Lane traffic state estimation method and device based on connected vehicle, terminal and medium

Abstract

The application provides a lane traffic state estimation method and device based on a connected vehicle, a terminal and a medium, and belongs to the technical field of intelligent traffic. The method comprises the following steps: constructing a space-time grid of a target lane, and setting a sensing range for each connected vehicle; obtaining driving data reported by each connected vehicle on the target lane; obtaining second driving data with data state marks based on the driving data and the space-time grid, wherein the data state comprises validity and absence; inputting the second driving data and a preset noise vector into a generator of a trained generative adversarial model, and outputting third driving data after processing; and estimating the lane traffic state based on the third driving data. The application collects driving data through the connected vehicle, introduces the driving data into the data state marks, and then uses the trained generator to complete the data with the missing data state marks to estimate the road condition, so that the accuracy of the traffic state estimation under the condition of low connected vehicle penetration rate can be effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent transportation technology, and in particular to a lane traffic state estimation method, device, terminal and medium based on a connected vehicle. Background Art

[0002] Accurately estimating traffic conditions is key to optimizing traffic management strategies, improving lane efficiency, and ensuring driving safety. Among existing lane traffic state estimation methods, the most common approach is to use fixed detectors, such as geomagnetic sensors and cameras, installed at specific locations along the lanes to collect traffic information and estimate lane traffic conditions based on this information. However, the limited number of fixed detectors makes it difficult to collect comprehensive and detailed data.

[0003] With the development of V2X (vehicle-to-everything) communication technology, connected vehicles can act as floating vehicles or mobile sensors, reporting their position, speed, and other relevant information at a preset frequency. This provides a new approach to estimating lane traffic states. However, currently, the majority of vehicles in lanes are non-connected, and the penetration rate of connected vehicles remains low. This situation leads to data sparsity and incompleteness in traffic state estimation, resulting in low accuracy in traffic state estimates based on this approach.

[0004] Therefore, the existing technology has defects and needs to be improved and developed. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a lane traffic state estimation method, device, terminal and medium based on a connected vehicle in response to the above-mentioned defects of the existing technology, aiming to solve the problem of low accuracy of traffic state estimation results in the existing technology.

[0006] The technical solutions adopted by the present invention to solve the technical problems are as follows:

[0007] In a first aspect, an embodiment of the present invention provides a method for estimating lane traffic status based on a connected vehicle, the method comprising:

[0008] Constructing a spatiotemporal grid for the target lane and setting a sensing range for each connected vehicle in the target lane within the spatiotemporal grid;

[0009] Obtaining driving data reported by each connected vehicle in the target lane, the driving data including the connected vehicle's own driving data and / or driving data collected by the connected vehicle from non-connected vehicles within its sensing range;

[0010] Based on the driving data and the spatiotemporal grid, obtaining second driving data with a data status mark, wherein the data status includes valid and missing;

[0011] inputting the second driving data and a preset noise vector into a generator of a trained generative adversarial model, and outputting, after processing, third driving data after completion;

[0012] estimating a lane traffic state based on the third driving data.

[0013] In an embodiment, a space-time grid of a target lane is constructed, including:

[0014] acquiring lane geometry data of the target lane;

[0015] generating a space-time grid of the target lane based on the lane geometry data and a preset time resolution and a preset space resolution.

[0016] In an embodiment, based on the driving data and the space-time grid, second driving data with data state labels is obtained, including:

[0017] after data preprocessing of the driving data, preprocessed driving data is obtained;

[0018] a mask matrix labeling data states of each grid is generated according to a coverage state of each preprocessed driving data in the space-time grid;

[0019] second driving data with data state labels is obtained based on the mask matrix and all preprocessed driving data.

[0020] In an embodiment, a mask matrix labeling data states of each grid is generated according to a coverage state of each preprocessed driving data in the space-time grid, including:

[0021] each preprocessed driving data is mapped to a corresponding grid of the space-time grid;

[0022] for each grid, it is judged whether the grid is covered by preprocessed driving data, and a data state of the grid is determined according to a judgment result;

[0023] a mask matrix is generated based on data states of all grids.

[0024] In an embodiment, for each grid, it is judged whether the grid is covered by preprocessed driving data, and a data state of the grid is determined according to a judgment result, including:

[0025] for each grid, it is judged whether there is self-driving data of a connected car, and if there is, a data state of the grid is determined as valid;

[0026] if the grid does not have self-driving data of a connected car, it is judged whether the grid is within a sensing range of any connected car;

[0027] If the grid is not within the sensing range of any connected vehicle, the data status of the grid is determined to be missing;

[0028] If the grid is within the sensing range of any connected vehicle, its status is determined to be valid or missing based on whether there is driving data of a non-connected vehicle.

[0029] In one embodiment, the training step of generating an adversarial model includes:

[0030] Obtaining a training dataset, wherein the training dataset includes a plurality of training subsets corresponding to preset connected vehicle penetration rates, each training subset including training driving data marked with a training data status, and benchmark driving data corresponding to each connected vehicle penetration rate, the benchmark driving data being derived from driving data collected by a traffic monitoring platform and connected vehicles, and the training data status including valid and missing;

[0031] Initializing a generative adversarial model, wherein the generative adversarial model includes a generator and a discriminator;

[0032] For each connected vehicle penetration rate, the generator and discriminator are trained for a preset number of rounds;

[0033] When the training of all connected vehicle penetration rates is completed, the data completion effect of the generative adversarial model is evaluated. After passing the evaluation, a trained generative adversarial model is obtained.

[0034] In one embodiment, for each connected vehicle penetration rate, the generator and discriminator are trained for a predetermined number of rounds, including:

[0035] Fixed the discriminator parameters and trained the generator for the first preset round. In each round, the noise training vector and the training driving data marked with the training data status at the current connected vehicle penetration rate were input into the generator. The training driving data with the missing training data status were supplemented and the supplemented driving data were output.

[0036] The completed driving data and the baseline driving data with the same connected vehicle penetration rate are fed into the discriminator, the generator loss is calculated, and the generator parameters are updated through backpropagation.

[0037] Fix the generator parameters, train the discriminator for the second preset round, input the completed driving data of the current connected vehicle penetration rate and the baseline driving data into the discriminator, calculate the discriminator loss and backpropagate to update the discriminator parameters.

[0038] In a second aspect, an embodiment of the present invention further provides a lane traffic state estimation device based on a connected vehicle, comprising:

[0039] A grid construction module is used to construct a spatiotemporal grid for the target lane and set a sensing range for each connected vehicle in the target lane within the spatiotemporal grid;

[0040] A data acquisition module is configured to acquire driving data reported by each connected vehicle in the target lane, wherein the driving data includes the connected vehicle's own driving data and / or driving data collected by the connected vehicle from non-connected vehicles within its sensing range;

[0041] A second driving data generating module is configured to obtain second driving data with a data status mark based on the driving data and the spatiotemporal grid, wherein the data status includes valid and missing;

[0042] a data completion module, configured to input the second driving data and a preset noise vector into a generator of a trained generative adversarial model, and output completed third driving data after processing;

[0043] A state estimation module is used to estimate the lane traffic state based on the third driving data.

[0044] In a third aspect, an embodiment of the present invention further provides a terminal comprising: a memory, a processor, and a lane traffic state estimation program based on a connected vehicle stored in the memory and runnable on the processor, wherein the lane traffic state estimation program based on a connected vehicle, when executed by the processor, implements the steps of the lane traffic state estimation method based on a connected vehicle as described above.

[0045] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, which stores a lane traffic state estimation program based on a connected vehicle, and the lane traffic state estimation program based on a connected vehicle can be executed to implement the steps of the lane traffic state estimation method based on a connected vehicle as described above.

[0046] Beneficial effects of the present invention: The present invention constructs a spatiotemporal grid of the target lane and sets a perception range for each connected vehicle; obtains driving data reported by each connected vehicle on the target lane; based on the driving data and the spatiotemporal grid, obtains second driving data with a data status mark, wherein the data status includes valid and missing; inputs the second driving data and a preset noise vector into a generator of a trained generative adversarial model, and outputs completed third driving data after processing; and estimates lane traffic status based on the third driving data. The present invention collects driving data from connected vehicles and introduces it into data status marks, and then uses a trained generator to complete the data marked as missing to estimate road conditions, which can effectively improve the accuracy of traffic status estimation under low connected vehicle penetration rates. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a flowchart of a preferred embodiment of the lane traffic state estimation method based on a connected vehicle in the present invention.

[0048] Figure 2 1 is a schematic diagram of the generation process of the mask matrix in the present invention.

[0049] Figure 3 It is a schematic diagram of the generator data processing flow in the present invention.

[0050] Figure 4 It is a schematic diagram of the generation process of the final mask training matrix in the present invention.

[0051] Figure 5 It is a schematic diagram of the discriminator data processing flow in the present invention.

[0052] Figure 6 It is a flowchart of the discriminator and generator jointly processing data in the present invention.

[0053] Figure 7 It is a structural diagram of a preferred embodiment of the lane traffic state estimation method based on a connected vehicle in the present invention.

[0054] Figure 8 It is a block diagram of the terminal principle of the present invention. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solutions and advantages of the present invention more clear and distinct, the present invention is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0056] Accurately estimating traffic conditions is key to optimizing traffic management strategies, improving lane efficiency, and ensuring driving safety. Among existing lane traffic state estimation methods, the most common approach is to use fixed detectors, such as geomagnetic sensors and cameras, installed at specific locations along the lanes to collect traffic information and estimate lane traffic conditions based on this information. However, the limited number of fixed detectors makes it difficult to collect comprehensive and detailed data.

[0057] With the development of V2X (vehicle-to-everything) communication technology, connected vehicles can act as floating vehicles or mobile sensors, reporting their position, speed, and other relevant information at a preset frequency. This provides a new approach to estimating lane traffic states. However, currently, the majority of vehicles in lanes are non-connected, and the penetration rate of connected vehicles remains low. This situation leads to data sparsity and incompleteness in traffic state estimation, resulting in low accuracy in traffic state estimates based on this approach.

[0058] In view of the above defects of the prior art, the present application provides a lane traffic state estimation method, device, terminal and medium based on a connected vehicle, belonging to the technical field of intelligent transportation, which comprises the following steps: constructing a space-time grid of a target lane, and setting a perception range for each connected vehicle; obtaining driving data reported by each connected vehicle on the target lane; obtaining second driving data with data state marks based on the driving data and the space-time grid, wherein the data state comprises valid and missing; inputting the second driving data and a preset noise vector into a generator of a trained generative adversarial model, and outputting third driving data after processing; and estimating the lane traffic state based on the third driving data. The present application can effectively improve the accuracy of traffic state estimation under low connected vehicle penetration rate by collecting driving data through connected vehicles, introducing the driving data into data state marks, and then using a trained generator to complete the data with missing data state marks to estimate the road condition.

[0059] See Figure 1 The lane traffic state estimation method based on a connected vehicle according to the present application comprises the following steps:

[0060] Step S100, constructing a space-time grid of a target lane, and setting a perception range of each connected vehicle on the target lane in the space-time grid.

[0061] Specifically, the space-time grid is a grid data form combining space and time. After obtaining the lane geometry data of the target lane, the space-time grid of the target lane can be generated based on the lane geometry data and preset time resolution and space resolution. The lane geometry data is the spatial position data of the lane, including parameters such as the shape, width, length and curvature of the lane, which are obtained from third-party map data. The time resolution refers to the accuracy of the space-time grid in the time dimension, i.e. the frequency of updating the grid data represented by each time unit (such as seconds, minutes, hours, etc.). The space resolution refers to the accuracy of the space-time grid in the space dimension, i.e. the actual space range (such as meters, kilometers, etc.) represented by each grid unit.

[0062] The connected vehicle is also called an intelligent connected vehicle or a vehicle internet vehicle, which comprises a speed sensor, a GPS positioning sensor, an acceleration sensor, a laser radar, a camera and other sensors, and a communication module and other parts. These sensors and modules enable the connected vehicle to obtain its own speed, position, acceleration and other information in real time, and exchange information with other vehicles or transportation infrastructure through the communication module. The connected vehicle can perceive the existence of other non-connected vehicles and collect driving data of other connected vehicles using these sensors. In the space-time grid, the perception distance of the connected vehicle is set at a preset distance from the center of the connected vehicle. The preset distance can be set according to the sensor performance of the connected vehicle, the traffic complexity of the target lane and other factors.

[0063] See Figure 1 The lane traffic state estimation method based on a connected vehicle according to an embodiment of the present invention includes the following steps:

[0064] Step S200: Acquire driving data reported by each connected vehicle on the target lane, wherein the driving data includes the connected vehicle's own driving data and / or driving data of non-connected vehicles collected by the connected vehicle within its perception range.

[0065] Specifically, connected vehicles can use their own sensors to report their own driving data and can also collect driving data from non-connected vehicles within their sensing range (i.e., non-connected vehicle driving data). Driving data includes vehicle speed, vehicle position, acceleration, driving direction, and timestamps.

[0066] See Figure 1 The lane traffic state estimation method based on a connected vehicle according to an embodiment of the present invention further includes the following steps:

[0067] Step S300: Based on the driving data and the space-time grid, obtain second driving data with a data status mark, where the data status includes valid and missing.

[0068] Specifically, based on the driving data and the spatiotemporal grid, driving data that identifies valid or missing data can be obtained. This second driving data is then fed into the generator of the trained generative adversarial model for completion, yielding the completed driving data. This approach can address the issue of sparse traffic data collected by connected vehicles at low penetration rates.

[0069] In one implementation, obtaining second driving data with a data status mark based on the driving data and the spatiotemporal grid includes:

[0070] After performing data preprocessing on the driving data, preprocessed driving data is obtained;

[0071] Generate a mask matrix marking the data state of each grid according to the coverage state of each pre-processed driving data in the space-time grid;

[0072] Based on the mask matrix and all pre-processed driving data, second driving data with a data status mark is obtained.

[0073] Specifically, after the driving data is obtained, it is first preprocessed. Preprocessing includes data cleaning and format unification. Data processing uses an outlier detection algorithm based on statistical analysis and traffic rules to remove obviously erroneous or abnormal data, such as data with speeds exceeding a reasonable range and location information that does not conform to the lane topology. Format unification is to unify the data collected by different sensors into a standard format for easy subsequent processing. The mask matrix is ​​a matrix with element values ​​0 and 1, which is used to mark the data status of each grid. When the element value of the mask matrix is ​​0, it indicates that the data status of the grid is missing. When the element value of the mask matrix is ​​1, it indicates that the data status of the grid is valid. The mask matrix is ​​used to simulate the data missing situation under different connected vehicle penetration rates to reflect the data sparsity caused by the limited number of connected vehicles in actual lanes.

[0074] In one implementation, a mask matrix marking the data state of each grid is generated according to the coverage state of each pre-processed driving data in the spatiotemporal grid, including:

[0075] Mapping each preprocessed driving data into a corresponding grid of the space-time grid;

[0076] For each grid, determine whether it is covered by the pre-processed driving data, and determine the data status of the grid according to the determination result;

[0077] Generates a mask matrix based on the data state of all rasters.

[0078] Specifically, the present invention converts discrete driving data into a continuous data representation in a spatiotemporal grid coordinate system and uses a mask matrix to identify which grids have valid driving data and which grids have missing data. The mask matrix and the preprocessed driving data are then used to generate a second set of driving data with data status markers, facilitating completion of the adversarial network model generator.

[0079] In one implementation, determining whether each grid is covered by the pre-processed driving data and determining the data status of the grid according to the determination result includes:

[0080] For each grid, it is determined whether there is any driving data of the connected vehicle. If so, the data status of the grid is determined to be valid.

[0081] If the grid does not contain any connected vehicle's own driving data, then determine whether the grid is within the sensing range of any connected vehicle;

[0082] If the grid is not within the sensing range of any connected vehicle, the data status of the grid is determined to be missing;

[0083] If the grid is within the sensing range of any connected vehicle, its status is determined to be valid or missing based on whether there is driving data of a non-connected vehicle.

[0084] Specifically, if the grid is within the sensing range of any connected vehicle, it will continue to determine whether there is non-connected vehicle driving data. If there is non-connected vehicle driving data, its status is determined to be valid. If there is no non-connected vehicle driving data, its status is determined to be missing. The schematic diagram of the mask matrix generation process is as follows: Figure 2 shown.

[0085] See Figure 1 The lane traffic state estimation method based on a connected vehicle according to an embodiment of the present invention includes the following steps:

[0086] Step S400: Input the second driving data and the preset noise vector into the generator of the trained generative adversarial model, and output the completed third driving data after processing.

[0087] Specifically, the generative adversarial model includes a generator and a discriminator. Figure 3 The figure is a schematic diagram of the generator data processing flow. The generator includes an input layer, a flattening layer, a fully connected layer, an output layer, and a reshaping layer connected in sequence. After the second driving data and the preset noise vector are input to the input layer, the data is spliced ​​in the last dimension (i.e., the number of features). The noise vector can be set according to the complexity of the lane-level traffic speed data and the generation diversity requirements. The flattening layer then processes the spliced ​​data into a one-dimensional vector for subsequent fully connected layer operations. The data then passes through multiple fully connected layers to extract global features, the number of neurons decreases, and the relu activation function is used to enhance the nonlinear expression ability of the model. Finally, it passes through the output layer, is activated by the linear activation function, and then the reshaping layer (i.e., the reshape layer) restores the data to its original shape.

[0088] In one implementation, the training step of generating an adversarial model includes:

[0089] Obtaining a training dataset, wherein the training dataset includes a plurality of training subsets corresponding to preset connected vehicle penetration rates, each training subset including training driving data marked with a training data status, and benchmark driving data corresponding to each connected vehicle penetration rate, the benchmark driving data being derived from driving data collected by a traffic monitoring platform and connected vehicles, and the training data status including valid and missing;

[0090] Initializing a generative adversarial model, wherein the generative adversarial model includes a generator and a discriminator;

[0091] For each connected vehicle penetration rate, the generator and discriminator are trained for a preset number of rounds;

[0092] When the training of all connected vehicle penetration rates is completed, the data completion effect of the generative adversarial model is evaluated. After passing the evaluation, a trained generative adversarial model is obtained.

[0093] Specifically, a set of driving data collected by the traffic monitoring platform and the target training connected vehicle on the training lane within a preset time range is obtained. This driving data set is then screened and simulated to generate a number of benchmark driving data corresponding to the penetration rate of connected vehicles.

[0094] The generation process of training driving data with training data status mark includes: dividing the target training lane into corresponding training time-space grids, and setting the perception range of each training networked vehicle on the target lane in the training time-space grid; obtaining the training driving data reported by each training networked vehicle on the target training lane, the training driving data including the training networked vehicle's own training driving data and / or the training driving data of non-networked vehicles collected by the training networked vehicle within its perception range; based on the coverage status of the training driving data in the training time-space grid, obtaining an initial mask matrix, the initial mask matrix is ​​a matrix composed of 0 and 1, where 0 represents that the data of the grid is missing and 1 represents that the data of the grid is valid; The baseline driving data and training driving data corresponding to the penetration rate are mapped onto the training spatiotemporal grid, respectively, to obtain the first and second traffic state matrices. The second traffic state matrix is ​​subtracted from the first traffic state matrix to obtain a difference matrix. The difference matrix and the masked training matrix have the same dimensions and correspond to the same training spatiotemporal grid. If a zero element exists in the difference matrix, it indicates that there is no difference between the baseline driving data and the training driving data. In this case, the corresponding element in the masked training matrix is ​​maintained at 1, indicating valid data. If a non-zero element exists in the difference matrix, it indicates that there is a difference between the baseline driving data and the training driving data. In this case, the corresponding element in the masked training matrix is ​​updated to 0, indicating missing data. After determining the element values ​​of all difference matrices and updating or maintaining the element values ​​of the masked training matrix, the final masked training matrix is ​​obtained. Based on the final masked training matrix and the driving training data at this connected vehicle penetration rate, training driving data with training data status labels are obtained. The resulting final masked training matrix accurately simulates the data sparsity at low penetration rates, providing realistic data samples for subsequent model training.

[0095] In one implementation, an initial mask matrix is ​​obtained based on the coverage of the training driving data on the training spatiotemporal grid, including:

[0096] Mapping the training driving data to corresponding grids of the training spatiotemporal grid;

[0097] For each grid, determine whether it is covered by the training driving data, and determine the data status of the grid according to the judgment result;

[0098] Generates an initial mask matrix based on the data state of all rasters.

[0099] Specifically, for each grid, determine whether there is self-training driving data of the connected vehicle. If so, determine that the data status of the grid is valid; if there is no self-training driving data of the connected vehicle for the grid, determine whether the grid is within the perception range of any connected vehicle; if the grid is not within the perception range of any connected vehicle, determine that the data status of the grid is missing; if the grid is within the perception range of any connected vehicle, determine whether its status is valid or missing based on whether there is non-connected vehicle driving training data. When there is non-connected vehicle driving training data, determine that its status is valid. If there is no non-connected vehicle driving training data, determine that its status is invalid. The final mask training matrix generation process diagram is shown in the figure below. Figure 4 shown.

[0100] In one implementation, dividing the target training lane into corresponding training spatiotemporal grids includes:

[0101] Obtain the lane geometry data of the target training lane;

[0102] A training spatiotemporal grid of the target training lane is generated based on the lane geometry data of the target training lane and a preset second temporal resolution and a second spatial resolution.

[0103] For each connected vehicle penetration rate, the generator and discriminator are trained for a preset number of rounds, including:

[0104] Fixed the discriminator parameters and trained the generator for the first preset round. In each round, the noise training vector and the training driving data marked with the training data status at the current connected vehicle penetration rate were input into the generator. The training driving data with the missing training data status were supplemented and the supplemented driving data were output.

[0105] The completed driving data and the baseline driving data with the same connected vehicle penetration rate are fed into the discriminator, the generator loss is calculated, and the generator parameters are updated through backpropagation.

[0106] Fix the generator parameters, train the discriminator for the second preset round, input the completed driving data of the current connected vehicle penetration rate and the baseline driving data into the discriminator, calculate the discriminator loss and backpropagate to update the discriminator parameters.

[0107] Specifically, the discriminator is built based on a convolutional neural network, and its data processing flow diagram is as follows: Figure 5As shown. First, the input training driving data with training data status marks and the benchmark driving data under the penetration rate of the connected vehicle are converted into three-dimensional form through the reshaping layer (i.e., the reshape layer) to adapt to the subsequent convolution layer operation. Then, the data passes through multiple convolution layers to extract deep-level features of the data. The padding method is same to ensure that the output size after the convolution operation is the same as the input. At the same time, the relu activation function is used to introduce nonlinear features. Through these multiple convolution layers, the discriminator can automatically learn the local features of the data in space, so as to better judge the authenticity of the data. After processing by the convolution layer, the data is processed into a one-dimensional vector by the flattening layer, and then the discrimination result is output through a fully connected layer (1 neuron, the activation function is sigmoid). The output range of the sigmoid function is between 0 and 1. Close to 1 means that the discriminator believes that the input data is real data, and close to 0 means that it is generated data. The discriminator extracts and distinguishes features from the baseline driving data and the supplemented driving data output by the generator, providing feedback to the generator, prompting the generator to continuously optimize its generation strategy, improve the quality of generated data, and make the generated data more difficult for the discriminator to distinguish, ultimately achieving a distribution similar to that of real data. The flowchart of the discriminator and generator jointly processing data is shown in the figure below. Figure 6 As shown in the figure, the raw data is the set of driving data collected from the traffic monitoring platform and the target training connected vehicles on the training lanes within a preset time range. The real samples are the baseline driving data corresponding to each connected vehicle penetration rate. The fake samples output by the generator are the supplemented driving data corresponding to each connected vehicle penetration rate. The supplemented driving data corresponding to each connected vehicle penetration rate and the baseline driving data are input into the discriminator for judgment. Continuous training leads to convergence of the loss function.

[0108] The training process uses the Adam optimizer and binary cross entropy loss function to iteratively optimize the GAN model.

[0109] The noise training vector is sampled from a normal distribution. The binary cross entropy loss function is calculated as follows: Among them, L represents the result of the binary cross entropy loss function; N represents the number of samples; y i Indicates the true label of sample i, which takes a value of 0 or 1. In the discriminator training scenario, for real data samples (i.e., benchmark driving data), y i =1, for generating data samples (i.e. completing driving data), y i =0;p i It represents the probability that the model predicts that sample i is a true data sample, and its value range is between [0,1].

[0110] The trained generative adversarial model can accurately learn the distribution of traffic data, complete the input traffic data with missing values, and output the completed traffic data for subsequent completion effect evaluation and practical application, providing reliable data support for traffic status estimation.

[0111] The steps for evaluating the generative adversarial model are as follows: from the driving data set collected by the traffic monitoring platform and the target training connected vehicle for the training lane within a preset time range, extract the missing reference data corresponding to the missing parts in the final mask training matrix as the benchmark for evaluating the completion effect; in the completed driving data output by the generative adversarial model, filter out the missing completion data corresponding to the missing parts according to the final mask matrix; calculate the root mean square error and mean absolute error between the missing reference data and the missing completion data. The formula for calculating the root mean square error is: The formula for calculating the mean absolute error is: Where n is the number of missing data points, y i is the value of the missing benchmark data, The value for missing data.

[0112] The root mean square error (RMSE) reflects the overall degree of deviation of the completed data from the original real data. A smaller RMSE indicates a more stable model. The mean absolute error (MAE) reflects the absolute deviation between the model's completed data and the original missing data. A smaller MAE indicates a higher prediction accuracy. By calculating the RMSE and MAE under different penetration rates, we comprehensively evaluate the model's completion performance under different levels of data sparsity, providing strong data support for model optimization and traffic state estimation.

[0113] When both indicators reach the corresponding preset thresholds, it is determined that the evaluation has passed and a trained generative adversarial model is obtained.

[0114] See Figure 1 The lane traffic state estimation method based on a connected vehicle according to an embodiment of the present invention includes the following steps:

[0115] Step S500: Estimate lane traffic status based on the third driving data.

[0116] Specifically, lane traffic conditions can be estimated based on the supplemented third driving data. This third driving data includes driving data from both connected and non-connected vehicles. Due to the supplementation, the data is more complete. This third driving data can be combined with other relevant traffic information, such as road grade, traffic light status, historical traffic flow data, and current weather conditions, to comprehensively determine lane-level traffic conditions.

[0117] In one implementation, the lane-level traffic status is comprehensively judged based on the vehicle speed in the third driving data, combined with the road grade, traffic light status, historical traffic flow data, and current weather conditions.

[0118] Specifically, if the completed vehicle speed remains below a preset speed threshold for a preset duration, and combined with the red light duration of the traffic light and historical congestion data, the lane is judged to be congested. If the completed vehicle speed is within the preset speed range with minimal fluctuations, and the lane's historical traffic data shows smooth flow during the same time period, and the current weather conditions are good, then the traffic condition is considered relatively smooth. This approach enables accurate estimation of lane-level traffic status for connected vehicles at low penetration rates, helping traffic management departments formulate scientific traffic diversion strategies, planning reasonable travel routes for travelers, and providing reliable decision-making basis for intelligent driving systems.

[0119] In summary, the present invention constructs a spatiotemporal grid for the target lane and sets a perception range for each connected vehicle; obtains driving data reported by each connected vehicle on the target lane; based on the driving data and the spatiotemporal grid, obtains second driving data with data status markers, where the data statuses include valid and missing; inputs the second driving data and a preset noise vector into a generator of a trained generative adversarial model, and outputs completed third driving data after processing; and estimates lane traffic status based on the third driving data. The present invention collects driving data from connected vehicles, introduces data status markers into it, and then uses a trained generator to complete data marked as missing to estimate road conditions. This can effectively improve the accuracy of traffic status estimation under low connected vehicle penetration rates.

[0120] In one embodiment, if Figure 7 As shown, based on the above-mentioned lane traffic state estimation method based on a connected vehicle, the present invention also provides a lane traffic state estimation device based on a connected vehicle, including:

[0121] A grid construction module 100 is used to construct a spatiotemporal grid for the target lane and set a sensing range for each connected vehicle in the target lane within the spatiotemporal grid;

[0122] A data acquisition module 200 is configured to acquire driving data reported by each connected vehicle in the target lane, wherein the driving data includes the connected vehicle's own driving data and / or driving data collected by the connected vehicle from non-connected vehicles within its sensing range;

[0123] A second driving data generating module 300 is configured to obtain second driving data with a data status mark based on the driving data and the spatiotemporal grid, wherein the data status includes valid and missing;

[0124] a data completion module 400 , configured to input the second driving data and a preset noise vector into a generator of a trained generative adversarial model, and output completed third driving data after processing;

[0125] The state estimation module 500 is configured to estimate the lane traffic state based on the third driving data.

[0126] It should be noted that the aforementioned explanation of the embodiment of the lane traffic state estimation method based on a connected vehicle is also applicable to the lane traffic state estimation device based on a connected vehicle in this embodiment, and will not be repeated here.

[0127] Based on the above embodiment, the present invention further provides a terminal, whose principle block diagram can be shown as follows: Figure 8 As shown. The terminal includes a processor, a memory, a network interface and a display screen connected via a device bus. The processor of the terminal is used to provide computing and control capabilities. The memory of the terminal includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating device and a lane traffic state estimation program based on a networked vehicle. The internal memory provides an environment for the operation of the operating device and the lane traffic state estimation program based on a networked vehicle in the non-volatile storage medium. The network interface of the terminal is used to communicate with an external terminal through a network connection. When the lane traffic state estimation program based on a networked vehicle is executed by the processor, the steps of any one of the above-mentioned lane traffic state estimation methods based on a networked vehicle are implemented. The display screen of the terminal may be a liquid crystal display or an electronic ink display.

[0128] Those skilled in the art will understand that Figure 8 The principle block diagram shown in the figure is only a block diagram of a partial structure related to the solution of the present invention, and does not constitute a limitation on the terminal to which the solution of the present invention is applied. The specific terminal may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0129] In one embodiment, a terminal is provided, comprising a memory, a processor, and a connected vehicle-based lane traffic state estimation program stored in the memory and executable on the processor. When the connected vehicle-based lane traffic state estimation program is executed by the processor, the steps of any one of the connected vehicle-based lane traffic state estimation methods provided in an embodiment of the present invention are implemented.

[0130] An embodiment of the present invention also provides a computer-readable storage medium, on which a lane traffic state estimation program based on a connected vehicle is stored. When the lane traffic state estimation program based on a connected vehicle is executed by a processor, the steps of any one of the lane traffic state estimation methods based on a connected vehicle provided in an embodiment of the present invention are implemented.

[0131] It should be understood that the sequence numbers of the steps in the above embodiments do not imply a specific order of execution; the order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0132] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the above-mentioned device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention. The specific working process of the units and modules in the above-mentioned device can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0133] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0134] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0135] In the embodiments provided by the present invention, it should be understood that the disclosed apparatus / terminal device and method can be implemented in other ways. For example, the apparatus / terminal device embodiments described above are merely illustrative. For example, the division of the modules or units described above is merely a logical functional division. In actual implementation, other division methods may be used. For example, multiple units or components may be combined or integrated into another device, or some features may be omitted or not implemented.

[0136] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, it should be understood by those skilled in the art that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the scope of protection of the present invention.

Claims

1. A lane traffic state estimation method based on a connected vehicle, characterized in that: The method comprises: Constructing a spatiotemporal grid for the target lane and setting a sensing range for each connected vehicle in the target lane within the spatiotemporal grid; Obtaining driving data reported by each connected vehicle in the target lane, the driving data including the connected vehicle's own driving data and / or driving data collected by the connected vehicle from non-connected vehicles within its sensing range; Based on the driving data and the spatiotemporal grid, obtaining second driving data with a data status mark, wherein the data status includes valid and missing; Inputting the second driving data and a preset noise vector into a generator of a trained generative adversarial model, and outputting completed third driving data after processing; estimating a lane traffic state based on the third driving data; The training steps of the generative adversarial model include: Obtaining a training dataset, wherein the training dataset includes a plurality of training subsets corresponding to preset connected vehicle penetration rates, each training subset including training driving data marked with a training data status, and benchmark driving data corresponding to each connected vehicle penetration rate, the benchmark driving data being derived from driving data collected by a traffic monitoring platform and connected vehicles, and the training data status including valid and missing; Initializing a generative adversarial model, wherein the generative adversarial model includes a generator and a discriminator; For each connected vehicle penetration rate, the generator and discriminator are trained for a preset number of rounds; When the training of all connected vehicle penetration rates is completed, the data completion effect of the generative adversarial model is evaluated. If it passes the evaluation, the trained generative adversarial model is obtained; For each connected vehicle penetration rate, the generator and discriminator are trained for a preset number of rounds, including: Fixed the discriminator parameters and trained the generator for the first preset round. In each round, the noise training vector and the training driving data marked with the training data status at the current connected vehicle penetration rate were input into the generator. The training driving data with the missing training data status were supplemented and the supplemented driving data were output. The completed driving data and the baseline driving data of the current connected vehicle penetration rate are input into the discriminator, the generator loss is calculated and backpropagation is used to update the generator parameters; Fix the generator parameters, train the discriminator for the second preset round, input the completed driving data of the current connected vehicle penetration rate and the baseline driving data into the discriminator, calculate the discriminator loss and backpropagate to update the discriminator parameters.

2. The lane traffic state estimation method based on a connected vehicle according to claim 1, characterized in that: Construct the spatiotemporal grid of the target lane, including: Obtain lane geometry data of the target lane; A spatiotemporal grid of the target lane is generated based on the lane geometry data and preset temporal resolution and spatial resolution.

3. The method for estimating lane traffic status based on a connected vehicle according to claim 1, characterized in that: Obtaining second driving data with a data status mark based on the driving data and the spatiotemporal grid includes: After performing data preprocessing on the driving data, preprocessed driving data is obtained; Generate a mask matrix marking the data state of each grid according to the coverage state of each pre-processed driving data in the space-time grid; Based on the mask matrix and all pre-processed driving data, second driving data with a data status mark is obtained.

4. The method for estimating lane traffic status based on a connected vehicle according to claim 3, characterized in that: According to the coverage state of each pre-processed driving data in the spatiotemporal grid, a mask matrix marking the data state of each grid is generated, including: Mapping each preprocessed driving data into a corresponding grid of the space-time grid; For each grid, determine whether it is covered by the pre-processed driving data, and determine the data status of the grid according to the determination result; Generates a mask matrix based on the data state of all rasters.

5. The method for estimating lane traffic status based on a connected vehicle according to claim 4, characterized in that: For each grid, determine whether it is covered by the pre-processed driving data, and determine the data status of the grid based on the judgment result, including: For each grid, it is determined whether there is any driving data of the connected vehicle. If so, the data status of the grid is determined to be valid. If the grid does not contain any connected vehicle's own driving data, then determine whether the grid is within the sensing range of any connected vehicle; If the grid is not within the sensing range of any connected vehicle, the data status of the grid is determined to be missing; If the grid is within the sensing range of any connected vehicle, its status is determined to be valid or missing based on whether there is driving data of a non-connected vehicle.

6. A lane traffic state estimation device based on a connected vehicle, characterized in that: include: A grid construction module is used to construct a spatiotemporal grid for the target lane and set a sensing range for each connected vehicle in the target lane within the spatiotemporal grid; A data acquisition module is configured to acquire driving data reported by each connected vehicle in the target lane, wherein the driving data includes the connected vehicle's own driving data and / or driving data collected by the connected vehicle from non-connected vehicles within its sensing range; A second driving data generating module is configured to obtain second driving data with a data status mark based on the driving data and the spatiotemporal grid, wherein the data status includes valid and missing; a data completion module, configured to input the second driving data and a preset noise vector into a generator of a trained generative adversarial model, and output completed third driving data after processing; The training steps of the generative adversarial model include: Obtaining a training dataset, wherein the training dataset includes a plurality of training subsets corresponding to preset connected vehicle penetration rates, each training subset including training driving data marked with a training data status, and benchmark driving data corresponding to each connected vehicle penetration rate, the benchmark driving data being derived from driving data collected by a traffic monitoring platform and connected vehicles, and the training data status including valid and missing; Initializing a generative adversarial model, wherein the generative adversarial model includes a generator and a discriminator; For each connected vehicle penetration rate, the generator and discriminator are trained for a preset number of rounds; When the training of all connected vehicle penetration rates is completed, the data completion effect of the generative adversarial model is evaluated. If it passes the evaluation, the trained generative adversarial model is obtained; For each connected vehicle penetration rate, the generator and discriminator are trained for a preset number of rounds, including: Fixed the discriminator parameters and trained the generator for the first preset round. In each round, the noise training vector and the training driving data marked with the training data status at the current connected vehicle penetration rate were input into the generator. The training driving data with the missing training data status were supplemented and the supplemented driving data were output. The completed driving data and the baseline driving data of the current connected vehicle penetration rate are input into the discriminator, the generator loss is calculated and backpropagation is used to update the generator parameters; Fixed the generator parameters and trained the discriminator for the second preset round. The completed driving data of the current connected vehicle penetration rate and the baseline driving data were input into the discriminator. The discriminator loss was calculated and backpropagated to update the discriminator parameters. A state estimation module is used to estimate the lane traffic state based on the third driving data.

7. A terminal, characterized in that: The terminal includes a memory, a processor, and a lane traffic state estimation program based on a connected vehicle stored in the memory and runnable on the processor. When the lane traffic state estimation program based on a connected vehicle is executed by the processor, the steps of the lane traffic state estimation method based on a connected vehicle are implemented.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a lane traffic state estimation program based on a connected vehicle. When the lane traffic state estimation program based on a connected vehicle is executed by a processor, the steps of the lane traffic state estimation method based on a connected vehicle as described in any one of claims 1 to 5 are implemented.

Images