Method, system, equipment and medium for calculating unequal red speed of real-time road section based on fusion of license plate snapshot by electronic police and real-time signal control data

Abstract

The invention relates to a method, a system, equipment and a medium for calculating a real-time road section unequal red speed based on fusion of a license plate captured by an electronic police and real-time signal control data. The method comprises the following steps: receiving vehicle passing data and traffic light state information from upstream and downstream electronic police checkpoints; according to the traffic light state information, performing time restoration on the capturing moments of each piece of vehicle passing data at the upstream and downstream electric police checkpoints to obtain initial travel time; performing correction according to the initial travel time, the free flow travel time and the downstream intersection period to obtain pure travel time; all the pure travel time is subjected to clustering analysis and divided into unequal red light samples and equal red light samples, the average travel speed of each sample is calculated, and the unequal red light travel speed of the real-time road section is obtained by introducing correction factors. According to the method, the license plate captured by the electronic police and the real-time signal control data are fused, so that the real-time unequal-red-light speed of the road section is accurately calculated, and the method has a remarkable practical application value.

Description

Technical Field

[0001] The present invention relates to the field of intelligent transportation technology, and in particular to a method, system, device and medium for calculating real-time road section unequal red speed based on the fusion of license plate captured by electronic police and real-time signal control data. Background Art

[0002] In the field of traffic management, real-time monitoring and analysis of road speed is of great significance for improving road traffic efficiency and optimizing signal control strategies. At present, there are many technical solutions for identifying the speed of vehicles on a road. These solutions usually rely on the electronic police system to capture license plate information to obtain the timestamp of the vehicle passing a specific point, and then calculate the speed. However, these solutions often face some limitations in practical applications.

[0003] Since existing technologies mainly focus on calculating vehicle speed directly through license plate capture data, they usually do not fully consider the impact of traffic light signal cycle on vehicle speed calculation. On urban roads, traffic lights are key elements in controlling traffic flow, and their signal cycle directly affects the driving status of vehicles, especially during red light periods, when vehicles are forced to stop and wait, resulting in a large amount of non-driving time in the vehicle speed calculation result, which cannot accurately reflect the actual driving speed of the road section.

[0004] In addition, existing solutions often ignore the interference of upstream and downstream waiting areas on vehicle speed calculation. As a measure to ease traffic pressure and improve the traffic capacity of intersections, waiting areas are set up in front of traffic lights to wait for vehicles. When a vehicle enters the waiting area and waits for passage, it is not actually driving on the road section, but the license plate capture system may count this time as driving time, resulting in a higher speed calculation, which cannot truly reflect the continuous driving speed of the vehicle on the road section.

[0005] Therefore, the existing vehicle speed recognition scheme based on the license plate information captured by the electronic police does not take into account the traffic light signal cycle and does not effectively eliminate the influence of the upstream and downstream waiting areas, resulting in a large deviation in the calculated vehicle speed data and failing to accurately reflect the actual driving conditions of the road section. This not only affects the traffic management department's accurate assessment of road traffic efficiency, but also restricts the optimization and adjustment of signal control strategies. Summary of the invention

[0006] 1. Technical issues to be resolved

[0007] In view of the above-mentioned shortcomings and deficiencies of the prior art, the present invention provides a method, system, device and medium for estimating real-time unequal red light speed on road sections based on the fusion of license plate captured by electronic police and real-time signal control data, which solves the technical problem that when calculating the speed of road sections in the prior art, the speed calculation result is inaccurate because the traffic light signal cycle is not integrated and the influence of the upstream and downstream waiting areas is not excluded.

[0008] (II) Technical solution

[0009] In order to achieve the above object, the main technical solutions adopted by the present invention include:

[0010] In a first aspect, an embodiment of the present invention provides a method for calculating the real-time road section unequal red light speed based on the fusion of license plate captured by electronic police and real-time signal control data, including:

[0011] Receive vehicle passing data and traffic light status information from upstream and downstream electronic police checkpoints;

[0012] According to the traffic light status information, the time of capture of each vehicle passing data at the upstream and downstream electric police checkpoints is repaired to obtain the preliminary travel time of each vehicle not affected by the red light;

[0013] Based on the preliminary travel time, correction is performed in combination with the preset free flow travel time and the downstream intersection cycle to obtain a pure travel time that is not affected by red lights and cycles;

[0014] A cluster analysis is performed on the pure travel time of all vehicles, and the samples are divided into samples without waiting for red lights and samples waiting for red lights. The average travel speed of each type of samples is calculated respectively, and the real-time travel speed without waiting for red lights is obtained by introducing correction factors.

[0015] Optionally, receiving vehicle passing data and traffic light status information from upstream and downstream electric police checkpoints includes:

[0016] Receive upstream vehicle passing data from the upstream electric police checkpoint, which includes the license plate number and the time when the upstream electric police captured the vehicle;

[0017] Receive downstream vehicle passing data from the downstream electric police checkpoint, the downstream vehicle passing data includes the license plate number and the time when the downstream electric police captured it;

[0018] Receive traffic light status information synchronized with upstream and downstream vehicle passing data, including the traffic light interval status of the upstream and downstream intersections at the corresponding capture time, and the actual earliest release time of each intersection;

[0019] The received upstream vehicle passing data and downstream vehicle passing data are matched to ensure that each pair of upstream and downstream data corresponds to the same license plate number.

[0020] Optionally, according to the traffic light status information, the capture time of each vehicle passing data at the upstream and downstream electric police checkpoints is time-repaired to obtain the preliminary travel time of each vehicle not affected by the red light, including:

[0021] For each vehicle, the color of the traffic light zone to which it belongs is determined at the moment when the electric police at the upstream intersection captures the vehicle;

[0022] If the traffic light interval at the upstream intersection where the electric alarm is captured is red, the earliest actual release time recorded at the upstream intersection is the start time of the next green light;

[0023] If the traffic light interval to which the electric police captures the traffic at the upstream intersection belongs is green, the electric police capture time at the upstream intersection is directly used as the earliest actual release time at the upstream intersection;

[0024] For each vehicle, the color of the traffic light zone to which it belongs is determined at the moment when the electric police at the downstream intersection captures the vehicle;

[0025] If the color of the traffic light interval to which the electric alarm captures the downstream intersection belongs is red, the actual earliest release time recorded at the downstream intersection is the start time of the next green light;

[0026] If the color of the traffic light section to which the electric police capture time at the downstream intersection belongs is green, the electric police capture time at the downstream intersection is directly used as the actual earliest release time of the downstream intersection;

[0027] The preliminary travel time of each vehicle without being affected by red lights is calculated based on the difference between the actual earliest release time of the upstream intersection and the actual earliest release time of the downstream intersection.

[0028] Optionally, the preliminary travel time is corrected in combination with the preset free flow travel time and the downstream intersection cycle to obtain a pure travel time that is not affected by red lights and cycles, including:

[0029] Determine a free-flow travel time and downstream intersection cycle;

[0030] For each pair of vehicles, determine whether the preliminary travel time minus the free-flow travel time is greater than n times the downstream intersection cycle;

[0031] If the initial travel time minus the free flow travel time is greater than n times the downstream intersection cycle, it means that the vehicle has waited at least n traffic light cycles at the downstream intersection. The initial actual travel time minus n times the downstream intersection cycle is subtracted to obtain the corrected pure travel time;

[0032] If the preliminary travel time minus the free flow travel time is not greater than n times the downstream intersection cycle, the preliminary actual travel time is directly used as the corrected pure travel time;

[0033] Among them, the free-flow travel time is the travel time of the road section without traffic congestion and traffic lights, which is obtained based on historical traffic data or actual measurements. The downstream intersection cycle is determined based on the traffic light settings and traffic flow conditions at the downstream intersection, which represents a complete cycle time of the traffic light.

[0034] Optionally, cluster analysis is performed on the pure travel time of all vehicles, and the samples are divided into samples without waiting for red lights and samples waiting for red lights. The average travel speed of each type of samples is calculated respectively. By introducing correction factors, the travel speed of the real-time road section without waiting for red lights includes:

[0035] Perform cluster analysis on the pure travel time of all vehicles, and divide the samples into samples without waiting for red lights and samples waiting for red lights according to the extracted features;

[0036] The vehicle travel time in the non-waiting red light samples is processed, and the average travel speed of the non-waiting red light samples is obtained by combining the obtained road section length information;

[0037] Identify the correction factors that affect the travel speed and establish the corresponding correction model;

[0038] Based on the correction model, the average travel speed of the samples without waiting for red lights is corrected to obtain the real-time travel speed of the road section when it is not affected by the red light.

[0039] Optionally, cluster analysis is performed on the pure travel time of all vehicles, and the samples are divided into non-red light waiting samples and red light waiting samples according to the extracted features, including:

[0040] For the pure travel time of all vehicles, the acceleration change feature vector of each vehicle is formed by extracting the speed statistical features including the acceleration peak value, acceleration valley value and acceleration change rate, and analyzing the acceleration spectrum features through fast Fourier transform;

[0041] By extracting the intersection of the downstream intersection to capture the traffic light interval, and encoding the traffic light state into discrete feature values, a traffic light interval feature vector is formed;

[0042] The driving time of the vehicle from the upstream intersection to the downstream intersection is calculated by extraction and used as the driving time feature vector;

[0043] The acceleration change feature vector, the traffic light interval feature vector and the driving time feature vector are integrated to obtain at least one optimized feature vector;

[0044] Based on the optimized feature vector, the similarity matrix is ​​constructed using the Gaussian kernel function;

[0045] Perform Laplace transform based on the similarity matrix to obtain a Laplace matrix;

[0046] Calculate the eigenvalues ​​and eigenvectors of the Laplace matrix, and select the eigenvectors corresponding to the first k smallest eigenvalues ​​to form the feature space;

[0047] Project the optimized feature vectors of all vehicle samples into this feature space to obtain the projected sample set;

[0048] The local density of each sample is determined by calculating the distance between each sample and its nearest neighbor sample in the sample set, and the k samples with the highest density are selected as the initial clustering centers;

[0049] Assign each sample to the nearest initial cluster center to obtain multiple clusters;

[0050] Recalculate the mean of all samples in the cluster as the new cluster center, check whether the cluster center has changed or whether the change is less than the set threshold. If it is less than the set threshold, clustering ends;

[0051] According to the clustering results, each sample is divided into samples that do not wait for the red light and samples that wait for the red light.

[0052] Optionally, identifying correction factors affecting the travel speed and establishing a corresponding correction model includes:

[0053] Collect multi-dimensional historical traffic data including weather conditions, road conditions, and time period data, and identify correction factors related to travel speed;

[0054] Determine the baseline state corresponding to each correction factor;

[0055] A random forest model was built and trained to predict the relationship between each correction factor and travel speed;

[0056] The correction coefficient is obtained by comparing the predicted travel speed of the trained random forest model under different conditions of each correction factor with the predicted travel speed under the baseline condition;

[0057] Based on each correction factor and the corresponding correction coefficient, a correction model is obtained;

[0058] The modified model is:

[0059]

[0060] In the formula, is the average travel speed, x1, x2, x3 are weather conditions, road conditions and time period data, ω1, ω2, ω3 are the weights of weather conditions, road conditions and time period data, f1(x1), f2(x2), f3(x3) are the nonlinear relationship functions between weather conditions, road conditions and time period data and travel speed, The value range is [0.6-1].

[0061] In a second aspect, an embodiment of the present invention provides a system for calculating the real-time road section speed at different red lights based on the fusion of license plate captured by electronic police and real-time signal control data, including:

[0062] A signal receiving module is used to receive vehicle passing data and traffic light status information from upstream and downstream electric police checkpoints;

[0063] The module for determining the preliminary travel time is used to perform time repair on the capture time of each vehicle passing data at the upstream and downstream electric police checkpoints according to the traffic light status information, so as to obtain the preliminary travel time of each vehicle not affected by the red light;

[0064] A pure travel time determination module is used to make corrections based on the preliminary travel time in combination with a preset free flow travel time and a downstream intersection cycle to obtain a pure travel time that is not affected by red lights and cycles;

[0065] The travel speed output module without waiting for red lights is used to perform cluster analysis on the pure travel time of all vehicles, divide the samples into samples without waiting for red lights and samples waiting for red lights, and calculate the average travel speed of each type of samples respectively. By introducing correction factors, the real-time travel speed of the road section without waiting for red lights is obtained.

[0066] In a third aspect, an embodiment of the present invention provides a device for estimating real-time unequal red light speeds on road sections based on the fusion of license plates captured by electronic police and real-time signal control data, comprising: at least one database; and a memory communicatively connected to at least one database; wherein the memory stores instructions executable by at least one database, and the instructions are executed by at least one database so that at least one database can execute the method for estimating real-time unequal red light speeds on road sections based on the fusion of license plates captured by electronic police and real-time signal control data as described above.

[0067] In a fourth aspect, an embodiment of the present invention provides a computer-readable medium having computer executable instructions stored thereon. When the executable instructions are executed by a processor, the method of calculating the real-time unequal red light speed on a road section based on the fusion of license plate captured by the electronic police and real-time signal control data as described above is implemented.

[0068] (III) Beneficial effects

[0069] The beneficial effects of the present invention are:

[0070] First, by receiving vehicle passing data and traffic light status information from upstream and downstream electric police checkpoints, comprehensive data collection is achieved, providing a solid foundation for subsequent analysis.

[0071] Next, the traffic light status information is used to perform time repair on the passing vehicle data. This step effectively eliminates the interference of the red light on the vehicle travel time, making the initially obtained travel time more accurate and reliable. Furthermore, the present invention also combines the preset free flow travel time and the downstream intersection cycle for correction. This correction process not only considers the speed of the vehicle under normal driving conditions, but also fully considers the impact of the intersection signal cycle on the vehicle travel, thereby obtaining a pure travel time that is closer to reality and is not affected by red lights and cycles. This processing makes the calculation of travel time more accurate and provides strong support for subsequent cluster analysis and speed calculation.

[0072] On this basis, cluster analysis was performed on the pure travel time of all vehicles, and the samples were divided into samples without waiting for red lights and samples waiting for red lights. This classification method not only improved the accuracy of data analysis, but also made the subsequent speed calculation more targeted. Subsequently, the average travel speed of each type of sample was calculated, and correction factors were introduced to finally obtain the travel speed of the real-time road section without waiting for red lights.

[0073] Therefore, the present invention realizes the accurate calculation of the real-time speed of different red lights on road sections by integrating the license plate captured by electronic police and the real-time signal control data, providing powerful data support for the fields of traffic management, road condition monitoring and travel planning, and has significant practical application value and social benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 A flow chart of a method for calculating the real-time road section speed at different red lights based on the fusion of license plate captured by electronic police and real-time signal control data provided by an embodiment of the present invention;

[0075] Figure 2 A specific flow chart of step S1 of a method for estimating real-time road section unequal red light speed based on the fusion of license plate captured by electronic police and real-time signal control data provided by an embodiment of the present invention;

[0076] Figure 3 A schematic diagram of a sample trip of a method for calculating the real-time road section speed at different red lights based on the fusion of license plate captured by electronic police and real-time signal control data provided by an embodiment of the present invention;

[0077] Figure 4 A schematic diagram of sample travel calculation data of a method for estimating real-time road section unequal red light speed based on the fusion of license plate captured by electronic police and real-time signal control data provided by an embodiment of the present invention;

[0078] Figure 5 A specific flow chart of step S2 of a method for estimating real-time road section unequal red light speed based on the fusion of license plate captured by electronic police and real-time signal control data provided by an embodiment of the present invention;

[0079] Figure 6 A specific flow chart of step S3 of a method for estimating real-time road section unequal red light speed based on the fusion of license plate captured by electronic police and real-time signal control data provided by an embodiment of the present invention;

[0080] Figure 7 A specific flow chart of step S4 of a method for estimating real-time road section unequal red light speed based on the fusion of license plate captured by electronic police and real-time signal control data provided by an embodiment of the present invention;

[0081] Figure 8 A schematic diagram of trip clustering of a method for calculating real-time road section speeds at different red lights based on the fusion of license plate captured by electronic police and real-time signal control data provided by an embodiment of the present invention;

[0082] Fig. 9 A specific flow chart of step S41 of a method for estimating real-time road section unequal red light speed based on the fusion of license plate captured by electronic police and real-time signal control data provided by an embodiment of the present invention;

[0083] Fig.10 A Gantt chart of samples of unequal red lights based on a method for estimating real-time road section unequal red light speeds by fusing license plate captured by electronic police with real-time signal control data provided by an embodiment of the present invention;

[0084] Fig.11 A specific flow chart of step S43 of a method for calculating real-time unequal red light speed on a road section based on the fusion of license plate captured by electronic police and real-time signal control data provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0085] 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.

[0086] like Figure 1 As shown, an embodiment of the present invention proposes a method for estimating the real-time road section unequal red light speed based on the fusion of license plate captured by electronic police and real-time signal control data, comprising: receiving vehicle passing data and traffic light status information from upstream and downstream electronic police checkpoints; performing time repair on the capture moment of each vehicle passing data at the upstream and downstream electronic police checkpoints according to the traffic light status information, to obtain a preliminary travel time of each vehicle not affected by red lights; performing correction based on the preliminary travel time and in combination with a preset free flow travel time and a downstream intersection cycle, to obtain a pure travel time not affected by red lights and cycles; performing cluster analysis on the pure travel time of all vehicles, dividing the samples into unequal red light samples and equal red light samples, and calculating the average travel speed of each type of sample respectively, and obtaining the real-time road section unequal red light travel speed by introducing correction factors.

[0087] 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.

[0088] Specifically, an embodiment of the present invention provides a method for calculating the real-time road section unequal red light speed based on the fusion of license plate captured by electronic police and real-time signal control data, which includes:

[0089] S1. Receive vehicle passing data and traffic light status information from upstream and downstream electric police checkpoints.

[0090] Furthermore, if Figure 2 As shown, step S1 includes:

[0091] S11. Receive upstream vehicle passing data from the upstream electric police checkpoint, where the upstream vehicle passing data includes the license plate number and the upstream electric police capture time.

[0092] S12, receiving downstream vehicle passing data from the downstream electric police checkpoint, the downstream vehicle passing data including the license plate number and the downstream electric police capture time.

[0093] S13, receiving traffic light status information synchronized with upstream vehicle passing data and downstream vehicle passing data, including the traffic light interval status of the upstream intersection and the downstream intersection at the corresponding capture time, and the actual earliest release time of each intersection.

[0094] S14. Match the received upstream vehicle passing data and downstream vehicle passing data to ensure that each pair of upstream and downstream data corresponds to the same license plate number.

[0095] In a specific embodiment, reference Figure 3 and Figure 4 , collect data on upstream and downstream vehicle pairs, including license plate number, electric police capture time at the upstream intersection (t1, t2, t4, t5), traffic light section (red / green) captured at the upstream intersection, actual earliest release time at the upstream intersection (t3, t4, t5), electric police capture time at the downstream intersection (t1', t2', t4', t5'), traffic light section (red / green) captured at the downstream intersection, actual earliest release time at the downstream intersection (t6, t2', t4', t7).

[0096] Next, match the upstream and downstream vehicle data according to the license plate numbers, and output the matching results: license plate 1: (t1, t1'), license plate 2: (t2, t2'), license plate 3: (t4, t4'), license plate 4: (t5, t5'). After that, for each pair of matching data, attach the corresponding traffic light status information and the actual earliest release time. Output the processed data: license plate 1: (t1, red, t3, t1', red, t6), license plate 2: (t2, red, t3, t2', green, t2'), license plate 3: (t4, green, t4, t4', green, t4'), license plate 4: (t5, green, t5, t5', red, t7).

[0097] S2. Based on the traffic light status information, the capture time of each vehicle passing data at the upstream and downstream electric police checkpoints is repaired to obtain the preliminary travel time of each vehicle not affected by the red light.

[0098] Furthermore, if Figure 4 and Figure 5 As shown, step S2 includes:

[0099] S21. For each vehicle, at the moment of being captured by the electric police at the upstream intersection, determine the color of the traffic light zone to which it belongs.

[0100] S22. If the traffic light interval to which the electric alarm captures the upstream intersection belongs is red, then the earliest actual release time of the upstream intersection is recorded as the start time of the next green light, recorded as t 上游_最早放行 At the same time, if the traffic light interval of the upstream intersection where the electric police captures the traffic light is green, the upstream intersection electric police capture time is directly used as the actual earliest release time of the upstream intersection, t 上游_最早放行 =t 上游 .

[0101] S23. For each vehicle, at the moment when the electric police at the downstream intersection captures the vehicle, determine the color of the traffic light zone to which it belongs.

[0102] S24. If the color of the traffic light interval to which the electric alarm captures the downstream intersection belongs is red, then the actual earliest release time of the downstream intersection is recorded as the start time of the next green light, recorded as t 下游_最早放行 If the color of the traffic light interval to which the electric police at the downstream intersection captures the time is green, the electric police capture time at the downstream intersection is directly used as the actual earliest release time of the downstream intersection, that is, t 下游_最早放行 =t 下游 .

[0103] S25. Calculate the initial travel time of each vehicle without being affected by the red light based on the difference between the actual earliest release time of the upstream intersection and the actual earliest release time of the downstream intersection, t 差 =t 下游_最早放行 -t 上游_最早放行 .

[0104] S3. Correct the preliminary travel time based on the preset free flow travel time and the downstream intersection cycle to obtain a pure travel time that is not affected by red lights and cycles.

[0105] Furthermore, if Figure 6 As shown, step S3 includes:

[0106] S31. Determine a free flow travel time and a downstream intersection cycle.

[0107] It should be noted that the free flow travel time t0 is obtained based on historical traffic data or actual measurements. It is the travel time of a road section without traffic congestion and traffic lights. The downstream intersection cycle t 10 It is determined based on the traffic light settings and traffic flow conditions at the downstream intersection, and represents a complete cycle time of the traffic light.

[0108] S32. For each pair of vehicles, determine whether the initial travel time minus the free flow travel time is greater than n times the downstream intersection cycle.

[0109] S33, if the initial travel time minus the free flow travel time is greater than n times the downstream intersection cycle, it means that the vehicle has waited at least n traffic light cycles at the downstream intersection, and the initial actual travel time minus n times the downstream intersection cycle is subtracted to obtain the corrected pure travel time. In this case, the corrected pure travel time t 纯净 is: t 纯净 =t 差 -n*t10. This means that the additional time due to waiting at traffic lights is subtracted from the preliminary travel time.

[0110] S34, if the initial travel time minus the free flow travel time is not greater than n times the downstream intersection cycle, it means that the vehicle did not wait for n or more traffic light cycles at the downstream intersection, then the initial actual travel time is directly used as the corrected pure travel time. In this case, the corrected pure travel time t 纯净 Directly equal to t 差 , that is, no additional correction is performed.

[0111] S4. Perform cluster analysis on the pure travel time of all vehicles, divide the samples into samples without waiting for red lights and samples waiting for red lights, and calculate the average travel speed of each type of samples respectively. By introducing correction factors, the real-time travel speed of the road section without waiting for red lights is obtained.

[0112] Furthermore, if Figure 7 As shown, step S4 includes:

[0113] S41. Perform cluster analysis on the pure travel time of all vehicles, and divide the samples into samples of not waiting for red lights and samples of waiting for red lights according to the extracted features.

[0114] refer to Figure 8 , cluster analysis was performed on the pure travel time data of all vehicles. Through continuous iteration and optimization of the algorithm, three clusters were finally obtained. According to the clustering results and feature analysis, these three clusters can be divided into two types of samples:

[0115] Cluster 1: Identified as non-waiting red light samples. The vehicle travel time data in this type of sample shows obvious characteristics of not waiting for red lights, such as relatively stable driving speed, fewer stops and less parking time.

[0116] Cluster 2 and Cluster 3: Both are identified as samples waiting for red lights. This shows that the vehicle trips in these two types of samples are significantly affected by red lights. However, when further distinguishing the two clusters, the red light waiting period represented by Cluster 3 is longer than that of Cluster 2. This means that the vehicles in Cluster 3 have longer stop waiting times due to longer red light cycles or more red lights.

[0117] Furthermore, if Fig. 9 As shown, step S41 includes:

[0118] S411. For the pure travel time of all vehicles, the acceleration change feature vector of each vehicle is formed by extracting the speed statistical features including the acceleration peak value, the acceleration valley value and the acceleration change rate, and analyzing the acceleration spectrum features by fast Fourier transform; the traffic light interval belonging to the intersection of the downstream intersection is captured by extracting the intersection, and the traffic light state is encoded into discrete feature values ​​to form a traffic light interval feature vector; the driving time of the vehicle from the upstream intersection to the downstream intersection is calculated by extraction and used as the driving time feature vector.

[0119] Specifically, the acceleration change sequence of each vehicle is calculated, and key feature points are extracted, such as acceleration peak value, acceleration valley value, acceleration change rate, etc. The acceleration change sequence is converted into a feature vector using methods such as discretization and Fourier transform for subsequent clustering analysis.

[0120] Next, the traffic light intervals are captured according to the intersections of the upstream and downstream intersections, and the traffic light interval features are constructed. Binary encoding or one-hot encoding can be used to represent the traffic light intervals, such as 1 for red light and 0 for green light, or other encoding methods can be used.

[0121] For travel time, the travel time and the earliest actual release time of the downstream intersection are directly used as features. These features can be standardized, such as using Z-score standardization or Min-Max standardization, to ensure that they have the same weight in the clustering algorithm.

[0122] S412: Fusing the acceleration change feature vector, the traffic light interval feature vector, and the driving time feature vector to obtain at least one optimized feature vector.

[0123] S413, based on the optimized feature vector, a similarity matrix is ​​constructed using a Gaussian kernel function, wherein the Gaussian kernel function formula is:

[0124]

[0125] In the formula, x i and x j are two data points and σ is the width parameter of the Gaussian kernel.

[0126] S414, perform Laplace transformation based on the similarity matrix to obtain a Laplace matrix. Wherein, construct a degree matrix D, D = ∑ j S ij , Laplace transform the matrix D into L = ID -1 S. I is the identity matrix.

[0127] S415, calculating the eigenvalues ​​and eigenvectors of the Laplacian matrix, and selecting the eigenvectors corresponding to the first k smallest eigenvalues ​​to form a feature space.

[0128] S416, projecting the optimized feature vectors of all vehicle samples into the feature space to obtain a projected sample set. Through projection, the original high-dimensional data can be reduced to a low-dimensional feature space while retaining as much information as possible. In this way, a projected sample set is obtained, which has better separability in the feature space.

[0129] S417, determining the local density of each sample by calculating the distance between each sample in the sample set and the nearest neighbor sample, and selecting k samples with the highest density as initial clustering centers.

[0130] S418. Assign each sample to the nearest initial cluster center to obtain multiple clusters.

[0131] S419, recalculate the mean of all samples in the cluster as the new cluster center, check whether the cluster center has changed or whether the change is less than the set threshold. If it is less than the set threshold, the clustering ends; according to the clustering results, divide each sample into unequal red light samples and equal red light samples.

[0132] Recalculate the mean of all samples in each cluster as the new cluster center. The mean is a good statistic that can reflect the average characteristics of samples in a cluster. Next, check whether the cluster center has changed or whether the change is less than the set threshold. If the change of the cluster center is less than the set threshold, it means that the clustering result is relatively stable and the clustering process can be ended. Otherwise, it is necessary to continue to iteratively update the cluster center until the stopping condition is met. Finally, these samples are clearly divided into two categories: samples that do not wait for red lights and samples that wait for red lights.

[0133] S42: Process the vehicle travel time in the non-waiting red light samples, and calculate the average travel speed of the non-waiting red light samples in combination with the acquired road section length information.

[0134] like Fig.10 The project Gantt chart shown intuitively shows the changing trend of individual unequal red light samples over time. In this Gantt chart, the acquired road section length information is combined. The road section length is a key parameter necessary for calculating the speed, which reflects the actual distance traveled by the vehicle. By combining the travel time of each vehicle with the corresponding road section length, the travel speed of each vehicle on a specific road section can be calculated. Finally, the vehicle travel speeds in all unequal red light samples were statistically analyzed and the average travel speed of these samples was obtained.

[0135] S43, identifying the correction factors that affect the travel speed and establishing a corresponding correction model.

[0136] Furthermore, if Fig.11 As shown, step S43 includes:

[0137] S431. Collect multi-dimensional historical traffic data including weather conditions, road conditions, and time period data, and identify correction factors related to travel speed.

[0138] S432: Determine the reference state corresponding to each correction factor.

[0139] S433. Establish and train a random forest model for predicting the relationship between each correction factor and travel speed.

[0140] S434. Obtain a correction coefficient by comparing the predicted travel speed of the trained random forest model under different states of each correction factor with the predicted travel speed under the reference state.

[0141] S435. Obtain a correction model based on each correction factor and the corresponding correction coefficient.

[0142] The modified model is:

[0143]

[0144] In the formula, is the average travel speed, x1, x2, x3 are weather conditions, road conditions and time period data, ω1, ω2, ω3 are weights of weather conditions, road conditions and time period data, which can be determined by historical data regression analysis or expert experience, and f1(x1), f2(x2), f3(x3) are nonlinear relationship functions between weather conditions, road conditions and time period data and travel speed.

[0145] For weather conditions, it is generally believed that high temperature will reduce travel speed, so a negatively correlated exponential function is chosen: (Assuming temperatures are in degrees Celsius).

[0146] For road conditions, the higher the congestion level, the lower the travel speed, so a negatively correlated linear function or quadratic function can be selected: f2(x2)=-0.5x2 (assuming that the congestion level is represented by some quantitative indicator).

[0147] For the time period, the travel speed will decrease during the morning and evening peak hours, so a piecewise function can be selected to represent it: f3(x3)=-0.2 (if it is the morning and evening peak hours), and the function takes 0 during non-peak hours.

[0148] thus, The value range is [0.6-1].

[0149] S44. Based on the correction model, the average travel speed of the samples without waiting for red lights is corrected to obtain the travel speed of the real-time road section when it is not affected by the red light.

[0150] In a specific embodiment, the average travel speed of the samples without waiting for red lights is The key factor state set obtained in real time is S = {s1, s2, ..., s n}, the corresponding correction coefficient set is C = {C 11 ,C 22 ,…,C nn}, the corrected average travel speed can be expressed as:

[0151]

[0152] In the formula, It is the product of the key factor correction coefficients.

[0153] In addition, an embodiment of the present invention provides a system for calculating the real-time road section speed difference when waiting for red lights based on the fusion of license plate captured by electronic police and real-time signal control data, including:

[0154] The signal receiving module is used to receive vehicle passing data and traffic light status information from upstream and downstream electric police checkpoints.

[0155] The preliminary travel time determination module is used to perform time repair on the capture moment of each vehicle passing data at the upstream and downstream electric police checkpoints according to the traffic light status information, so as to obtain the preliminary travel time of each vehicle not affected by the red light.

[0156] A pure travel time determination module is used to make corrections based on the preliminary travel time in combination with a preset free flow travel time and a downstream intersection cycle to obtain a pure travel time that is not affected by red lights and cycles;

[0157] The travel speed output module without waiting for red lights is used to perform cluster analysis on the pure travel time of all vehicles, divide the samples into samples without waiting for red lights and samples waiting for red lights, and calculate the average travel speed of each type of samples respectively. By introducing correction factors, the real-time travel speed of the road section without waiting for red lights is obtained.

[0158] Furthermore, an embodiment of the present invention provides a device for estimating real-time unequal red light speeds on road sections based on the fusion of license plates captured by electronic police and real-time signal control data, comprising: at least one database; and a memory communicatively connected to at least one database; wherein the memory stores instructions executable by at least one database, and the instructions are executed by at least one database so that at least one database can execute the method for estimating real-time unequal red light speeds on road sections based on the fusion of license plates captured by electronic police and real-time signal control data as described above.

[0159] First, the device is equipped with at least one database. As the core of data storage and management, this database is responsible for receiving, storing and managing key data such as license plate data captured by electronic police and traffic light status information provided by real-time signal control solutions. The database's powerful storage capacity and efficient data processing capabilities provide a solid foundation for subsequent data fusion and speed estimation.

[0160] Secondly, the device establishes a communication connection with the at least one database and is equipped with a corresponding memory. The memory, as a carrier of instructions, stores specific instruction sets that can be executed by the database. These instruction sets are the core of the device to realize its functions. They are carefully designed to ensure that the database can efficiently execute the method of calculating the real-time red light speed of the road section based on the fusion of the license plate captured by the electronic police and the real-time signal control data according to the predetermined logic and algorithm.

[0161] When the instructions are executed by the database, the device begins to exert its powerful data processing and analysis capabilities. It first uses the license plate data captured by the electronic police, combined with the traffic light status information of the real-time signal control solution, to accurately repair and correct the vehicle's driving data. Then, through advanced data processing techniques such as cluster analysis, the processed data is divided into samples of non-waiting red lights and samples of waiting for red lights, and the average travel speed of each type of sample is calculated separately. Finally, by introducing correction factors, the device can accurately calculate the non-waiting red light speed of the real-time road section.

[0162] The device provided by the embodiment of the present invention, which is based on the fusion of electronic police-captured license plates and real-time signal control data to estimate real-time unequal red light speeds on road sections, integrates an efficient database and a carefully designed instruction set to achieve accurate monitoring and analysis of real-time road section traffic conditions, and provides strong technical support for fields such as traffic management, road condition prediction, and travel planning.

[0163] Furthermore, an embodiment of the present invention provides a computer-readable medium having computer executable instructions stored thereon. When the executable instructions are executed by a processor, the method of calculating the real-time unequal red light speed on a road section based on the fusion of license plate captured by electronic police and real-time signal control data as described above is implemented.

[0164] The embodiment of the present invention also covers an innovative computer-readable medium that carries computer-executable instructions and has high technical practicality and application value. Specifically, when the instructions stored on the computer-readable medium are recognized and executed by the processor, an advanced function can be fully realized - based on the fusion of the license plate captured by the electronic police and the real-time signal control data, the real-time road section's different red light speed can be accurately calculated.

[0165] This computer-readable medium may be a storage device in various forms such as a hard disk, flash drive, optical disk, SD card, etc. Its core value lies in the stored instruction set. These instruction sets are carefully designed and contain complex data processing logic and algorithms, which can guide the processor to efficiently process the license plate capture data from the electronic police and the traffic light status information provided by the real-time signal control solution.

[0166] When the processor executes these instructions, it will follow the steps defined by the instruction set to gradually complete a series of operations such as data reception, time repair, correction, cluster analysis, and speed estimation. Ultimately, the processor can output the different red light speeds of real-time road sections, providing accurate and timely data support for multiple fields such as traffic management, road condition monitoring, and travel planning.

[0167] Therefore, the computer-readable medium provided by the embodiment of the present invention not only enriches the application field of computer technology, but also injects new vitality into the development of intelligent transportation systems. Through this medium, the traffic conditions of the road section can be more accurately grasped, providing a strong technical guarantee for optimizing traffic management and improving travel efficiency.

[0168] Since the system / device described in the above embodiments of the present invention is a system / device used to implement the method of the above embodiments of the present invention, a person skilled in the art can understand the specific structure and deformation of the system / device based on the method described in the above embodiments of the present invention, and thus will not be described in detail here. All systems / devices used in the method of the above embodiments of the present invention belong to the scope of protection of the present invention.

[0169] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0170] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present invention. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions.

[0171] It should be noted that in the claims, any reference numerals placed between brackets shall not be construed as limiting the claims. The word "comprising" does not exclude the presence of components or steps not listed in the claims. The word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The invention may be implemented by means of hardware comprising several different components and by means of a suitably programmed computer. In the claims enumerating several means, several of these means may be embodied by the same hardware. The use of the words first, second, third, etc., is for convenience of expression only and does not indicate any order. These words may be understood as part of the component name.

[0172] In addition, it should be noted that, 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.

[0173] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments after knowing the basic creative concept. Therefore, the claims should be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present invention.

[0174] Obviously, those skilled in the art can make various modifications and variations 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 equivalents, the present invention should also include these modifications and variations.

Claims

1. A method for calculating the real-time speed of different sections of roads at different red lights based on the fusion of license plate captured by electronic police and real-time signal control data, characterized in that: include: Receive vehicle passing data and traffic light status information from upstream and downstream electronic police checkpoints; According to the traffic light status information, the time of capture of each vehicle passing data at the upstream and downstream electric police checkpoints is repaired to obtain the preliminary travel time of each vehicle not affected by the red light; Based on the preliminary travel time, correction is performed in combination with the preset free flow travel time and the downstream intersection cycle to obtain a pure travel time that is not affected by red lights and cycles; A cluster analysis is performed on the pure travel time of all vehicles, and the samples are divided into samples without waiting for red lights and samples waiting for red lights. The average travel speed of each type of samples is calculated respectively, and the real-time travel speed without waiting for red lights is obtained by introducing correction factors.

2. The method for calculating the real-time road section speed at different red lights based on the fusion of the license plate captured by the electronic police and the real-time signal control data as claimed in claim 1, characterized in that: Receive vehicle passing data and traffic light status information from upstream and downstream electronic police checkpoints, including: Receive upstream vehicle passing data from the upstream electric police checkpoint, which includes the license plate number and the time when the upstream electric police captured the vehicle; Receive downstream vehicle passing data from the downstream electric police checkpoint, the downstream vehicle passing data includes the license plate number and the time when the downstream electric police captured it; Receive traffic light status information synchronized with upstream and downstream vehicle passing data, including the traffic light interval status of the upstream and downstream intersections at the corresponding capture time, and the actual earliest release time of each intersection; The received upstream vehicle passing data and downstream vehicle passing data are matched to ensure that each pair of upstream and downstream data corresponds to the same license plate number.

3. The method for calculating the real-time road section speed at different red lights based on the fusion of license plate captured by electronic police and real-time signal control data as claimed in claim 1, characterized in that: According to the traffic light status information, the time of capture of each vehicle passing data at the upstream and downstream electric police checkpoints is repaired, and the preliminary travel time of each vehicle not affected by the red light is obtained, including: For each vehicle, the color of the traffic light zone to which it belongs is determined at the moment when the electric police at the upstream intersection captures the vehicle; If the traffic light interval at the upstream intersection where the electric alarm is captured is red, the earliest actual release time recorded at the upstream intersection is the start time of the next green light; If the traffic light interval to which the electric police captures the traffic at the upstream intersection belongs is green, the electric police capture time at the upstream intersection is directly used as the earliest actual release time at the upstream intersection; For each vehicle, the color of the traffic light zone to which it belongs is determined at the moment when the electric police at the downstream intersection captures the vehicle; If the color of the traffic light interval to which the electric alarm captures the downstream intersection belongs is red, the actual earliest release time recorded at the downstream intersection is the start time of the next green light; If the color of the traffic light section to which the electric police capture time at the downstream intersection belongs is green, the electric police capture time at the downstream intersection is directly used as the actual earliest release time of the downstream intersection; The preliminary travel time of each vehicle without being affected by red lights is calculated based on the difference between the actual earliest release time of the upstream intersection and the actual earliest release time of the downstream intersection.

4. The method for calculating the real-time road section speed at different red lights based on the fusion of license plate captured by electronic police and real-time signal control data as claimed in claim 1, characterized in that: Based on the preliminary travel time, combined with the preset free flow travel time and the downstream intersection cycle, the pure travel time without the influence of red lights and cycles is corrected to obtain: Determine a free-flow travel time and downstream intersection cycle; For each pair of vehicles, determine whether the preliminary travel time minus the free-flow travel time is greater than n times the downstream intersection cycle; If the initial travel time minus the free flow travel time is greater than n times the downstream intersection cycle, it means that the vehicle has waited at least n traffic light cycles at the downstream intersection. The initial actual travel time minus n times the downstream intersection cycle is subtracted to obtain the corrected pure travel time; If the preliminary travel time minus the free flow travel time is not greater than n times the downstream intersection cycle, the preliminary actual travel time is directly used as the corrected pure travel time; Among them, the free-flow travel time is the travel time of the road section without traffic congestion and traffic lights, which is obtained based on historical traffic data or actual measurements. The downstream intersection cycle is determined based on the traffic light settings and traffic flow conditions at the downstream intersection, which represents a complete cycle time of the traffic light.

5. The method for calculating the real-time road section speed at different red lights based on the fusion of license plate captured by electronic police and real-time signal control data as claimed in claim 1, characterized in that: Cluster analysis is performed on the pure travel time of all vehicles, and the samples are divided into samples without waiting for red lights and samples waiting for red lights. The average travel speed of each type of samples is calculated respectively. By introducing correction factors, the travel speed of the real-time road section without waiting for red lights is obtained: Perform cluster analysis on the pure travel time of all vehicles, and divide the samples into samples without waiting for red lights and samples waiting for red lights according to the extracted features; The vehicle travel time in the non-waiting red light samples is processed, and the average travel speed of the non-waiting red light samples is obtained by combining the obtained road section length information; Identify the correction factors that affect the travel speed and establish the corresponding correction model; Based on the correction model, the average travel speed of the samples without waiting for red lights is corrected to obtain the real-time travel speed of the road section when it is not affected by the red light.

6. The method for calculating the real-time road section speed at different red lights based on the fusion of license plate captured by electronic police and real-time signal control data as claimed in claim 5, characterized in that: Cluster analysis is performed on the pure travel time of all vehicles. According to the extracted features, the samples are divided into samples that do not wait for red lights and samples that wait for red lights, including: For the pure travel time of all vehicles, the acceleration change feature vector of each vehicle is formed by extracting the speed statistical features including the acceleration peak value, acceleration valley value and acceleration change rate, and analyzing the acceleration spectrum features through fast Fourier transform; By extracting the intersection of the downstream intersection to capture the traffic light interval, and encoding the traffic light state into discrete feature values, a traffic light interval feature vector is formed; The driving time of the vehicle from the upstream intersection to the downstream intersection is calculated by extraction and used as the driving time feature vector; The acceleration change feature vector, the traffic light interval feature vector and the driving time feature vector are integrated to obtain at least one optimized feature vector; Based on the optimized feature vector, the similarity matrix is ​​constructed using the Gaussian kernel function; Perform Laplace transform based on the similarity matrix to obtain a Laplace matrix; Calculate the eigenvalues ​​and eigenvectors of the Laplace matrix, and select the eigenvectors corresponding to the first k smallest eigenvalues ​​to form the feature space; Project the optimized feature vectors of all vehicle samples into this feature space to obtain the projected sample set; The local density of each sample is determined by calculating the distance between each sample and its nearest neighbor sample in the sample set, and the k samples with the highest density are selected as the initial clustering centers; Assign each sample to the nearest initial cluster center to obtain multiple clusters; Recalculate the mean of all samples in the cluster as the new cluster center, check whether the cluster center has changed or whether the change is less than the set threshold. If it is less than the set threshold, clustering ends; According to the clustering results, each sample is divided into samples that do not wait for the red light and samples that wait for the red light.

7. The method for calculating the real-time road section speed at different red lights based on the fusion of license plate captured by electronic police and real-time signal control data as claimed in claim 5, characterized in that: Identify the correction factors that affect the travel speed and establish the corresponding correction model including: Collect multi-dimensional historical traffic data including weather conditions, road conditions, and time period data, and identify correction factors related to travel speed; Determine the baseline state corresponding to each correction factor; A random forest model was built and trained to predict the relationship between each correction factor and travel speed; The correction coefficient is obtained by comparing the predicted travel speed of the trained random forest model under different conditions of each correction factor with the predicted travel speed under the baseline condition; Based on each correction factor and the corresponding correction coefficient, a correction model is obtained; The modified model is: Where v is the average travel speed, x1, x2, x3 are the weather conditions, road conditions and time period data, ω1, ω2, ω3 are the weights of weather conditions, road conditions and time period data, f1(x1), f2(x2), f3(x3) are the nonlinear relationship functions between weather conditions, road conditions and time period data and travel speed, (1+ω1·(-e 0.1x The value range of 1)+ω2·(-0.5x2)+ω3·f3·(x3)) is [0.6-1].

8. A system for calculating the real-time speed of different sections of roads at different red lights based on the fusion of license plate captured by electronic police and real-time signal control data, characterized in that: include: A signal receiving module is used to receive vehicle passing data and traffic light status information from upstream and downstream electric police checkpoints; The module for determining the preliminary travel time is used to perform time repair on the capture time of each vehicle passing data at the upstream and downstream electric police checkpoints according to the traffic light status information, so as to obtain the preliminary travel time of each vehicle not affected by the red light; A pure travel time determination module is used to make corrections based on the preliminary travel time in combination with a preset free flow travel time and a downstream intersection cycle to obtain a pure travel time that is not affected by red lights and cycles; The travel speed output module without waiting for red lights is used to perform cluster analysis on the pure travel time of all vehicles, divide the samples into samples without waiting for red lights and samples waiting for red lights, and calculate the average travel speed of each type of samples respectively. By introducing correction factors, the real-time travel speed of the road section without waiting for red lights is obtained.

9. A device for calculating the real-time speed of different sections of roads at different red lights based on the fusion of license plate captured by electronic police and real-time signal control data, characterized in that: include: at least one database; and a memory in communication with at least one database; Among them, the memory stores instructions that can be executed by at least one database, and the instructions are executed by at least one database so that at least one database can execute the method of calculating the real-time unequal red light speed on a road section based on the fusion of license plate captured by electronic police and real-time signal control data as described in any one of claims 1-7.

10. A computer-readable medium having computer-executable instructions stored thereon, characterized in that: When the executable instructions are executed by the processor, a method for calculating the real-time unequal red light speed on a road section based on the fusion of license plate captured by electronic police and real-time signal control data is implemented as described in any one of claims 1-7.

Images