A method and device for analyzing roadside data traffic based on a connected vehicle

Abstract

The application provides a roadside data traffic analysis method and device based on a connected vehicle, comprising: acquiring video data, millimeter wave radar point cloud data and connected vehicle trajectory data, and establishing a traffic flow state estimation model; after time and space synchronization of the video data, the millimeter wave radar point cloud data and the connected vehicle trajectory data, fusing the video data and the millimeter wave radar point cloud data, inputting to the traffic flow state estimation model, and obtaining first traffic flow state characteristic parameters; fusing the video data, the millimeter wave radar point cloud data and the connected vehicle trajectory data, inputting to the traffic flow state estimation model, and obtaining second traffic flow state characteristic parameters; controlling a preset number of characteristic parameters input to a preset adaptive signal control framework according to the first traffic flow state characteristic parameters and the second traffic flow state characteristic parameters, and obtaining a passing efficiency corresponding to each characteristic parameter. The application realizes the purpose of analyzing characteristic parameters on traffic signal control.

Description

Technical Field

[0001] This invention relates to the field of roadside data analysis technology, and specifically to a roadside data traffic analysis method and apparatus based on connected vehicles. Background Technology

[0002] Traffic information perception is a crucial function of traffic information infrastructure, providing essential data and decision support for traffic situation prediction, signal control, and other traffic applications. Currently, common traffic situation perception methods include geomagnetic coils, radar, video, and infrared, but these individual traffic information perception devices generally lack comprehensive information perception and high precision. Roadside sensors have a wide detection range and are equipped with powerful edge computing capabilities. With the advent of 5G communication technology, the high bandwidth and low latency significantly improve information transmission rates, making it possible for intelligent connected vehicles to provide real-time roadside perception information beyond their visual range.

[0003] Current technologies rely on roadside millimeter-wave radar and cameras to acquire vehicle trajectory data near intersections, but the accuracy of this data is insufficient for practical needs. Furthermore, the penetration rate of connected vehicles on roads is relatively low, and the deployment of vehicle-to-infrastructure (V2I) communication is limited. Existing research has largely focused on establishing traffic state estimation models based on varying connected vehicle penetration rates, such as shockwave models and vehicle kinematics models, to estimate vehicle arrival information. However, any estimation model contains errors, resulting in low accuracy of intersection traffic data.

[0004] Therefore, there is an urgent need to propose a roadside data traffic analysis method and device based on connected vehicles to solve the technical problem that the estimation models of roadside millimeter-wave radar, cameras and connected vehicles in the existing technology cannot obtain traffic information at intersections with sufficient accuracy. Summary of the Invention

[0005] In view of this, it is necessary to provide a roadside data traffic analysis method and device based on connected vehicles to solve the technical problem that the estimation models of roadside millimeter-wave radar, cameras and connected vehicles in the prior art cannot obtain traffic information of intersections with sufficient accuracy.

[0006] On the one hand, this invention provides a roadside data traffic analysis method based on connected vehicles, including: Acquire video data from fixed-point observation cameras, millimeter-wave radar point cloud data, and connected vehicle trajectory data to establish a traffic flow state estimation model; After synchronizing the video data, the millimeter-wave radar point cloud data, and the connected vehicle trajectory data in time and space, the video data and the millimeter-wave radar point cloud data are fused and input into the traffic flow state estimation model to obtain the first traffic flow state feature parameters. The video data, the millimeter-wave radar point cloud data, and the connected vehicle trajectory data are fused and input into the traffic flow state estimation model to obtain the second traffic flow state feature parameters. Based on the first traffic flow state characteristic parameters and the second traffic flow state characteristic parameters, a preset number of characteristic parameters input to the preset adaptive signal control framework are controlled to obtain the traffic efficiency corresponding to each characteristic parameter.

[0007] In some possible implementations, after acquiring video data from fixed-point observation cameras, millimeter-wave radar point cloud data, and connected vehicle trajectory data, and establishing a traffic flow state estimation model, the method further includes: The checkerboard pattern captured by the camera at multiple angles was determined according to Zhang Zhengyou's calibration method, and the intrinsic and extrinsic parameters of the camera were calibrated. Based on the chessboard grid from multiple angles, the intrinsic parameters, and the extrinsic parameters, the world coordinate system in the video data is transformed to the pixel coordinate system to obtain pixel video data.

[0008] In some possible implementations, after acquiring video data from fixed-point observation cameras, millimeter-wave radar point cloud data, and connected vehicle trajectory data, and establishing a traffic flow state estimation model, the method further includes: The pixel video data is detected using a deep learning algorithm to obtain the target vehicle, and the target vehicle is tracked to obtain the first vehicle trajectory. Based on the trajectory of the first vehicle, the first feature parameter is obtained; Clustering preprocessing is performed on the millimeter-wave radar point cloud data to obtain clustered point cloud data; The target vehicle in the clustered point cloud data is tracked based on the vehicle kinematics model to obtain the second vehicle trajectory, and a Kalman state transition matrix is ​​constructed. The second feature parameter is obtained based on the second vehicle trajectory and the Kalman state transition matrix; The connected vehicle trajectory data is detected to obtain a third vehicle trajectory and a third feature parameter.

[0009] In some possible implementations, the temporal and spatial synchronization of the video data, the millimeter-wave radar point cloud data, and the connected vehicle trajectory data includes: The first vehicle trajectory of the video data, the second vehicle trajectory of the millimeter-wave radar point cloud data, and the third vehicle trajectory of the connected vehicle trajectory data are transformed to the pixel coordinate system; Based on the sampling frequency of the second vehicle trajectory, the first vehicle trajectory and the third vehicle trajectory are extracted frame by frame to obtain the first extracted trajectory corresponding to the first vehicle trajectory and the second extracted trajectory corresponding to the third vehicle trajectory.

[0010] In some possible implementations, the fusion of the video data and the millimeter-wave radar point cloud data, and the input to the traffic flow state estimation model to obtain first traffic flow state feature parameters, includes: The data from the second vehicle trajectory and the corresponding time points in the first extracted trajectory are fused to obtain the first fused trajectory; The first fused trajectory, the first feature parameter, and the second feature parameter are input into the traffic flow state estimation model to obtain the first traffic flow state feature parameter.

[0011] In some possible implementations, the fusion of the video data, the millimeter-wave radar point cloud data, and the connected vehicle trajectory data, and the input to the traffic flow state estimation model to obtain second traffic flow state feature parameters, includes: The data at corresponding times from the second vehicle trajectory, the first extracted trajectory, and the second extracted trajectory are fused to obtain the second fused trajectory. The second fused trajectory, the first feature parameter, the second feature parameter, and the third feature parameter are input into the traffic flow state estimation model to obtain the second traffic flow state feature parameter.

[0012] In some possible implementations, the step of controlling a preset number of feature parameters input to the preset adaptive signal control framework based on the first traffic flow state feature parameters and the second traffic flow state feature parameters to obtain the traffic efficiency corresponding to each feature parameter includes: By comparing and analyzing the first traffic flow state characteristic parameters and the second traffic flow state characteristic parameters, the analysis results of the effect of the connected vehicle trajectory data on the fixed-point observation are obtained. Based on the analysis results, the preset number of feature parameters input to the preset adaptive signal control framework are controlled to obtain the traffic efficiency corresponding to each feature parameter.

[0013] In some possible implementations, controlling the preset number of feature parameters input to the preset adaptive signal control framework based on the analysis results to obtain the traffic efficiency corresponding to each feature parameter includes: Based on the analysis results, the target feature parameters are determined; The value of each feature parameter is modified to obtain the parameter value corresponding to each feature parameter; The parameter values ​​corresponding to each feature parameter are input into the preset adaptive signal control framework to obtain the changes in the target feature parameters, and the traffic efficiency corresponding to each feature parameter is obtained based on the changes.

[0014] In some possible implementations, the step of inputting the parameter value corresponding to each feature parameter into the preset adaptive signal control framework to obtain the change of the target feature parameter, and obtaining the passage efficiency corresponding to each feature parameter based on the change, includes: The parameter value corresponding to the first feature parameter among the preset number of feature parameters is input into the preset adaptive signal control framework to obtain the change of the target feature parameter; Based on the changes, the traffic efficiency corresponding to the first feature parameter is obtained; Determine whether there is a second feature parameter among the preset number of feature parameters that has not been input to the preset adaptive signal control framework; If so, the parameter value corresponding to the second feature parameter is input into the preset adaptive signal control framework to obtain the passage efficiency corresponding to the second feature parameter; If not, then the traffic efficiency corresponding to each feature parameter is obtained.

[0015] On the other hand, the present invention also provides a roadside data traffic analysis device based on connected vehicles, comprising: The data acquisition module is used to acquire video data from fixed-point observation cameras, millimeter-wave radar point cloud data, and connected vehicle trajectory data to establish a traffic flow state estimation model. The first data fusion module is used to fuse the video data and the millimeter-wave radar point cloud data after synchronizing the video data, the millimeter-wave radar point cloud data and the connected vehicle trajectory data in time and space, and input the fusion data into the traffic flow state estimation model to obtain the first traffic flow state feature parameters. The second data fusion module is used to fuse the video data, the millimeter-wave radar point cloud data, and the connected vehicle trajectory data, and input them into the traffic flow state estimation model to obtain the second traffic flow state feature parameters. The efficiency determination module is used to control a preset number of feature parameters input to the preset adaptive signal control framework based on the first traffic flow state feature parameters and the second traffic flow state feature parameters, so as to obtain the traffic efficiency corresponding to each feature parameter.

[0016] The beneficial effects of the above embodiments are as follows: The roadside data traffic analysis method based on connected vehicles provided by this invention synchronizes and then fuses video data, millimeter-wave radar point cloud data, and connected vehicle trajectory data. This allows for mutual supplementation of data from cameras, millimeter-wave radar, and connected vehicle trajectories. Furthermore, a traffic flow state estimation model is established, and by estimating the traffic flow state estimation model using the traffic flow state estimation model, the accuracy of traffic flow state characteristic parameters is improved. Further, a preset adaptive signal control framework is set up, which controls the signal based on a preset number of characteristic parameters input from the first and second traffic flow state characteristic parameters. This allows for the determination of the traffic efficiency corresponding to each characteristic parameter, achieving the goal of analyzing different characteristic parameters and obtaining the effect of each characteristic parameter on traffic signal control. Attached Figure Description

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

[0018] Figure 1 A schematic flowchart of an embodiment of the roadside data traffic analysis method based on connected vehicles provided by the present invention; Figure 2 This is a schematic diagram of an embodiment of multi-sensor fusion in a networked environment provided by the present invention; Figure 3 A schematic diagram of an embodiment of the roadside data traffic analysis device based on connected vehicles provided by the present invention; Figure 4 A schematic diagram of an embodiment of the electronic device provided by the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0020] Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor systems and / or microcontroller systems.

[0021] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0022] This invention provides a method and apparatus for traffic analysis based on roadside data from connected vehicles, which will be described below.

[0023] Figure 1 This is a schematic flowchart of an embodiment of the roadside data traffic analysis method based on connected vehicles provided by the present invention, as shown below. Figure 1 As shown, the roadside data traffic analysis method based on connected vehicles includes: S101. Acquire video data from fixed-point observation cameras, millimeter-wave radar point cloud data, and connected vehicle trajectory data to establish a traffic flow state estimation model; S102. After synchronizing the video data, millimeter-wave radar point cloud data and connected vehicle trajectory data in time and space, the video data and millimeter-wave radar point cloud data are fused and input into the traffic flow state estimation model to obtain the first traffic flow state characteristic parameters. S103. The video data, millimeter-wave radar point cloud data and connected vehicle trajectory data are fused and input into the traffic flow state estimation model to obtain the second traffic flow state characteristic parameters. S104. Control the preset number of feature parameters input to the preset adaptive signal control framework according to the first traffic flow state feature parameter and the second traffic flow state feature parameter to obtain the traffic efficiency corresponding to each feature parameter.

[0024] Compared with existing technologies, the roadside data traffic analysis method based on connected vehicles provided by this invention synchronizes and then fuses video data, millimeter-wave radar point cloud data, and connected vehicle trajectory data. This allows for mutual supplementation of data from cameras, millimeter-wave radar, and connected vehicle trajectories. Furthermore, a traffic flow state estimation model is established, and by estimating the traffic flow state model using the traffic flow state estimation model, the accuracy of traffic flow state characteristic parameters is improved. Further, a preset adaptive signal control framework is set up, controlling the signal based on a preset number of characteristic parameters input from the first and second traffic flow state characteristic parameters. This allows for the determination of the traffic efficiency corresponding to each characteristic parameter, achieving the goal of analyzing different characteristic parameters and understanding the effect of each characteristic parameter on traffic signal control.

[0025] It should be understood that the methods for acquiring video data, millimeter-wave radar point cloud data, and connected vehicle trajectory data in step S101 can be different. It can be video data acquired from an image acquisition device, or it can be historically stored video data retrieved from a storage medium. The same applies to millimeter-wave radar point cloud data and connected vehicle trajectory data.

[0026] It should be noted that after acquiring the video data, preprocessing is required. In some embodiments of the present invention, the preprocessing may include the following steps after step S101: The checkerboard pattern captured by the camera at multiple angles was determined using Zhang Zhengyou's calibration method, and the camera's internal and external parameters were calibrated. Based on the checkerboard pattern from multiple angles, intrinsic parameters, and extrinsic parameters, the world coordinate system in the video data is transformed to the pixel coordinate system to obtain pixel video data.

[0027] It should be noted that after acquiring the video data, the Zhang Zhengyou calibration method can be used to photograph the chessboard from multiple angles. The intrinsic and extrinsic parameters of the camera can also be calibrated using the MATLAB toolbox, thereby transforming the world coordinate system in the video data to the pixel coordinate system to obtain pixel video data.

[0028] In a specific embodiment of the present invention, the Zhang Zhengyou chessboard calibration method is adopted. A 9*7 chessboard grid is photographed at 20 angles, with each grid measuring 3cm*3cm. Calibration is performed using MATLAB's Toolbox. Based on the calibration results, the camera's extrinsic parameters (rotation matrix and translation vector) and intrinsic parameters (pixel far point coordinates, camera focal length, and unit pixel length) are calculated. The world coordinate system can be first transformed to the camera coordinate system, then to the image coordinate system, and finally to the pixel coordinate system. It is assumed that the radar is installed at a height of [missing information - likely a height in the original text]. The position, the radar beam at the tilt angle When the radar illuminates the ground, the normal to the radar plane is... The radar detected the target at a radial distance of... azimuth angle is With the radar as the origin, the radar normal plane line... for The axis is downward, with the positive direction being downward; the horizontal direction to the left of the radar plane is... Positive direction of the axis, perpendicular to Axial upward Establish in the positive direction of the axis The transformation relationship between the three-dimensional world coordinate system and the camera coordinate system is shown in formula (1): (1) In the formula, Represented as Rotation matrix, , , , Represented as revolving around the world coordinate system axis, axis, shaft Rotation matrix, Represented as Translation vector, x c , y c , z c These are represented as coordinate axes in the camera coordinate system.

[0029] The transformation relationship between the camera coordinate system and the image coordinate system (perspective projection transformation) is shown in formula (2): (2) In the formula, f It is expressed as focal length.

[0030] The common feature of the image coordinate system and the pixel coordinate system is that both are on the imaging plane. However, since the calibration points of their coordinate axes and the units of the two coordinate systems are different, it is necessary to transform the coordinate systems between them. The calibration point of the image coordinate system is the intersection of the camera and the imaging plane, and its coordinate unit is distance unit, representing the physical distance to the intersection point; while the pixel coordinate system divides an image into RGB matrices, and its unit is pixels. Its representation is based on the origin as the calibration point and is expressed in matrix form. Therefore, the transformation between the image coordinate system and the pixel coordinate system is shown in formula (3): (3) In the formula ，dA unit distance representing one pixel. u , v Represented as coordinate axes in a pixel coordinate system.

[0031] Therefore, the transformation from the world coordinate system to the pixel coordinate system of a point is obtained by the transformation of the four coordinate systems as shown in formula (4): (4) In some embodiments of the present invention, the method further includes the following step after step S101: The target vehicle is detected by using a deep learning algorithm to detect pixel video data, and the target vehicle is tracked to obtain the trajectory of the first vehicle. Based on the trajectory of the first vehicle, the first feature parameters are obtained; Clustering preprocessing is performed on millimeter-wave radar point cloud data to obtain clustered point cloud data; The target vehicle in the clustered point cloud data is tracked based on the vehicle kinematics model to obtain the second vehicle trajectory, and the Kalman state transition matrix is ​​constructed. The second feature parameter is obtained based on the second vehicle trajectory and the Kalman state transition matrix; By detecting the trajectory data of connected vehicles, a third vehicle trajectory and a third feature parameter are obtained.

[0032] In specific embodiments of the present invention, pixel video data can be detected using object detection algorithms. Object detection methods for pixel video data mainly include traditional object detection methods and deep learning-based object detection methods. Traditional object detection algorithms primarily rely on filtering between preceding and following frames or background filtering, which has low robustness. Deep learning-based methods, on the other hand, extract target features through deep learning networks to detect targets, resulting in higher detection accuracy and better robustness. This invention specifically describes a single-stage YOLOv5 object detection method based on deep learning. For target tracking, after the target is detected by the object detection algorithm, it is tracked. The Deepsort algorithm can be used to track targets and identify vehicle IDs. Initially, the Deepsort algorithm for target tracking mainly combines the Hungarian algorithm and Kalman filter to achieve target association and target position estimation between preceding and following frames. The Hungarian algorithm can determine whether a target in the current frame is the same as a target in the previous frame. Kalman filtering predicts the tracking position at the current moment based on the target's tracking position at the previous moment, and updates the predicted position based on the detected position at the current moment to obtain the optimal estimated position, thus obtaining the first vehicle trajectory. The first feature parameter is extracted from the first vehicle trajectory. The first feature parameter may include information such as vehicle type and number of vehicles. However, in the actual target detection and tracking process, the camera sensor may miss or falsely detect targets, and the detection results are easily affected by lighting conditions, weather, etc. It can provide information such as vehicle type and number of vehicles, but cannot provide accurate perception information such as vehicle speed and target distance.

[0033] In a specific embodiment of the present invention, the data acquired by the millimeter-wave radar is in the format of distance and angle, that is, the data is presented in two-dimensional polar coordinates. Other data information, such as velocity, needs to be calculated. After obtaining the millimeter-wave radar point cloud data, preprocessing is required. Preprocessing can involve transforming the millimeter-wave radar coordinate system to the world coordinate system. The millimeter-wave radar coordinate system is... The world coordinate system is ,in, p The point is a point in the millimeter-wave radar coordinate system, and it needs to be... p Point projection to world coordinate system In a plane. Assume... p The coordinates of the point in the millimeter-wave radar coordinate system are The coordinates in the world coordinate system are Then p The specific transformation relationship from the millimeter-wave radar coordinate system to the world coordinate system is shown in formula (5): (5) In the formula, H represents the height of unity. The matrix.

[0034] After preprocessing, point clouds of the same vehicle need to be clustered to obtain clustered point cloud data. Then, the classic Kalman filter tracking algorithm is used to track the target, obtaining a second vehicle trajectory. During tracking, the vehicle kinematics model is considered, and a Kalman state transition matrix is ​​constructed. Features are then extracted based on the second vehicle trajectory and the Kalman state transition matrix to obtain second feature parameters. Millimeter-wave radar target detection and tracking can detect vehicle position and speed, but not vehicle type. Therefore, the second feature parameters can include vehicle position and speed. Similarly, connected vehicle trajectory data can be detected to obtain a third vehicle trajectory and third feature parameters. Connected vehicle OBUs and Roadside Equipment (RSE) The implementation of information interaction and sharing by equipment can obtain real-time high-precision and high-resolution trajectory data, but it can only obtain continuous trajectory data when the vehicle is stationary. Therefore, the third feature parameter can include trajectory data when the vehicle is stationary. The feature parameters obtained by camera sensors, millimeter-wave radar and connected vehicles are only a part. If applied alone, the accuracy is not enough. Therefore, the data obtained by camera sensors, millimeter-wave radar and connected vehicles can be fused to improve the accuracy of the acquired data.

[0035] It should be noted that camera sensors, millimeter-wave radar, and connected vehicles use different data formats and coordinate systems when acquiring data. To enable data fusion, in some embodiments of this invention, step S102 includes: The first vehicle trajectory from video data, the second vehicle trajectory from millimeter-wave radar point cloud data, and the third vehicle trajectory from connected vehicle trajectory data are transformed into pixel coordinates. Based on the sampling frequency of the second vehicle trajectory, the first vehicle trajectory and the third vehicle trajectory are extracted frame by frame to obtain the first extracted trajectory corresponding to the first vehicle trajectory and the second extracted trajectory corresponding to the third vehicle trajectory.

[0036] In a specific embodiment of the present invention, before fusing the data of the camera sensor, millimeter-wave radar and connected vehicle, it is necessary to synchronize the first vehicle trajectory of the video data, the second vehicle trajectory of the millimeter-wave radar point cloud data and the third vehicle trajectory of the connected vehicle trajectory data in time and space. The first vehicle trajectory of the video data has already been converted to the pixel coordinate system. Now it is necessary to perform coordinate system transformation on the second vehicle trajectory of the millimeter-wave radar point cloud data and the third vehicle trajectory of the connected vehicle trajectory data. The second vehicle trajectory of the millimeter-wave radar point cloud data is the second vehicle trajectory after preprocessing and transforming the millimeter-wave radar coordinate system to the world coordinate system. Then, according to formulas (1)-(4), the second vehicle trajectory of the millimeter-wave radar point cloud data in the pixel coordinate system can be obtained. Similarly, the third vehicle trajectory of the connected vehicle trajectory data can be transformed to the pixel coordinate system to obtain the third vehicle trajectory in the pixel coordinate system, thus realizing the spatial synchronization of the camera sensor, millimeter-wave radar and connected vehicle.

[0037] Because millimeter-wave radar, camera sensors, and connected vehicles each have their own inherent output data time intervals, i.e., sampling frequencies, the data output frequencies of multiple sensors are inconsistent. To ensure that data from multiple sensors can be obtained at the same time, time synchronization of the collected data is necessary. For example, if the sampling frequency of the millimeter-wave radar is 20Hz, the sampling frequency of the camera is 30Hz, and the sampling frequency of the connected vehicle is 28Hz, then because the image frequency acquired by the camera vision sensor is higher, the images acquired by the camera sensor and the connected vehicle can be extracted frame by frame according to the sampling frequency of the millimeter-wave radar. This allows us to obtain the first extracted trajectory corresponding to the first vehicle trajectory and the second extracted trajectory corresponding to the third vehicle trajectory.

[0038] In some embodiments of the present invention, step S102 includes: The data from the second vehicle trajectory and the corresponding time points in the first extracted trajectory are fused to obtain the first fused trajectory; The first fused trajectory, the first feature parameter, and the second feature parameter are input into the traffic flow state estimation model to obtain the first traffic flow state feature parameter.

[0039] In a specific embodiment of the present invention, the data at the corresponding time in the second vehicle trajectory and the first extracted trajectory can be fused to obtain the fused first fused trajectory. The first fused trajectory, the first feature parameter and the second feature parameter can be input into the traffic flow state estimation model. The traffic flow state estimation model can obtain the first traffic flow state feature parameter, such as traffic volume, queue length, travel time and other information.

[0040] In some embodiments of the present invention, step S103 includes: The data at corresponding times from the second vehicle trajectory, the first extracted trajectory, and the second extracted trajectory are fused to obtain the second fused trajectory; The second fused trajectory, the first feature parameter, the second feature parameter, and the third feature parameter are input into the traffic flow state estimation model to obtain the second traffic flow state feature parameter.

[0041] In a specific embodiment of the present invention, the data at corresponding times in the second vehicle trajectory, the first extracted trajectory, and the second extracted trajectory can be fused to obtain the fused second fused trajectory. The second fused trajectory, the first feature parameter, the second feature parameter, and the third feature parameter can then be input into the traffic flow state estimation model. The traffic flow state estimation model can obtain the second traffic flow state feature parameter, which can be information such as flow rate, queue length, and travel time. The second traffic flow state feature parameter is a more accurate traffic flow state feature parameter, which compensates for the shortcomings of camera and millimeter-wave radar fusion in detecting jitter and trajectory dispersion when fusion is used for stationary vehicles.

[0042] In specific embodiments of the present invention, such as Figure 2 As shown, multi-sensor fusion in a networked environment can include a sensor data acquisition module, a fusion preprocessing module, a fusion module, and a feature parameter extraction module. The sensor data acquisition module can include steps for extracting and tracking image data features from video data from camera sensors, acquiring millimeter-wave radar feature data (millimeter-wave radar point cloud data) from millimeter-wave radar, and acquiring connected vehicle data (connected vehicle trajectory data) from roadside RSUs. The fusion preprocessing module synchronizes the data acquired from multiple sensors in time and space, and then the fusion module performs data association (image fusion) and state association (feature fusion) on the synchronized data. The feature parameter extraction module extracts feature parameter information such as flow rate, queue length, and travel time obtained from the traffic flow state estimation model.

[0043] In some embodiments of the present invention, step S104 includes: A comparative analysis of the first and second traffic flow state characteristic parameters was conducted to obtain the analysis results of the effect of connected vehicle trajectory data on fixed-point observation. Based on the analysis results, a preset number of feature parameters are controlled by the input of the preset adaptive signal control framework to obtain the passage efficiency corresponding to each feature parameter.

[0044] In a specific embodiment of the present invention, the first traffic flow state characteristic parameter and the second traffic flow state characteristic parameter can be compared and analyzed to obtain the analysis results of the effect of connected vehicle trajectory data on fixed-point observation, and the impact of connected vehicle data on the effect of fixed-point observation can be obtained. Furthermore, an adaptive signal control framework can be built based on the data acquired by camera sensors, millimeter-wave radar and connected vehicles. The value of the third characteristic parameter of the connected vehicle trajectory data can also be changed, and the changed characteristic parameter can be input into the adaptive signal control framework, thereby obtaining the traffic efficiency corresponding to each characteristic parameter.

[0045] It should be noted that, in order to determine the impact of each feature parameter on the traffic efficiency of a signalized intersection, in some embodiments of the present invention, a preset number of feature parameters input to a preset adaptive signal control framework are controlled based on the analysis results to obtain the traffic efficiency corresponding to each feature parameter, including: Based on the analysis results, the target feature parameters are determined; The values ​​of each feature parameter are modified individually to obtain the corresponding parameter values ​​for each feature parameter; The parameter values ​​corresponding to each feature parameter are input into the preset adaptive signal control framework to obtain the changes in the target feature parameters, and based on the changes, the passage efficiency corresponding to each feature parameter is obtained.

[0046] In a specific embodiment of the present invention, target feature parameters can be determined based on the analysis results. For example, vehicle delay can be determined as a signal control evaluation index (target feature parameter). Then, the value of each feature parameter can be modified. Each feature parameter can be a feature parameter of environmental factors, such as connected vehicle penetration rate, intersection saturation, and traffic flow at different times. Then, the changes in the signal control evaluation index (target feature parameter) can be observed, and the traffic efficiency corresponding to each feature parameter can be obtained based on the changes.

[0047] In some embodiments of the present invention, the parameter values ​​corresponding to each feature parameter are input to a preset adaptive signal control framework to obtain the changes in the target feature parameters, and the passage efficiency corresponding to each feature parameter is obtained based on the changes, including: The parameter value corresponding to the first feature parameter among the preset number of feature parameters is input into the preset adaptive signal control framework to obtain the change of the target feature parameter; Based on the changes, the traffic efficiency corresponding to the first feature parameter is obtained; Determine whether there is a second feature parameter among the preset number of feature parameters that has not been input to the preset adaptive signal control framework; If so, the parameter value corresponding to the second feature parameter is input into the preset adaptive signal control framework to obtain the passage efficiency corresponding to the second feature parameter; If not, then the passage efficiency corresponding to each feature parameter is obtained.

[0048] In a specific embodiment of the present invention, a preset number of feature parameters can be modified one by one and input into a preset adaptive signal control framework. For example, the modified parameter value of the first feature parameter can be input into the preset adaptive signal control framework. By observing the changes in the signal control evaluation index (target feature parameter), the traffic efficiency corresponding to the first feature parameter can be obtained, that is, the influence of the first feature parameter on the traffic efficiency of the signalized intersection. Then, it can be determined whether there are any second feature parameters among the preset number of feature parameters that have not been input into the preset adaptive signal control framework. If so, the step of "inputting the parameter value corresponding to the first feature parameter among the preset number of feature parameters into the preset adaptive signal control framework to obtain the changes in the target feature parameter" is repeated for the second feature parameter, and the loop continues until there are no second feature parameters among the preset number of feature parameters that have not been input into the preset adaptive signal control framework. Then, the traffic efficiency corresponding to each feature parameter can be obtained, that is, the influence of each feature parameter on the traffic efficiency of the signalized intersection.

[0049] To better implement the roadside data traffic analysis method based on connected vehicles in the embodiments of the present invention, the embodiments of the present invention also provide a roadside data traffic analysis device based on connected vehicles, such as... Figure 3 As shown, the roadside data traffic analysis device based on connected vehicles includes: The data acquisition module 301 is used to acquire video data from fixed-point observation cameras, millimeter-wave radar point cloud data, and connected vehicle trajectory data to establish a traffic flow state estimation model. The first data fusion module 302 is used to fuse video data and millimeter-wave radar point cloud data after synchronizing video data, millimeter-wave radar point cloud data and connected vehicle trajectory data in time and space, and input them into the traffic flow state estimation model to obtain the first traffic flow state feature parameters. The second data fusion module 303 is used to fuse video data, millimeter-wave radar point cloud data and connected vehicle trajectory data, and input them into the traffic flow state estimation model to obtain the second traffic flow state feature parameters. The efficiency determination module 304 is used to control a preset number of characteristic parameters input to the preset adaptive signal control framework based on the first traffic flow state characteristic parameters and the second traffic flow state characteristic parameters, so as to obtain the traffic efficiency corresponding to each characteristic parameter.

[0050] The roadside data traffic analysis device based on connected vehicles provided in the above embodiments can realize the technical solutions described in the above embodiments of the roadside data traffic analysis method based on connected vehicles. The specific implementation principles of each module or unit can be found in the corresponding content in the above embodiments of the roadside data traffic analysis method based on connected vehicles, and will not be repeated here.

[0051] like Figure 4 As shown, the present invention also provides an electronic device 400. The electronic device 400 includes a processor 401, a memory 402, and a display 403. Figure 4 Only some components of the electronic device 400 are shown, but it should be understood that it is not required to implement all the components shown, and more or fewer components may be implemented instead.

[0052] In some embodiments, memory 402 may be an internal storage unit of electronic device 400, such as a hard disk or memory of electronic device 400. In other embodiments, memory 402 may also be an external storage device of electronic device 400, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc. equipped on electronic device 400.

[0053] Furthermore, the memory 402 may include both internal storage units of the electronic device 400 and external storage devices. The memory 402 is used to store application software and various types of data installed on the electronic device 400.

[0054] In some embodiments, processor 401 may be a central processing unit (CPU), microprocessor, or other data processing chip, used to run program code stored in memory 402 or process data, such as the roadside data traffic analysis method based on connected vehicles in this invention.

[0055] In some embodiments, display 403 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen. Display 403 is used to display information from electronic device 400 and to display a visual user interface. Components 401-403 of electronic device 400 communicate with each other via a system bus.

[0056] In some embodiments of the present invention, when the processor 401 executes the roadside data traffic analysis program based on connected vehicles in the memory 402, the following steps can be implemented: Acquire video data from fixed-point observation cameras, millimeter-wave radar point cloud data, and connected vehicle trajectory data to establish a traffic flow state estimation model; After synchronizing video data, millimeter-wave radar point cloud data, and connected vehicle trajectory data in time and space, the video data and millimeter-wave radar point cloud data are fused and input into the traffic flow state estimation model to obtain the first traffic flow state characteristic parameters. Video data, millimeter-wave radar point cloud data, and connected vehicle trajectory data are fused and input into the traffic flow state estimation model to obtain the second traffic flow state characteristic parameters; Based on the first traffic flow state characteristic parameters and the second traffic flow state characteristic parameters, a preset number of characteristic parameters input to the preset adaptive signal control framework are controlled to obtain the traffic efficiency corresponding to each characteristic parameter.

[0057] It should be understood that when the processor 401 executes the roadside data traffic analysis program based on connected vehicles in the memory 402, in addition to the functions mentioned above, it can also perform other functions, as can be found in the description of the corresponding method embodiments above.

[0058] Furthermore, this embodiment of the invention does not specifically limit the type of electronic device 400 mentioned. Electronic device 400 can be a mobile phone, tablet computer, personal digital assistant (PDA), wearable device, laptop computer, or other portable electronic device. Exemplary embodiments of portable electronic devices include, but are not limited to, portable electronic devices running iOS, Android, Microsoft, or other operating systems. The aforementioned portable electronic device can also be other portable electronic devices, such as a laptop computer with a touch-sensitive surface (e.g., a touch panel). It should also be understood that in some other embodiments of the invention, electronic device 400 may not be a portable electronic device, but rather a desktop computer with a touch-sensitive surface (e.g., a touch panel).

[0059] Accordingly, this application also provides a computer-readable storage medium for storing computer-readable programs or instructions. When the programs or instructions are executed by a processor, they can implement the steps or functions of the roadside data traffic analysis method based on connected vehicles provided in the above-described method embodiments.

[0060] Those skilled in the art will understand that all or part of the processes of the methods described in the above embodiments can be implemented by a computer program instructing related hardware (such as a processor, controller, etc.), and the computer program can be stored in a computer-readable storage medium. The computer-readable storage medium may be a disk, optical disk, read-only memory, or random access memory, etc.

[0061] The above provides a detailed description of the roadside data traffic analysis method and device based on connected vehicles provided by this invention. Specific examples have been used to illustrate the principles and implementation methods of this invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this invention. At the same time, those skilled in the art will recognize that there will be changes in the specific implementation methods and application scope based on the ideas of this invention. Therefore, the content of this specification should not be construed as a limitation of this invention.

Claims

1. A traffic analysis method based on roadside data from connected vehicles, characterized in that, include: Acquire video data from fixed-point observation cameras, millimeter-wave radar point cloud data, and connected vehicle trajectory data to establish a traffic flow state estimation model; After synchronizing the video data, the millimeter-wave radar point cloud data, and the connected vehicle trajectory data in time and space, the video data and the millimeter-wave radar point cloud data are fused and input into the traffic flow state estimation model to obtain the first traffic flow state feature parameters. The video data, the millimeter-wave radar point cloud data, and the connected vehicle trajectory data are fused and input into the traffic flow state estimation model to obtain the second traffic flow state feature parameters. Based on the first traffic flow state characteristic parameters and the second traffic flow state characteristic parameters, a preset number of characteristic parameters input to the preset adaptive signal control framework are controlled to obtain the traffic efficiency corresponding to each characteristic parameter; After acquiring video data from fixed-point observation cameras, millimeter-wave radar point cloud data, and connected vehicle trajectory data, and establishing a traffic flow state estimation model, the method further includes: The target vehicle is detected by using a deep learning algorithm to detect pixel video data, and the target vehicle is tracked to obtain the first vehicle trajectory. Based on the trajectory of the first vehicle, the first feature parameter is obtained; Clustering preprocessing is performed on the millimeter-wave radar point cloud data to obtain clustered point cloud data; The target vehicle in the clustered point cloud data is tracked based on the vehicle kinematics model to obtain the second vehicle trajectory, and a Kalman state transition matrix is ​​constructed. The second feature parameter is obtained based on the second vehicle trajectory and the Kalman state transition matrix; The connected vehicle trajectory data is detected to obtain a third vehicle trajectory and a third feature parameter.

2. The roadside data traffic analysis method based on connected vehicles according to claim 1, characterized in that, After acquiring video data from fixed-point observation cameras, millimeter-wave radar point cloud data, and connected vehicle trajectory data, and establishing a traffic flow state estimation model, the method further includes: The checkerboard pattern captured by the camera at multiple angles was determined according to Zhang Zhengyou's calibration method, and the intrinsic and extrinsic parameters of the camera were calibrated. Based on the chessboard grid from multiple angles, the intrinsic parameters, and the extrinsic parameters, the world coordinate system in the video data is transformed to the pixel coordinate system to obtain pixel video data.

3. The roadside data traffic analysis method based on connected vehicles according to claim 1, characterized in that, The temporal and spatial synchronization of the video data, the millimeter-wave radar point cloud data, and the connected vehicle trajectory data includes: The first vehicle trajectory of the video data, the second vehicle trajectory of the millimeter-wave radar point cloud data, and the third vehicle trajectory of the connected vehicle trajectory data are transformed to the pixel coordinate system; Based on the sampling frequency of the second vehicle trajectory, the first vehicle trajectory and the third vehicle trajectory are extracted frame by frame to obtain the first extracted trajectory corresponding to the first vehicle trajectory and the second extracted trajectory corresponding to the third vehicle trajectory.

4. The roadside data traffic analysis method based on connected vehicles according to claim 3, characterized in that, The process of fusing the video data and the millimeter-wave radar point cloud data and inputting them into the traffic flow state estimation model yields first traffic flow state feature parameters, including: The data from the second vehicle trajectory and the corresponding time points in the first extracted trajectory are fused to obtain the first fused trajectory; The first fused trajectory, the first feature parameter, and the second feature parameter are input into the traffic flow state estimation model to obtain the first traffic flow state feature parameter.

5. The roadside data traffic analysis method based on connected vehicles according to claim 3, characterized in that, The process of fusing the video data, the millimeter-wave radar point cloud data, and the connected vehicle trajectory data, and inputting them into the traffic flow state estimation model, yields second traffic flow state feature parameters, including: The data at corresponding times from the second vehicle trajectory, the first extracted trajectory, and the second extracted trajectory are fused to obtain the second fused trajectory. The second fused trajectory, the first feature parameter, the second feature parameter, and the third feature parameter are input into the traffic flow state estimation model to obtain the second traffic flow state feature parameter.

6. The roadside data traffic analysis method based on connected vehicles according to claim 1, characterized in that, The step of controlling a preset number of feature parameters input to the preset adaptive signal control framework based on the first traffic flow state feature parameters and the second traffic flow state feature parameters to obtain the traffic efficiency corresponding to each feature parameter includes: By comparing and analyzing the first traffic flow state characteristic parameters and the second traffic flow state characteristic parameters, the analysis results of the effect of the connected vehicle trajectory data on the fixed-point observation are obtained. Based on the analysis results, the preset number of feature parameters input to the preset adaptive signal control framework are controlled to obtain the traffic efficiency corresponding to each feature parameter.

7. The roadside data traffic analysis method based on connected vehicles according to claim 6, characterized in that, The step of controlling the preset number of feature parameters input to the preset adaptive signal control framework based on the analysis results to obtain the traffic efficiency corresponding to each feature parameter includes: Based on the analysis results, the target feature parameters are determined; The value of each feature parameter is modified to obtain the parameter value corresponding to each feature parameter; The parameter values ​​corresponding to each feature parameter are input into the preset adaptive signal control framework to obtain the changes in the target feature parameters, and the traffic efficiency corresponding to each feature parameter is obtained based on the changes.

8. The roadside data traffic analysis method based on connected vehicles according to claim 7, characterized in that, The step of inputting the parameter value corresponding to each feature parameter into the preset adaptive signal control framework to obtain the change of the target feature parameter, and obtaining the passage efficiency corresponding to each feature parameter based on the change, includes: The parameter value corresponding to the first feature parameter among the preset number of feature parameters is input into the preset adaptive signal control framework to obtain the change of the target feature parameter; Based on the changes, the traffic efficiency corresponding to the first feature parameter is obtained; Determine whether there is a second feature parameter among the preset number of feature parameters that has not been input to the preset adaptive signal control framework; If so, the parameter value corresponding to the second feature parameter is input into the preset adaptive signal control framework to obtain the passage efficiency corresponding to the second feature parameter; If not, then the traffic efficiency corresponding to each feature parameter is obtained.

9. A roadside data traffic analysis device based on connected vehicles, characterized in that, The device is used to implement the roadside data traffic analysis method based on connected vehicles as described in any one of claims 1 to 8. The device includes: a data acquisition module, used to acquire video data from fixed-point observation cameras, millimeter-wave radar point cloud data, and connected vehicle trajectory data, and to establish a traffic flow state estimation model. The first data fusion module is used to fuse the video data and the millimeter-wave radar point cloud data after synchronizing the video data, the millimeter-wave radar point cloud data and the connected vehicle trajectory data in time and space, and input the fusion data into the traffic flow state estimation model to obtain the first traffic flow state feature parameters. The second data fusion module is used to fuse the video data, the millimeter-wave radar point cloud data, and the connected vehicle trajectory data, and input them into the traffic flow state estimation model to obtain the second traffic flow state feature parameters. The efficiency determination module is used to control a preset number of feature parameters input to the preset adaptive signal control framework based on the first traffic flow state feature parameters and the second traffic flow state feature parameters, so as to obtain the traffic efficiency corresponding to each feature parameter.

Images