of a traffic hazard identification method based on multi-source data fusion according to the present invention;

[0094] Figure 2 This is a connection diagram of the detection and identification device on the traffic unit in the second embodiment of the present invention. DETAILED DESCRIPTION

[0095] In order to better explain the present invention and facilitate understanding, the present invention is described in detail below through specific implementation modes in conjunction with the accompanying drawings.

[0096] In order to better understand the above technical solution, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided to enable a clearer and more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0097] Embodiment 1

[0098] See also Figure 1 This embodiment provides a traffic hazard identification method based on multi-source data fusion, including:

[0099] S100, obtaining traffic influencing factors within a first specified time period corresponding to any traffic unit in the urban traffic network;

[0100] Wherein, the traffic unit is a road crossroad in the urban traffic network;

[0101] Traffic influencing factors in the first specified time period corresponding to the traffic unit include: pedestrian saturation X 1 , Non-motor vehicle saturation X 2 , Vehicle saturation X 3 、Number of pedestrian violations X 4 、Number of non-motor vehicle violation incidents X 5 、Number of motor vehicle violations X 6 , signal equipment impact value X 7 , Monitoring equipment impact value X 8 、Sign and marking influence value X 9 , Road construction impact value X 10 、Rain and snow weather impact value X 11 , morning and evening peak impact value X 12 、Holiday impact value X 13 , Building environmental impact value X 14 .

[0102] A total of 14 traffic influencing factors are included (such as pedestrian saturation, number of motor vehicle violation incidents, signal equipment impact value, etc.). The integration of multi-dimensional data can more comprehensively reflect the potential hazards of traffic units.

[0103] Pedestrian Saturation X 1 It is calculated by formula (1); the formula (1) is:

[0104] Np is the passenger flow of the traffic unit in the first specified time period; C p =M p *B p ;M p The maximum number of pedestrians allowed to pass through the crosswalk section of the traffic unit; B p The number of pedestrian crossings in the traffic unit;

[0105] The non-motor vehicle saturation is calculated by formula (2); the formula (2) is:

[0106]

[0107] N m is the number of non-motor vehicles passing through the traffic unit in the first specified time period; C m =M m *B m ;M m B is the traffic capacity of the lane section at the intersection in the traffic unit; m The number of non-motorized vehicle lanes at the intersection of the traffic unit;

[0108] The motor vehicle saturation is calculated by formula (3); the formula (3) is:

[0109]

[0110] N v is the number of motor vehicles in the traffic unit in the first specified time period; C v =M v *B v ;M v B is the traffic capacity of the lane section at the traffic unit intersection; v is the number of lanes at the traffic unit intersection;

[0111] Number of pedestrian violations X 4 is: the number of pedestrian violation events that occurred in the traffic unit within the first specified time period; the pedestrian violation events include: pedestrians waiting for traffic violations, pedestrians not walking on the sidewalk, and pedestrians running red lights; it should be noted that pedestrians waiting for traffic violations means that pedestrians do not stand in the waiting area as required when waiting for traffic, and may occupy the motor vehicle lane or non-motor vehicle lane. Pedestrians not walking on the sidewalk means that pedestrians do not follow the prescribed path, such as crossing the road in places without zebra crossings or sidewalks. Pedestrians running red lights means that pedestrians forcefully cross the road during the red light signal and fail to comply with traffic signal rules.

[0112] Number of non-motor vehicle violations X 5is: the number of non-motor vehicle violation events that occurred in the traffic unit within the first specified time period; the non-motor vehicle violation events include: riding an electric scooter without a helmet, riding an electric scooter with a passenger, non-motor vehicle running a red light, and non-motor vehicle driving in the wrong direction;

[0113] Number of motor vehicle violations X 6 It is: the number of motor vehicle violation events that occurred in the traffic unit within the first specified time period; the motor vehicle violation events include: motor vehicles not driving according to guidance, motor vehicles changing lanes illegally, motor vehicles not giving way to pedestrians, and motor vehicles running red lights.

[0114] Signal equipment impact value X 7 is the ratio of the number of first traffic accidents counted by the traffic unit in the first time period before the first specified time period to the total number of all traffic accidents counted by the traffic unit in the first time period before the first specified time period; the first time period is 7 days; the first traffic accident is a traffic accident unrelated to the signal device; in this embodiment, the signal device influence value X 7 It is used to measure the influence of traffic signal equipment in a certain period of time, that is, the normal operation and effectiveness of the equipment. Among them, signal equipment specifically refers to the hardware equipment that controls traffic signals, such as traffic lights, traffic signal controllers, signal poles, light bulbs, signal sensors, etc. They indicate the right of way and prohibited time for vehicles and pedestrians by setting different signal cycles (red light, green light, yellow light) to improve traffic flow and safety.

[0115] Monitoring equipment impact value X 8 It is the ratio of the number of the second traffic accidents counted in the first time period before the first specified time period of the traffic unit to the total number of all traffic accidents counted in the first time period before the first specified time period of the traffic unit; the second traffic accident is a traffic accident unrelated to the monitoring equipment; in this embodiment, the monitoring equipment refers to the equipment installed in the traffic unit (such as an intersection or road area) for traffic monitoring and management. These devices are mainly used for real-time monitoring of traffic flow, recording violations, collecting traffic data and assisting in accident analysis. An accident unrelated to the monitoring equipment refers to a traffic accident in which the effectiveness or existence of the monitoring equipment has nothing to do with the occurrence of the accident. For example, an accident caused by the driver's personal error or an area that the equipment cannot cover.

[0116] Sign and marking influence value X 9It is the ratio of the number of the third traffic accidents counted in the first time period before the corresponding designated time period of the traffic unit to the total number of all traffic accidents counted in the first time period before the corresponding designated time period of the traffic unit; the third traffic accident is a traffic accident unrelated to signs and markings; signs and markings refer to various traffic signs and road markings on the road used to convey information and guide traffic participants (such as drivers, pedestrians, and non-motor vehicle drivers). These facilities play a vital role in ensuring traffic safety, optimizing traffic flow, and guiding traffic behavior. Traffic accidents unrelated to signs and markings refer to situations where the setting or status of signs and markings has no direct relationship with the occurrence of the accident. For example, the driver did not drive according to the traffic signal, but the signs and markings were clear and functioned normally.

[0117] Road construction impact value X 10 It is the ratio of the number of the fourth traffic accidents counted in the first time period before the corresponding designated time period of the traffic unit to the total number of all traffic accidents counted in the first time period before the corresponding designated time period of the traffic unit; the fourth traffic accident is a traffic accident unrelated to road construction; in this embodiment, the traffic accident unrelated to road construction refers to an accident that has no direct relationship with the road construction activities. For example, the accident occurred in an area without construction, or was caused by the driver's own violation. Road construction refers to the activities of building, repairing, renovating or maintaining roads. Road construction usually involves the construction or maintenance of pavements, bridges, tunnels or other traffic facilities, with the aim of improving road conditions, enhancing traffic safety or extending the life of roads.

[0118] Rain and snow weather impact value X 11 It is the ratio of the number of the fifth traffic accidents counted in the first time period before the corresponding designated time period of the traffic unit to the total number of all traffic accidents counted in the first time period before the corresponding designated time period of the traffic unit; the fifth traffic accident is a traffic accident unrelated to rain or snow; traffic accidents unrelated to rain or snow refer to traffic accidents that are not directly affected by rain or snow. The causes of these accidents are mainly related to the driver, vehicle conditions, road conditions or other factors, rather than the influence of rain or snow.

[0119] Morning and evening peak impact value X 12It is the ratio of the number of the sixth traffic accidents counted in the first time period before the corresponding designated time period of the traffic unit to the total number of all traffic accidents counted in the first time period before the corresponding designated time period of the traffic unit; the sixth traffic accident is a traffic accident unrelated to the morning and evening peaks; traffic accidents unrelated to the morning and evening peaks refer to traffic accidents that occur during non-morning and evening peaks with low traffic flow. The causes of these accidents are unrelated to typical peak characteristics such as traffic congestion, heavy traffic flow or rushing to catch up during the morning and evening peaks.

[0120] Holiday impact value X 13 is the ratio of the number of the seventh traffic accidents counted in the first time period before the corresponding designated time period of the traffic unit to the total number of all traffic accidents counted in the first time period before the corresponding designated time period of the traffic unit; the seventh traffic accident is a traffic accident not related to holidays;

[0121] Building environmental impact value X 14 It is the ratio of the number of the seventh traffic accidents counted in the first time period before the corresponding designated time period of the traffic unit to the total number of all traffic accidents counted in the first time period before the corresponding designated time period of the traffic unit; the seventh traffic accident is a traffic accident in which the location of occurrence is more than a preset distance from a designated building; the designated buildings include schools, hospitals, and supermarkets.

[0122] S200, obtaining a hidden danger score for a first specified time period corresponding to the traffic unit according to traffic influencing factors in a first specified time period corresponding to the traffic unit;

[0123] The S200 specifically includes:

[0124] According to the traffic influencing factors in the first specified time period corresponding to the traffic unit, the hidden danger score in the first specified time period of the traffic unit is obtained using formula (4);

[0125] The formula (4) is:

[0126] Da=w a ×(q 1 +...q i ...+q 14 );

[0127] Among them, Da is the hidden danger score of the ath traffic unit in the urban traffic network in the first specified time period;

[0128] w a is the weight value of the importance corresponding to the ath traffic unit in the urban traffic network; the weight value of importance (w a), and make personalized adjustments based on the historical data and actual conditions of each traffic unit, thus improving the pertinence and applicability of the evaluation results.

[0129] q i X is the factor affecting traffic i Corresponding normalized value; i∈(1, 2, ..., 14).

[0130] Traffic influencing factors X i The corresponding normalized value q i It is calculated by formula (5);

[0131] The formula (5) is:

[0132]

[0133] X imin It is the hidden danger influencing factor X in the second time period before the first specified time period of the traffic unit i The minimum value of

[0134] X imax It is the hidden danger influencing factor X in the second time period before the first specified time period of the traffic unit i The maximum value of .

[0135] The weight value w of the importance corresponding to the a-th traffic unit in the urban traffic network a It is calculated by formula (6);

[0136] The formula (6) is:

[0137]

[0138] Where n is the number of transportation units in the urban transportation network;

[0139] e a =b 1 F a +b 2 G a +b 3 A a ;

[0140] e a is the first parameter value corresponding to the ath traffic unit in the urban traffic network;

[0141] F a is the total number of vehicles passing through the ath traffic unit in the urban traffic network in a preset historical time period;

[0142] G ais the road congestion index corresponding to the ath traffic unit in the urban traffic network counted in advance within a preset historical time period;

[0143] G a =1-(average speed of vehicles passing through the traffic unit in the historical time period / maximum allowed speed corresponding to the traffic unit);

[0144] A a It is the average waiting time of vehicles in a pre-set historical time period corresponding to the ath traffic unit in the urban traffic network when passing through the traffic unit under red light or congestion conditions.

[0145] In this embodiment, specific mathematical formulas (Formulas 4, 5, and 6) are used to calculate the hidden danger scores of traffic units, making the evaluation process more objective and repeatable, and reducing the influence of human factors. By comprehensively considering multiple traffic influencing factors (such as the total number of vehicles, road congestion indicators, waiting time, etc.) and normalizing them, fair comparison and comprehensive evaluation between different factors are ensured.

[0146] The hidden danger score is calculated using formula (4), and the unit and magnitude differences are eliminated through normalization formula (5). The importance weight (formula 6) is added to adjust the score according to the historical data and actual situation of the traffic unit. The weight value is adjusted according to the actual situation to ensure that key areas or high-risk traffic units are evaluated more accurately.

[0147] S300, comparing the hidden danger score of the first specified time period corresponding to the traffic unit with a plurality of hidden danger score intervals obtained in advance, and determining the hidden danger score interval to which the hidden danger score of the first specified time period corresponding to the traffic unit belongs;

[0148] In this embodiment, the hidden danger scores are classified into different risk levels through pre-defined hidden danger scoring intervals, thereby realizing the automation and standardization of hidden danger assessment and greatly reducing the errors and inconsistencies of manual subjective judgment.

[0149] S400: Determine the preset traffic handling plan corresponding to the traffic unit based on the hidden danger score interval to which the hidden danger score of the first specified time period corresponding to the traffic unit belongs. In this embodiment, the hidden danger score is directly associated with the handling plan, and appropriate response measures can be quickly selected without temporary analysis or judgment. This shortens the time from risk assessment to plan implementation, and improves the response speed of the traffic management department to hidden dangers.

[0150] Specifically, before S100, it also includes:

[0151] S01. Obtaining traffic influencing factors within a second specified time period corresponding to any traffic unit in the urban traffic network;

[0152] The second specified time period is before the first specified time period;

[0153] S02. Obtaining a hidden danger score for the second specified time period of the traffic unit according to the traffic influencing factors for the second specified time period corresponding to the traffic unit;

[0154] S03. Obtain multiple hidden danger score intervals according to the hidden danger score of the second specified time period corresponding to each traffic unit in the urban traffic network.

[0155] In the practical application of this embodiment, S03 specifically includes:

[0156] S03-1. Obtain an initial sample set D according to the hidden danger score of each traffic unit in the urban traffic network in the second specified time period;

[0157] D={D1, D2, D3,..., Dn};

[0158] Where Dn is the hidden danger score of the nth traffic unit in the urban traffic network in the second specified time period;

[0159] S03-2, randomly selecting t traffic units' hidden danger scores within the second specified time period from the initial sample set D and duplicating them respectively, and putting the duplicated hidden danger scores into the initial sample set D to obtain a new training set;

[0160] The new training set includes the hidden danger scores of n+t traffic units in the second specified time period;

[0161] S03-3, performing step S03-2 r times, and obtaining r new training sets respectively;

[0162] Where r = h - 1;

[0163] S03-4, combining the new training set obtained in step S03-2 and the r new training sets obtained in step S03-3 to obtain a sample set group;

[0164] S03-5. Based on the sample set, multiple hidden danger score intervals are obtained;

[0165] Wherein, the S03-5 specifically includes:

[0166] Using a clustering algorithm to divide all hidden danger scores in the sample set into k groups;

[0167] For each group's hidden danger score, obtain the hidden danger score interval corresponding to the group;

[0168] Among them, the hidden danger score interval corresponding to each group is from the minimum value of the hidden danger score of the group to the maximum value of the hidden danger score of the group.

[0169] In this embodiment, by clustering analysis of the hidden danger scores of the traffic units, the scores are divided into multiple groups according to similarity, and the hidden danger score interval is generated based on the distribution range of each group of hidden danger scores.

[0170] The clustering algorithm in this embodiment can identify the natural clustering pattern of hidden danger score data (for example, the difference between low risk scores and high risk scores) according to the distribution characteristics of hidden danger scores. The generated intervals can accurately reflect the actual distribution of hidden danger scores, thereby avoiding the inaccuracy that may be caused by artificial division of intervals. The accuracy of the intervals helps to more accurately assess the hidden danger risk level of the traffic unit and support the reasonable formulation of subsequent treatment plans. In this embodiment, by randomly extracting and copying the scoring data in the sample set (steps S03-2 and S03-3), the data sample is expanded and multiple training sets are generated, which helps to improve the robustness of the model.

[0171] Random sampling and replication increase the diversity of data and reduce the bias that may be caused by small sample problems. Generating new training sets multiple times (i.e., multiple random sampling and expansion) further reduces the impact of individual outliers on the results. The diversity of the sample set group enhances the stability of the scoring interval and avoids the instability of the division results caused by insufficient or single samples. By generating multiple new training sets (step S03-4), a sample set group is constructed, so that the division results of the hidden danger scoring interval are more generalized. The sample set group combines the characteristics of multiple training sets and covers a variety of possible distributions of traffic unit hidden danger scores. This method can better cope with possible changes in the distribution of hidden danger scores in the future and improve the universality of the hidden danger interval division results. Hidden danger scoring intervals with strong generalization ability can more effectively guide the hidden danger assessment of multiple traffic units. By using a clustering algorithm to automatically divide the hidden danger scoring interval, the need for human intervention is reduced and errors caused by subjectivity are avoided. The clustering algorithm groups according to the characteristics of the data itself, without the need to manually specify the division rules of the scoring interval. The automated interval division process can significantly improve efficiency while maintaining the objectivity and consistency of the interval division. Reducing manual involvement can effectively reduce operating costs and improve the repeatability of segmentation results.

[0172] On the other hand, this embodiment also provides a traffic hazard identification system based on multi-source data fusion, comprising: at least one processor; and at least one memory communicatively connected to the processor, wherein the memory stores program instructions executable by the processor, and the processor calls the program instructions to execute the traffic hazard identification method based on multi-source data fusion described in Embodiment 1.

[0173] A traffic hazard identification method using multi-source data fusion in this embodiment adopts a hazard scoring and zoning method based on multi-source data fusion technology. By obtaining the traffic influencing factors of the traffic unit within a specified time period and performing scoring and zoning, compared with the existing technology, it can comprehensively consider multiple traffic influencing factors, improve the accuracy and real-time performance of hazard identification, and achieve the effect of comprehensively evaluating traffic hazards and providing targeted treatment plans.

[0174] Embodiment 2

[0175] This embodiment provides a traffic hazard identification method based on multi-source data fusion, including:

[0176] Step 1: Obtain traffic influencing factors within a first specified time period corresponding to any traffic unit in the urban traffic network;

[0177] Wherein, the traffic unit is a road crossroad in the urban traffic network;

[0178] Traffic influencing factors in the first specified time period corresponding to the traffic unit include: pedestrian saturation X 1 , Non-motor vehicle saturation X 2 , Vehicle saturation X 3 、Number of pedestrian violations X 4 、Number of non-motor vehicle violation incidents X 5 、Number of motor vehicle violations X 6 , signal equipment impact value X 7 , Monitoring equipment impact value X 8 、Sign and marking influence value X 9 , Road construction impact value X 10 、Rain and snow weather impact value X 11 , morning and evening peak impact value X 12 、Holiday impact value X 13 , Building environmental impact value X 14 .

[0179] The traffic influencing factors in this embodiment can be obtained through the detection and identification device installed on the traffic unit, wherein the detection and identification device installed on the traffic unit includes a detection camera and an edge computing station. The edge station performs intelligent structured analysis on the real-time video stream, extracts structured information of the active target from it, and analyzes various traffic events such as pedestrian, non-motor vehicle flow, motor vehicle flow, and pedestrian running red lights, non-motor vehicle running red lights, and non-motor vehicle reverse driving.

[0180] The principle of the detection device installation and illumination direction is to have the highest possible vehicle recognition rate on the entrance and exit lanes, and to detect the direction of the opposite intersection entrance and exit lanes. The detection camera can be installed on the electric police pole or traffic light pole at the intersection, depending on the actual situation. Since there is a blind spot of about 10-15 meters below its installation position, it is recommended that the opposite illumination camera can complement and cover the blind spot. Installation angle: To minimize the blind spot, the installation angle should be directly facing the entrance and exit lanes of the intersection. When installed on the traffic light pole, consider pressing it down by about 5 degrees.

[0181] In order to increase the data transmission speed, the traffic unit uses a combination of local area network and device-specific network. Figure 2 As shown, the detection camera is configured with a LAN switch, the edge computing station is configured with a LAN and a device-specific network, and the video data collected by the detection camera is uploaded to the edge computing station. After the edge computing station calculates and processes the data, it is uploaded to the central cloud platform through the public network switch.

[0182] Step 2: According to the traffic influencing factors in the first specified time period corresponding to the traffic unit, obtain the hidden danger score of the first specified time period corresponding to the traffic unit;

[0183] Step 3: Compare the hidden danger score of the first specified time period corresponding to the traffic unit with a plurality of hidden danger score intervals obtained in advance, and determine the hidden danger score interval to which the hidden danger score of the first specified time period corresponding to the traffic unit belongs;

[0184] Step 4: Determine a preset traffic handling plan corresponding to the traffic unit based on the hidden danger score interval to which the hidden danger score of the first specified time period corresponding to the traffic unit belongs.

[0185] The following are some examples:

[0186] The detection equipment (including detection cameras and edge computing stations) installed on traffic unit A collects traffic influencing factors within the first specified time period corresponding to traffic unit A. Then, based on the traffic influencing factors within the first specified time period corresponding to the traffic unit, the hidden danger score of the first specified time period corresponding to the traffic unit is obtained. The hidden danger score is compared with the preset score interval. If it is found that the hidden danger score belongs to one of the score intervals, the preset traffic handling plan corresponding to this score interval is automatically selected, for example: Traffic light optimization: Increase the duration of green lights during the morning rush hour to improve traffic efficiency. Pedestrian management: Strengthen voice reminder devices in pedestrian areas to reduce red light running. Non-motor vehicle regulation: Deploy traffic police or drone patrols to reduce non-motor vehicle violations, etc.

[0187] A traffic hazard identification method using multi-source data fusion in this embodiment can automatically trigger traffic handling plans through hazard scoring and interval matching. Compared with the existing technology, it can reduce manual intervention, improve the degree of automation of traffic management, and achieve the effect of quickly responding to traffic hazards.

[0188] In the description of the present invention, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0189] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0190] In the present invention, unless otherwise clearly specified and limited, when a first feature is “on” or “below” a second feature, it may be that the first and second features are in direct contact, or that the first and second features are in indirect contact through an intermediate medium. Moreover, when a first feature is “above”, “above” or “above” a second feature, it may be that the first feature is directly above or obliquely above the second feature, or it may simply mean that the first feature is higher in level than the second feature. When a first feature is “below”, “below” or “below” a second feature, it may be that the first feature is directly below or obliquely below the second feature, or it may simply mean that the first feature is lower in level than the second feature.

[0191] In the description of this specification, the description of the terms "one embodiment", "some embodiments", "embodiment", "example", "specific example" or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

[0192] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may alter, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A traffic hazard identification method based on multi-source data fusion, characterized in that: include: S100, obtaining traffic influencing factors within a first specified time period corresponding to any traffic unit in the urban traffic network; Wherein, the traffic unit is a road crossroad in the urban traffic network; S200, obtaining a hidden danger score for a first specified time period corresponding to the traffic unit according to traffic influencing factors in a first specified time period corresponding to the traffic unit; S300, comparing the hidden danger score of the first specified time period corresponding to the traffic unit with a plurality of hidden danger score intervals obtained in advance, and determining the hidden danger score interval to which the hidden danger score of the first specified time period corresponding to the traffic unit belongs; S400: Determine a preset traffic handling plan corresponding to the traffic unit based on the hidden danger score interval to which the hidden danger score of the first specified time period corresponding to the traffic unit belongs.

2. The traffic hazard identification method based on multi-source data fusion according to claim 1 is characterized in that: The traffic influencing factors in the first specified time period corresponding to the traffic unit include: pedestrian saturation X1, non-motor vehicle saturation X2, motor vehicle saturation X3, pedestrian violation event number X4, non-motor vehicle violation event number X5, motor vehicle violation event number X6, signal equipment impact value X7, monitoring equipment impact value X8, sign and marking impact value X9, road construction impact value X10, and road traffic flow rate X11. 10 、Rain and snow weather impact value X 11 , morning and evening peak impact value X 12 、Holiday impact value X 13 , Building environmental impact value X 14 .

3. The traffic hazard identification method based on multi-source data fusion according to claim 2 is characterized in that: The pedestrian saturation X1 is calculated by formula (1); The formula (1) is: N p is the volume of people in the traffic unit during the first specified time period; C p =M p *B p ; M p The maximum number of pedestrians allowed to pass through the crosswalk section of the traffic unit; B p The number of pedestrian crossings in the traffic unit; The non-motor vehicle saturation is calculated by formula (2); The formula (2) is: N m is the number of non-motor vehicles that pass through the traffic unit during the first specified time period; C m =M m *B m ; M m is the traffic capacity of the lane section at the intersection in the traffic unit; B m The number of non-motorized vehicle lanes at the intersection of the traffic unit; The motor vehicle saturation is calculated by formula (3); The formula (3) is: N v is the number of motor vehicles in the first specified time period for the traffic unit; C v =M v *B v ; M v is the traffic capacity of the lane section at the traffic unit intersection; B v is the number of lanes at the traffic unit intersection; The number of pedestrian violation events X4 is: the number of pedestrian violation events that occurred in the traffic unit during the first specified time period; The pedestrian violation incidents include: pedestrians waiting to cross illegally, pedestrians not walking on the sidewalk, and pedestrians running red lights; The number of non-motor vehicle violation events X5 is: the number of non-motor vehicle violation events that occurred in the traffic unit within the first specified time period; the non-motor vehicle violation events include: riding an electric scooter without a helmet, riding an electric scooter with a passenger, non-motor vehicle running a red light, and non-motor vehicle driving in the wrong direction; The number of motor vehicle violation events X6 is: the number of motor vehicle violation events that occurred in the traffic unit during the first specified time period; The motor vehicle violations include: motor vehicles not driving according to guidance, motor vehicles changing lanes illegally, motor vehicles not giving way to pedestrians, and motor vehicles running red lights.

4. The method for identifying traffic hazards by multi-source data fusion according to claim 3 is characterized in that: The signal equipment influence value X7 is the ratio of the number of first traffic accidents counted by the traffic unit in the first time period before the first specified time period to the total number of all traffic accidents counted by the traffic unit in the first time period before the first specified time period; the first time period is 7 days; the first traffic accident is a traffic accident unrelated to the signal equipment; The monitoring device influence value X8 is the ratio of the number of second traffic accidents counted by the traffic unit in the first time period before the first specified time period to the total number of all traffic accidents counted by the traffic unit in the first time period before the first specified time period; the second traffic accident is a traffic accident unrelated to the monitoring device; The sign and marking influence value X9 is the ratio of the number of third traffic accidents counted by the traffic unit in the first time period before the corresponding designated time period to the total number of all traffic accidents counted by the traffic unit in the first time period before the corresponding designated time period; the third traffic accident is a traffic accident unrelated to the sign and marking; Road construction impact value X 10 is the ratio of the number of fourth traffic accidents counted by the traffic unit in the first time period before the corresponding designated time period to the total number of all traffic accidents counted by the traffic unit in the first time period before the corresponding designated time period; the fourth traffic accident is a traffic accident unrelated to road construction; Rain and snow weather impact value X 11 is the ratio of the number of the fifth traffic accidents counted by the traffic unit in the first time period before the corresponding designated time period to the total number of all traffic accidents counted by the traffic unit in the first time period before the corresponding designated time period; the fifth traffic accident is a traffic accident not related to rain or snow; Morning and evening peak impact value X 12 The ratio of the number of the sixth traffic accidents counted by the traffic unit in the first time period before the corresponding designated time period to the total number of all traffic accidents counted by the traffic unit in the first time period before the corresponding designated time period; the sixth traffic accident is a traffic accident unrelated to the morning and evening rush hours; Holiday impact value X 13 is the ratio of the number of the seventh traffic accidents counted in the first time period before the corresponding designated time period of the traffic unit to the total number of all traffic accidents counted in the first time period before the corresponding designated time period of the traffic unit; The seventh traffic accident is a traffic accident unrelated to holidays; Building environmental impact value X 14 It is the ratio of the number of the seventh traffic accidents counted in the first time period before the corresponding designated time period of the traffic unit to the total number of all traffic accidents counted in the first time period before the corresponding designated time period of the traffic unit; the seventh traffic accident is a traffic accident in which the location of occurrence is more than a preset distance from a designated building; the designated buildings include schools, hospitals, and supermarkets.

5. The traffic hazard identification method based on multi-source data fusion according to claim 4 is characterized in that: The S200 specifically includes: According to the traffic influencing factors in the first specified time period corresponding to the traffic unit, the hidden danger score in the first specified time period of the traffic unit is obtained using formula (4); The formula (4) is: Da=w a ×(q1+...q i ...+q 14 ); Among them, Da is the hidden danger score of the ath traffic unit in the urban traffic network in the first specified time period; w a is the weight value of the importance corresponding to the ath traffic unit in the urban traffic network; q i X is the factor affecting traffic i Corresponding normalized value; i∈(1, 2, ..., 14).

6. The traffic hazard identification method based on multi-source data fusion according to claim 5 is characterized in that: Traffic influencing factors X i The corresponding normalized value q i It is calculated by formula (5); The formula (5) is: X imin It is the hidden danger influencing factor X in the second time period before the first specified time period of the traffic unit i The minimum value of X imax It is the hidden danger influencing factor X in the second time period before the first specified time period of the traffic unit i The maximum value of .

7. The method for identifying traffic hazards by multi-source data fusion according to claim 5 is characterized in that: The weight value w of the importance corresponding to the a-th traffic unit in the urban traffic network a It is calculated by formula (6); The formula (6) is: Where n is the number of transportation units in the urban transportation network; and a =b1F a +b2G a +b3A a ; e a is the first parameter value corresponding to the ath traffic unit in the urban traffic network; F a is the total number of vehicles passing through the ath traffic unit in the urban traffic network in a preset historical time period; G a is the road congestion index corresponding to the ath traffic unit in the urban traffic network counted in advance within a preset historical time period; G a =1-(average speed of vehicles passing through the traffic unit in the historical time period / maximum allowed speed corresponding to the traffic unit); A a It is the average waiting time of vehicles in a pre-set historical time period corresponding to the ath traffic unit in the urban traffic network when passing through the traffic unit under red light or congestion conditions.

8. The method for identifying traffic hazards by multi-source data fusion according to claim 7 is characterized in that: Prior to S100, it also included: S01. Obtaining traffic influencing factors within a second specified time period corresponding to any traffic unit in the urban traffic network; The second specified time period is before the first specified time period; S02. Obtaining a hidden danger score for the second specified time period of the traffic unit according to the traffic influencing factors for the second specified time period corresponding to the traffic unit; S03. Obtain multiple hidden danger score intervals according to the hidden danger score of the second specified time period corresponding to each traffic unit in the urban traffic network.

9. The method for identifying traffic hazards by multi-source data fusion according to claim 4 is characterized in that: The S03 specifically includes: S03-1. Obtain an initial sample set D according to the hidden danger score of each traffic unit in the urban traffic network in the second specified time period; D={D1, D2, D3,..., Dn}; Where Dn is the hidden danger score of the nth traffic unit in the urban traffic network in the second specified time period; S03-2, randomly selecting t traffic units' hidden danger scores within the second specified time period from the initial sample set D and duplicating them respectively, and putting the duplicated hidden danger scores into the initial sample set D to obtain a new training set; The new training set includes the hidden danger scores of n+t traffic units in the second specified time period; S03-3, performing step S03-2 r times, and obtaining r new training sets respectively; Where r = h - 1; S03-4, combining the new training set obtained in step S03-2 and the r new training sets obtained in step S03-3 to obtain a sample set group; S03-5. Based on the sample set, multiple hidden danger score intervals are obtained; Wherein, the S03-5 specifically includes: Using a clustering algorithm to divide all hidden danger scores in the sample set into k groups; For each group's hidden danger score, obtain the hidden danger score interval corresponding to the group; Among them, the hidden danger score interval corresponding to each group is from the minimum value of the hidden danger score of the group to the maximum value of the hidden danger score of the group.

10. A multi-source data fusion traffic hazard identification system, characterized in that: include: at least one processor; as well as At least one memory is communicatively connected to the processor, wherein the memory stores program instructions executable by the processor, and the processor calls the program instructions to execute the traffic hazard identification method based on multi-source data fusion as described in any one of claims 1-9.

Images