A Traffic Flow State Recognition Method Based on Multi-Source Heterogeneous Data Sources

Abstract

This invention provides a traffic flow state recognition method based on multi-source heterogeneous data sources, including collecting and classifying road traffic information data; preprocessing the traffic information data; inputting the preprocessed traffic information data into corresponding classification models according to their types; outputting intermediate traffic states for each type; assigning weights to the intermediate traffic states for each type; and calculating the final traffic state. The beneficial effects of this invention are that it adopts customized feature extraction and transformation methods tailored to the characteristics of different types of sensor data, which can better preserve the semantic information of different types of data and improve the quality and effect of data fusion; it integrates the performance of different models on the validation set and dynamically adjusts the voting weights according to the model's accuracy, making the final prediction results more accurate and reliable.

Description

Technical Field

[0001] This invention belongs to the field of road traffic artificial intelligence technology, and in particular relates to a traffic flow state recognition method based on multi-source heterogeneous data sources. Background Technology

[0002] With the acceleration of urbanization and the rapid increase in vehicle ownership, highways, as vital transportation arteries, are experiencing ever-increasing traffic flow, leading to frequent traffic congestion and accidents, posing a severe challenge to the safe and efficient operation of highways. Traditional traffic flow status identification methods mostly rely on data from fixed vehicle detectors or single data sources, resulting in incomplete data acquisition, poor real-time performance, and weak anti-interference capabilities, making it difficult to meet the needs of modern highway emergency command systems. Therefore, developing an efficient, accurate, and multi-source data fusion method for highway traffic flow status identification is particularly important. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention provides a traffic flow state identification method based on multi-source heterogeneous data sources, which is particularly suitable for processing traffic flow state identification methods with multi-source heterogeneous data sources.

[0004] The technical solution adopted in this invention is as follows: Firstly, a traffic flow state identification method based on multi-source heterogeneous data sources is provided, comprising the following steps:

[0005] Collect and classify road traffic information data;

[0006] The traffic information data is preprocessed;

[0007] The preprocessed traffic information data is input into the corresponding classification model according to its type;

[0008] Output the intermediate traffic status corresponding to each type;

[0009] Assign weights to each type of intermediate traffic state;

[0010] The final traffic state is calculated.

[0011] Furthermore, it also includes the following steps:

[0012] The road network is modeled using a spatiotemporal graph convolutional network model;

[0013] The final traffic state of the target road is comprehensively corrected by combining the real-time traffic conditions of other roads in the road network;

[0014] The final traffic status is uploaded to the traffic management system and / or user terminal.

[0015] Furthermore, the types of traffic information data include numerical data and image data; the traffic information data includes average vehicle speed, traffic flow, and lane occupancy rate.

[0016] Furthermore, the preprocessing of the traffic information data includes the following steps:

[0017] Data cleaning is performed using spatial clustering algorithms;

[0018] Perform data transformation;

[0019] Perform unified representation of attribute domains;

[0020] Centroid clustering analysis was performed using the K-means algorithm.

[0021] Data dimensionality reduction is performed using principal component analysis algorithms.

[0022] Data compression is performed using differential coding.

[0023] Furthermore, the unified representation of attribute domains includes the following steps:

[0024] For the numerical data in the traffic information data, their numerical attributes are directly retained;

[0025] The image data in the traffic information data is converted into numerical vector form after feature extraction.

[0026] Establish a data dictionary to clarify the name, data type, value range, and meaning of each feature field.

[0027] Furthermore, in the central cluster analysis, through equations Calculate the centroid, where Let the centroid of the k-th cluster be . Let k be the set of data points for the k-th cluster. For clusters Number of data points For clusters One of the data points.

[0028] Furthermore, the training of the classification model includes the following steps:

[0029] Collect historical traffic information data;

[0030] The historical traffic information data is preprocessed;

[0031] The historical traffic information data is tagged according to classification rules;

[0032] The classification model includes an SVM model and a decision tree model. The SVM model is constructed for the average vehicle speed, and the decision tree model is constructed for the traffic flow and the lane occupancy rate.

[0033] The historical traffic information data is input into the SVM model and the decision tree model respectively according to the classification for training;

[0034] The classification results of the SVM model and the decision tree model are integrated using an ensemble learning method.

[0035] Furthermore, the SVM model optimizes the classification boundary by maximizing the objective function, which is: ,in For Lagrange multipliers, For sample labels, For kernel function, This represents the number of samples.

[0036] Furthermore, the lane occupancy rate Through equations The average lane occupancy rate is obtained through the equation. Received, among which Let Y be the number of vehicles in the i-th lane, Y be the number of vehicles in one direction, and N be the number of lanes.

[0037] Furthermore, the final traffic state Through equations ,in The weight of the average speed of the vehicle. The weight of the traffic flow, The weight of the lane occupancy rate, The intermediate traffic state refers to the average speed of the vehicles. The intermediate traffic state is the traffic flow. The intermediate traffic state is the lane occupancy rate.

[0038] The advantages and positive effects of this invention are as follows: By adopting the above technical solution and employing customized feature extraction and transformation methods tailored to the characteristics of different types of sensor data, the semantic information of different types of data can be better preserved, improving the quality and effectiveness of data fusion; it integrates the performance of different models on the validation set and dynamically adjusts the voting weights based on the model's accuracy, making the final prediction results more accurate and reliable; it improves the accuracy and real-time performance of traffic flow state identification, providing fast and accurate data support for highway emergency command systems; it enhances anti-interference capabilities, reducing errors and interference from single data sources; and it optimizes the formulation and implementation of traffic management strategies, improving the efficiency of traffic resource allocation and management level. Attached Figure Description

[0039] Figure 1 This is a flowchart illustrating a traffic flow state identification method based on multi-source heterogeneous data sources according to an embodiment of the present invention. Detailed Implementation

[0040] The present disclosure will now be described more fully with reference to the accompanying drawings, which illustrate exemplary embodiments of the present disclosure. The technical solutions of the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present disclosure, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present disclosure without creative effort are within the scope of protection of the present disclosure.

[0041] like Figure 1 As shown, this invention provides a traffic flow state identification method based on multi-source heterogeneous data sources, comprising the following steps:

[0042] S100. Collect and classify road traffic information data;

[0043] Classification algorithms are used to initially screen potential data sources and select parameters that are highly correlated with traffic flow conditions, including but not limited to vehicle speed, traffic volume, and lane occupancy rate.

[0044] S200: Preprocess traffic information data;

[0045] S300. Input the preprocessed traffic information data into the corresponding classification model according to the type;

[0046] S400 outputs the intermediate traffic status corresponding to each type, including smooth traffic, slow traffic and congestion;

[0047] S500: Assign weights to various types of intermediate traffic states;

[0048] Different weights are assigned to the three classification results: vehicle speed, traffic flow, and lane occupancy. The methods for assigning classification weights include those based on historical data, expert experience, and those designed according to specific road conditions. The sum of all weights is 1.

[0049] S600, calculate the final traffic state.

[0050] Using the above method, the optimal processing path is adapted for different data types, and key information of the original data is preserved; independent processing by category avoids the pollution of the global result by a single data source; and parallel processing of each data stream effectively reduces system latency.

[0051] To address the problem that traditional methods ignore the traffic flow transmission effect between road segments and cannot overcome the spatial limitations of single-point state recognition, this embodiment provides an implementation method.

[0052] In one embodiment, the following steps are also included:

[0053] The road network is modeled using a spatiotemporal graph convolutional network model;

[0054] The final traffic state of the target road is comprehensively adjusted by combining the real-time traffic conditions of other roads in the road network.

[0055] The final traffic status will be uploaded to the traffic management system and / or user terminal.

[0056] The identification results are fed back to the traffic management system or user terminal in real time. At the same time, time series analysis algorithms are used to predict the trend of traffic status changes in the future, providing a scientific basis for traffic scheduling, route planning, information dissemination, etc.

[0057] Using the above method, the spatiotemporal propagation pattern of traffic flow is captured by ST-GCN to predict the trend of congestion spread; the results of this road segment are corrected by comprehensively considering the status of adjacent road segments, which effectively reduces the false alarm rate; and a certain early warning window is provided for traffic management systems and end users.

[0058] In one embodiment, the types of traffic information data include numerical and image types; the traffic information data includes average vehicle speed, traffic flow, and lane occupancy rate.

[0059] To address the issue of scale incompatibility among multi-source data, this embodiment provides an implementation method.

[0060] In one embodiment, preprocessing traffic information data includes the following steps:

[0061] Data cleaning is performed using spatial clustering algorithms;

[0062] A density-based spatial clustering algorithm is used to identify and remove noise points and outliers from the dataset. The spatial clustering algorithm is applied to identify noise points, and a neighborhood radius is defined. and the minimum number of points, MinPts, data points It is a noise point if and only if ,in for of - A set of points within the domain.

[0063] Perform data transformation;

[0064] Data transformation techniques, including but not limited to linear transformation and normalization, are used to ensure that data from different data sources are compared and clustered on the same scale.

[0065] Perform unified characterization of the attribute domain;

[0066] Unify the data collected from different sensors (such as fixed vehicle detectors, GPS devices, video cameras) in the attribute domain, including but not limited to unifying all speed data to kilometers per hour (km / h), unifying all traffic flow data to vehicles per hour (veh / h), and unifying the lane occupancy rate data to a percentage (%).

[0067] Perform central clustering analysis through the K-means algorithm;

[0068] Perform data dimensionality reduction through the principal component analysis algorithm;

[0069] Assume that the dimension of the original data is d, and the dimension of the data after dimensionality reduction is d′ (d′ < d). Then the goal of PCA is to minimize the reconstruction error. Among them, V is the transformation matrix. is the projection matrix projected onto the principal component space.

[0070] Perform data compression through differential coding.

[0071] The preprocessing methods also include spatio-temporal synchronization of multi-source data. For the GPS data delay problem, a timestamp-based compensation method is adopted. Record the timestamp of each GPS data point, compare it with the current system time, calculate the delay time, and perform interpolation processing on the GPS data according to the delay time to estimate the position of the GPS data at the correct time point. Use the linear interpolation method to calculate the position coordinates at the correct time point based on two known GPS data points before and after the delay time.

[0072] The preprocessing methods also include the video sampling rate and vehicle detector data alignment strategy. For the problem of misalignment between the video sampling rate and vehicle detector data, a timestamp alignment and resampling method is adopted. Add timestamps to the video frames and vehicle detector data to ensure that each data point has accurate time information. Align the video frames and vehicle detector data according to the timestamps. For the case of inconsistent sampling rates, adopt the resampling method to resample the high-sampling-rate data to the same sampling rate as the low-sampling-rate data, or adopt the interpolation method to estimate the values of missing data points based on the values of adjacent data points to ensure the consistency of the data in the time dimension.

[0073] Adopt the above methods to establish a comparable data benchmark through unified characterization of the attribute domain; differential coding compression reduces the transmission bandwidth.

[0074] To solve the problems of feature explosion caused by directly inputting original pixels in traditional methods and the semantic gap between video data and numerical vectors, an implementation method is provided in this embodiment.

[0075] In one embodiment, performing a unified representation of attribute domains includes the following steps:

[0076] For numerical data in traffic information data, its numerical attributes are directly retained;

[0077] Image data in traffic information data is converted into numerical vector form after feature extraction.

[0078] Establish a data dictionary to clarify the name, data type, value range, and meaning of each feature field, ensuring data consistency and understandability.

[0079] Using the above method, the vehicle's motion trajectory is analyzed from video frames to capture micro-behaviors such as rapid acceleration or lane changes; the influence of brightness fluctuations caused by weather is eliminated through feature vector transformation.

[0080] To address the issue of computational efficiency limitations under massive datasets and resource waste caused by processing all sensor data, this embodiment provides an implementation method.

[0081] In one embodiment, center clustering is selected. (k=1,2,...,K) serves as a representative data source, where X={x1,x2,...,x} n} represents the original dataset, and n represents the number of data points. In centroid clustering analysis, the equation is used... Calculate the centroid, where Let the centroid of the k-th cluster be . Let k be the set of data points for the k-th cluster. For clusters Number of data points For clusters One of the data points.

[0082] Using the above method, representative centroid points are selected to filter redundant data and reduce computation time; the K-means algorithm can more accurately select representative data sources based on the inherent characteristics of the data during clustering.

[0083] To address the irreconcilable conflict between rule-driven and data-driven approaches, the poor interpretability of pure machine learning models, and the insufficient flexibility of pure rule-based models, this embodiment provides an implementation method.

[0084] In one embodiment, training the classification model includes the following steps:

[0085] Collect historical traffic information data;

[0086] Preprocess historical traffic information data;

[0087] Historical traffic information data is labeled according to classification rules;

[0088] Based on three indicators—vehicle speed, traffic flow, and lane occupancy rate—corresponding thresholds are set to construct classification rules. If the average vehicle speed is above a certain threshold (e.g., 80 km / h), the lane occupancy rate is low (e.g., less than 20%), and the traffic flow is moderate (e.g., less than 60% of the road's design capacity), then the traffic is considered to be in a smooth flow condition. If the vehicle speed is below a certain threshold (e.g., 40 km / h), the traffic flow exceeds a certain percentage of the road's design capacity (e.g., 80%), and the lane occupancy rate is above a certain threshold (e.g., 40%), then the traffic is considered to be in a congested condition.

[0089] The classification models include SVM models and decision tree models. An SVM model is built for average vehicle speed, and a decision tree model is built for traffic flow and lane occupancy.

[0090] The decision tree model uses the C4.5 decision tree algorithm to construct a binary tree structure, selects the best splitting attribute based on the information gain ratio, and makes independent judgments on each indicator.

[0091] Historical traffic information data is input into the SVM model and decision tree model respectively according to the classification for training;

[0092] The classification results of SVM and decision tree models are integrated using an ensemble learning method.

[0093] The SVM model and decision tree model are used in parallel, and the results are integrated using a voting method. The training dataset is simultaneously input into both the SVM and decision tree models for training. During the prediction phase, for each new input sample, the trained SVM and decision tree models are used to make predictions, respectively, yielding the prediction results from both models. Then, a vote is performed based on the prediction results of the two models. If the prediction results of the two models are consistent, this result is directly used as the final prediction result; if the prediction results of the two models are inconsistent, a weighted vote is performed based on the accuracy of the two models on the validation set. The model with the higher accuracy has a larger weight, thus obtaining the final prediction result.

[0094] Using the above method, decision trees explicitly embed traffic domain knowledge, while SVM implicitly learns complex speed patterns; the voting mechanism enables the overall accuracy to exceed that of any single model.

[0095] To address the problem that traditional threshold methods cannot identify linearly inseparable scenarios such as "high-speed but congested" and speed abrupt changes, this embodiment provides an implementation method.

[0096] In one embodiment, the SVM model optimizes the classification boundary by maximizing an objective function, where the objective function is: ,in For Lagrange multipliers, For sample labels, For kernel function, This represents the number of samples.

[0097] For classifying vehicle speeds, the objective function is expressed as follows: ,in For Lagrange multipliers, For the sample Label, For kernel function, The sample size is used to calculate the sample size. and Similarity between kernels is used to determine the similarity between kernels. Commonly used kernel functions include linear kernels, polynomial kernels, and radial basis function (RBF) kernels.

[0098] Using the above method, the hypersurface classification boundary is constructed through the RBF kernel function, which improves the recognition rate of sudden deceleration; grid search ensures that the model is continuously optimized as road conditions change.

[0099] To address the problem of traditional methods neglecting lane differences and distorting the average occupancy rate across the entire road segment, this embodiment provides an implementation method.

[0100] In one embodiment, lane occupancy rate Through equations The average lane occupancy rate is obtained through the equation. Received, among which Let Y be the number of vehicles in the i-th lane, Y be the number of vehicles in one direction, and N be the number of lanes.

[0101] Object detection algorithms (such as the YOLO series) are used to detect vehicles in the video and obtain their positional information (such as bounding box coordinates). Then, based on pre-defined lane line positions, it is determined whether the detected vehicles are within the lanes, and the number of vehicles within the lanes is counted. Simultaneously, combined with the video's frame rate, the number of vehicles passing through a particular lane per unit time is calculated, thus yielding the lane occupancy rate. Lane occupancy data primarily comes from video analysis, using video data for clearing and lane marking recognition.

[0102] Using the above method, lane-specific calculations avoid the obscuring of overall smooth traffic flow by lanes with high truck density; and the source lanes of congestion are identified by average lane occupancy rate.

[0103] To address the issue of insufficient accuracy in single-dimensional traffic status, this embodiment provides an implementation method.

[0104] In one embodiment, the final traffic state Through equations ,in The weighting for the average vehicle speed. As a weight for traffic flow, The weight for lane occupancy rate, This represents the intermediate traffic conditions with average vehicle speed. This represents an intermediate traffic state with varying vehicle flow. This represents the intermediate traffic state based on lane occupancy rate.

[0105] Using the above method, the weighted results are more interpretable than the output of the black-box model.

[0106] The following description, in conjunction with a preferred embodiment, illustrates the content involved in the above embodiments.

[0107] On a certain section of highway, data was acquired including data from fixed vehicle detectors, mobile detection data (GPS data), and video image data. At time T1, the fixed vehicle detector data was v1 = 40 km / h, q1 = 120 vehicles / hour, the mobile detection data was GPS data 1, and the video image data o1 = 30%. At time T2, the fixed vehicle detector data was v2 = 20 km / h, q2 = 180 vehicles / hour, the mobile detection data was GPS data 2, and the video image data o2 = 45%. At time Tn, the fixed vehicle detector data was v... n =10km / h, q n =250 vehicles / hour, mobile detection data includes GPS data n and video image data o. n=60%. Density-based noise spatial clustering algorithm was used to identify and remove noise points and outliers in the dataset. Anomalies were found in GPS data 2 at time T2, so it was removed. Principal component analysis was used to reduce the dimensionality of the data while preserving key information. Assuming the original data dimension was 3 (speed, traffic flow, lane occupancy), the dimensionality of the reduced data remained 3, but the data had been optimized. Thresholds for vehicle speed, traffic flow, and lane occupancy were set to determine traffic flow status. If the average vehicle speed was below a certain threshold (e.g., 30 km / h), the lane occupancy was high (e.g., greater than 50%), and the traffic flow was large (e.g., exceeding 80% of the road's design capacity), it was judged as a congested state. An SVM classification model was used to independently judge the three indicators of vehicle speed, traffic flow, and lane occupancy. The preprocessed data was input into the SVM model to obtain preliminary classification results. The intermediate result obtained at time T1 was: vehicle speed was above the threshold, lane occupancy was below the threshold, and traffic flow was moderate, indicating a smooth flow state. Time T2 (outlier data removed): Vehicle speed is below the threshold, lane occupancy is above the threshold, and traffic volume is high, initially indicating congestion (further fusion analysis required). Time Tn: Vehicle speed is far below the threshold, lane occupancy is far above the threshold, and traffic volume is extremely high, initially indicating congestion (further fusion analysis required). Based on historical data and expert experience, different weights are assigned to the three classification results: vehicle speed, traffic volume, and lane occupancy (e.g., speed 0.4, traffic volume 0.3, lane occupancy 0.3). The three classification results and their weights are weighted and summed to obtain the final traffic flow state probability. The weighted results at times T2 and Tn both point to congestion. The identification results are fed back to the traffic management system in real time. The system immediately transmits the information to the emergency command system and triggers corresponding emergency response measures, such as issuing traffic guidance information.

[0108] Based on embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0109] An electronic device includes at least one processor and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the traffic flow state identification method based on multi-source heterogeneous data sources provided in this disclosure.

[0110] Electronic devices are intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic devices can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0111] A non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to execute the traffic flow state identification method based on multi-source heterogeneous data sources provided in this disclosure.

[0112] The various embodiments of this disclosure can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0113] A computer program product includes a computer program / instructions, which are executed by a processor to provide a traffic flow state recognition method based on multi-source heterogeneous data sources.

[0114] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0115] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0116] The embodiments of the present invention have been described in detail above, but the content described is only a preferred embodiment of the present invention and should not be considered as limiting the scope of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the patent coverage of the present invention.

Claims

1. A traffic flow state identification method based on multi-source heterogeneous data sources, characterized in that, Includes the following steps: Collect and classify road traffic information data; The traffic information data is preprocessed; The preprocessed traffic information data is input into the corresponding classification model according to its type; Output the intermediate traffic status corresponding to each type; Assign weights to each type of intermediate traffic state; The final traffic state is calculated; The traffic information data includes numerical and image types; the traffic information data includes average vehicle speed, traffic flow, and lane occupancy rate. The training of the classification model includes the following steps: Collect historical traffic information data; The historical traffic information data is preprocessed; The historical traffic information data is tagged according to classification rules; The classification model includes an SVM model and a decision tree model. The SVM model is constructed for the average vehicle speed, and the decision tree model is constructed for the traffic flow and the lane occupancy rate. The historical traffic information data is input into the SVM model and the decision tree model respectively according to the classification for training; The classification results of the SVM model and the decision tree model are integrated using an ensemble learning method.

2. The traffic flow state identification method based on multi-source heterogeneous data sources according to claim 1, characterized in that, It also includes the following steps: The road network is modeled using a spatiotemporal graph convolutional network model; The final traffic state of the target road is comprehensively corrected by combining the real-time traffic conditions of other roads in the road network; The final traffic status is uploaded to the traffic management system and / or user terminal.

3. The traffic flow state identification method based on multi-source heterogeneous data sources according to claim 1, characterized in that, Preprocessing the traffic information data includes the following steps: Data cleaning is performed using spatial clustering algorithms; Perform data transformation; Perform unified representation of attribute domains; Centroid clustering analysis was performed using the K-means algorithm. Data dimensionality reduction is performed using principal component analysis algorithms. Data compression is performed using differential coding.

4. The traffic flow state identification method based on multi-source heterogeneous data sources according to claim 3, characterized in that, Performing a unified representation of attribute domains includes the following steps: For the numerical data in the traffic information data, their numerical attributes are directly retained; The image data in the traffic information data is converted into numerical vector form after feature extraction. Establish a data dictionary to clarify the name, data type, value range, and meaning of each feature field.

5. The traffic flow state identification method based on multi-source heterogeneous data sources according to claim 3, characterized in that: In centroid cluster analysis, the equation is used Calculate the centroid, where Let the centroid of the k-th cluster be... Let k be the set of data points in the k-th cluster. For clusters Number of data points For clusters One of the data points.

6. The traffic flow state identification method based on multi-source heterogeneous data sources according to claim 1, characterized in that: The SVM model optimizes the classification boundary by maximizing the objective function, which is: ,in For Lagrange multipliers, For sample labels, For kernel function, This represents the number of samples.

7. The traffic flow state identification method based on multi-source heterogeneous data sources according to claim 1, characterized in that: The lane occupancy rate Through equations The average lane occupancy rate is obtained through the equation. Received, among which Let Y be the number of vehicles in the i-th lane, Y be the number of vehicles in one direction, and N be the number of lanes.

8. The traffic flow state identification method based on multi-source heterogeneous data sources according to claim 1, characterized in that: The final traffic state Through equations ,in The weight of the average speed of the vehicle. The weight of the traffic flow, The weight of the lane occupancy rate, The intermediate traffic state refers to the average speed of the vehicles. The intermediate traffic state is the traffic flow. The intermediate traffic state is the lane occupancy rate.

Images