Intelligent traffic flow management system and method for unmanned vehicle

Abstract

The invention belongs to the technical field of unmanned driving, and particularly relates to an intelligent traffic flow management system and method for an unmanned vehicle, and the system comprises the steps: firstly collecting environment data and traffic flow data around the vehicle, and carrying out the preprocessing; then obstacle detection and recognition are carried out, feature index detection and recognition are carried out on the traffic flow data, the probability that a current vehicle passes through an obstacle is predicted through an output result by using a prediction model, and the future traffic flow dynamic state is predicted; then, calling an adaptive algorithm to calculate a dynamic passing threshold value of the target vehicle according to the obstacle passing probability of the vehicle, an obstacle detection result and a traffic index recognition result; whether the dynamic passing threshold value of the target vehicle exceeds a preset passing threshold value or not is judged, and if yes, an early warning command is generated and transmitted to a background; and if not, dynamically generating the dynamic behavior of the target vehicle according to the predicted future traffic flow. According to the invention, the problem that the unmanned driving technology in the prior art is inaccurate in evaluation of the function of influencing the passage by the obstacle can be solved.

Description

Technical Field

[0001] The present invention belongs to the field of unmanned driving technology, and in particular relates to an intelligent traffic flow management system and method for unmanned vehicles. Background Art

[0002] With the rapid development of artificial intelligence and automation technology, driverless vehicles are gradually becoming an important part of the modern transportation system. Driverless vehicles refer to vehicles that can autonomously complete tasks such as navigation, driving, and parking without the need for direct operation by human drivers. They use advanced sensors (such as lidar, cameras, radar, and ultrasonic sensors), GPS positioning, map data, and complex algorithms to perceive the surrounding environment and make decisions. There are currently many types of driverless vehicles on the market, including but not limited to driverless food delivery vehicles, driverless express delivery vehicles, and driverless taxis.

[0003] Despite the remarkable progress in driverless technology, there are still some challenges for these vehicles when facing complex and changing real-world traffic conditions. First, driverless vehicles may not be smart or flexible enough to respond to certain obstacles, such as:

[0004] 1. When a small obstacle (such as a bag) appears in front of the vehicle, the driverless vehicle may take overly conservative actions, such as stopping immediately, because it cannot accurately assess its impact on traffic, which may cause unnecessary traffic disruptions;

[0005] 2. When encountering a roadblock while attempting to make a turn (such as a left turn), the driverless vehicle may choose to wait where it is instead of looking for an alternative path, causing local traffic congestion;

[0006] 3. When a detour is required (e.g. a road is partially closed due to illegal parking), the driverless vehicle may block the entire road due to excessive caution and courtesy to other vehicles;

[0007] 4. Under severe weather conditions (such as heavy rain), driverless vehicles may find it difficult to correctly judge the depth of water and its impact on safe driving, and sometimes even risk moving forward, increasing potential risks.

[0008] The reason for the above problems is that the existing unmanned driving technology is not accurate enough in assessing the impact of obstacles on traffic functions, and lacks effective integration with real-time traffic information. Summary of the invention

[0009] The technical problem solved by the present invention is to provide an intelligent traffic flow management system and method for unmanned vehicles, so as to solve the problem of inaccurate assessment of the impact of obstacles on the traffic function by unmanned driving technology in the prior art.

[0010] The basic solution provided by the present invention is an intelligent traffic flow management system for unmanned vehicles, comprising:

[0011] A data acquisition module is used to collect environmental data around the vehicle and the vehicle's current traffic flow data;

[0012] A preprocessing module, used to preprocess the collected environmental data and traffic flow data to generate preprocessed environmental data and preprocessed traffic flow data respectively;

[0013] The target detection module is used to detect and identify obstacles on the pre-processed environmental data according to a preset target detection algorithm and output obstacle detection results; and to detect and identify characteristic indicators on the pre-processed traffic flow data and output traffic indicator identification results;

[0014] The traffic prediction module is used to use the prediction model to predict the probability of the current vehicle passing the obstacle based on the obstacle detection results, and to predict the future traffic flow dynamics based on the traffic indicator recognition results;

[0015] The pass threshold calculation module calls the adaptive algorithm to calculate the dynamic pass threshold of the target vehicle based on the predicted probability of the current vehicle passing the obstacle, the obstacle detection result, and the traffic index recognition result;

[0016] The traffic judgment module is used to judge whether the dynamic traffic threshold of the target vehicle exceeds the preset traffic threshold. If so, a warning command is generated; if not, the dynamic behavior of the target vehicle is generated according to the predicted future traffic flow dynamics;

[0017] The execution control module is used to generate control commands according to the dynamic behavior of the target vehicle and transmit them to the control system of the target vehicle, and transmit warning commands to the control background of the target vehicle.

[0018] Further, the data acquisition module includes a vehicle environment data acquisition unit and a traffic monitoring unit, and the vehicle environment data acquisition unit is used to collect environmental data around the vehicle through a sensor unit and a camera unit deployed on the vehicle body;

[0019] The traffic monitoring unit is used to generate traffic flow data of the vehicle's current road by receiving real-time traffic data transmitted by the traffic monitoring equipment.

[0020] Furthermore, the traffic threshold calculation module includes a threshold setting unit and a threshold adjustment unit, wherein the threshold setting unit is used to use the traffic threshold that meets the preset requirements during the training of the prediction model in the traffic prediction module as the preliminary preset traffic threshold;

[0021] The threshold adjustment unit is used to dynamically adjust the initially preset pass threshold according to the predicted probability of the current vehicle passing an obstacle, the obstacle detection result, and the traffic index recognition result to generate a dynamic pass threshold.

[0022] Furthermore, the threshold adjustment unit dynamically adjusts the initially preset pass threshold according to the predicted probability of the current vehicle passing an obstacle, the obstacle detection result, and the traffic index recognition result, and generates a dynamic pass threshold specifically as follows:

[0023] Obtain the currently predicted probability of the current vehicle passing through an obstacle, obstacle detection results, and assign risk scores to different types of obstacles;

[0024] Retrieving environmental parameters from the traffic index identification results of the target vehicle's passage area, quantifying and weighting the environmental parameters, generating a target vehicle environmental complexity value, and mapping the generated target vehicle environmental complexity value to a preset risk score range to generate a target vehicle environmental complexity risk score;

[0025] Based on the obstacle risk score and the target vehicle environment complexity risk score, the initially preset passage threshold is dynamically adjusted according to the preset adaptive threshold algorithm.

[0026] Furthermore, the currently predicted probability of the current vehicle passing an obstacle and the obstacle detection result are obtained, and risk scores are assigned to different types of obstacles as follows:

[0027] Obtain obstacle detection results and preset risk level ranges for different types of obstacles;

[0028] Assigning a passing probability threshold of different types of obstacles from the current probability of the vehicle passing through the obstacle;

[0029] The pass probability thresholds of different types of obstacles are mapped to the preset risk level ranges of different types of obstacles to generate risk scores of different types of obstacles. The expression is:

[0030] P barrier = L min +[G×(L max -L min )]

[0031] Among them, P obstacle represents the obstacle risk score, L min Indicates the lowest risk level of the preset risk level range of the obstacle, L max It represents the highest risk level in the preset risk level range of the obstacle, and G represents the obstacle passage probability threshold.

[0032] Furthermore, the environmental parameters in the traffic index identification result of the target vehicle passage area are retrieved, the environmental parameters are quantified and weighted, a target vehicle environmental complexity value is generated, and the generated target vehicle environmental complexity value is mapped to a preset risk score range, and the target vehicle environmental complexity risk score is generated as follows:

[0033] Obtaining environmental parameters in the traffic index recognition result within the target vehicle passage area, wherein the environmental parameters include vehicle flow, road conditions, and route complexity;

[0034] Quantify the traffic volume, road conditions and route complexity, and assign corresponding weights to the quantified traffic volume, road conditions and route complexity according to the preset parameter importance;

[0035] The environment score range is preset, and the corresponding scores of the weighted traffic volume, road conditions, and route complexity are calculated to generate the environment complexity value. The calculation formula is:

[0036] H ring = w 1 ×P car+w 2 ×D road + w 3 ×L Road

[0037] Among them, H ring represents the environment complexity value, P car represents the traffic flow score, and w 1 represents the traffic flow parameter weight, Ddao represents the road condition score, and w 2 represents the weight of the road condition parameter, L represents the route complexity score, and w 3 Indicates the weight of route complexity parameter;

[0038] The risk score range is preset, and the environmental complexity value is mapped into the preset risk score range to generate the target vehicle environmental complexity risk score. The expression is:

[0039] P ring = (H ring - H ring -min ) / (H ring -max -H Ring -min )×(S max -S min )+S min

[0040] Among them, the P ring represents the environmental complexity risk score, the H ring represents the environmental complexity value, and the H ring -min Indicates the minimum value of the preset environmental rating range, H ring -max Indicates the maximum value of the preset environment score range, S max Indicates the maximum value of the preset risk score range, S min Indicates the minimum value of the preset risk score range.

[0041] Furthermore, the expression for dynamically adjusting the initially preset passage threshold according to the preset adaptive threshold algorithm by combining the obstacle risk score and the target vehicle environment complexity risk score is:

[0042] Z=Z 0 +w 4 ×P barrier + w 5 ×P Ring

[0043] Among them, Z represents the dynamic pass threshold, Z 0 represents the initial preset pass threshold, w 4 represents the weight of obstacle risk score, w 5 Indicates the weight of the environmental complexity risk score.

[0044] Further, the passage judgment module includes a judgment unit, a warning trigger unit and a vehicle action adjustment unit, and the judgment unit is used to judge whether the dynamic passage threshold of the target vehicle exceeds the preset passage threshold and output the judgment result;

[0045] The warning trigger unit is used to trigger the preset warning instruction to generate a warning command when the dynamic passage threshold of the target vehicle exceeds the preset passage threshold in the judgment result;

[0046] The vehicle action adjustment unit is used to retrieve the predicted future traffic flow dynamics and analyze and generate the dynamic behavior of the target vehicle when the dynamic passage threshold of the target vehicle in the judgment result does not exceed the preset passage threshold.

[0047] An intelligent traffic flow management method for unmanned vehicles, applied to the above-mentioned intelligent traffic flow management system for unmanned vehicles, comprises:

[0048] S1: Collect environmental data around the vehicle and the current traffic flow data of the vehicle, and perform preprocessing to generate preprocessed environmental data and preprocessed traffic flow data;

[0049] S2: calling the preset target detection algorithm to perform obstacle detection and identification on the preprocessed environmental data, and to perform feature index detection and identification on the preprocessed traffic flow data, and outputting obstacle detection results and traffic index identification results respectively;

[0050] S3: Call the prediction model to predict the probability of the current vehicle passing the obstacle based on the obstacle detection results, and predict the future traffic flow dynamics based on the traffic indicator recognition results;

[0051] S4: calling the adaptive algorithm to calculate the dynamic passage threshold of the target vehicle according to the predicted probability of the current vehicle passing the obstacle, the obstacle detection result, and the traffic index recognition result;

[0052] S5: Determine whether the dynamic passing threshold of the target vehicle exceeds the preset passing threshold. If not, generate a warning command; if so, generate the dynamic behavior of the target vehicle according to the predicted future traffic flow dynamics;

[0053] S6: Transmitting the dynamic behavior generation control command of the target vehicle to the control system of the target vehicle for execution, and transmitting the warning command to the control background of the target vehicle.

[0054] The principle and advantage of the present invention are as follows: in the present invention, firstly, environmental data around the vehicle and current traffic flow data are collected. For environmental data, it is possible to detect and identify whether there are obstacles near the vehicle. For traffic flow data, it is possible to predict the traffic flow dynamics near the current vehicle. For the identified obstacles, the passing probability is predicted by the prediction model. At this time, the initial passing probability is not sufficient to meet the traffic risk assessment of the real vehicle environment. For this, an adaptive algorithm is adopted to comprehensively consider the environmental parameters of the target vehicle, the obstacle detection results and the initial passing probability, and generate a dynamic passing threshold to meet the real passing threshold of the target vehicle under the influence of the outside world under real operation. When the dynamic passing threshold does not meet the preset passing threshold, the running state of the vehicle is controlled in combination with the traffic flow dynamics, and the passing instruction of the vehicle in the future is generated. When the dynamic passing threshold is met, an early warning command is generated and transmitted to the background. The staff at the background can remotely control the vehicle according to the content of the early warning command transmitted back, so that the vehicle can better deal with the situation where it cannot pass through the obstacle by itself.

[0055] During the entire process described above, the principle of the dynamically adjusted passage threshold set in the present application is upward feedback. By combining the vehicle passing probability and the dynamic prediction of traffic flow, different operation instructions under different states are implemented, making the intelligent passage function of unmanned vehicles more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 is a functional block diagram of an embodiment of the present invention;

[0057] Figure 2 The figure is a flowchart of an embodiment of the present invention. DETAILED DESCRIPTION

[0058] The following is further described in detail through specific implementation methods:

[0059] With the development of intelligence, more and more driverless vehicles are appearing on the market, such as driverless food delivery vehicles, driverless express delivery vehicles, driverless taxis, etc. Among these driverless vehicles, when faced with complex and changeable traffic conditions, it is often impossible to accurately assess the actual impact of obstacles on traffic. For example, when a roadblock appears at a turning intersection, driverless vehicles often wait on the spot with their lights and double flashes on; another example is when encountering illegally parked vehicles and need to detour, but the road is congested because of giving way to oncoming vehicles; there are many such situations because driverless vehicles cannot accurately judge road obstacles and lack real-time information on traffic and road conditions in various regions. Under the influence of these two factors, there are deficiencies in the intelligent passage of driverless vehicles.

[0060] In this regard, the technical solution established by this application is an intelligent traffic flow management system for unmanned vehicles, such as Figure 1 As shown, it is composed of a data acquisition module, a preprocessing module, a target detection module, a traffic prediction module, a traffic judgment module and an execution control module, wherein the data acquisition module is used to collect environmental data around the vehicle and the current traffic flow data of the vehicle; specifically, the data acquisition module is composed of a vehicle environmental data acquisition unit and a traffic monitoring unit, and the vehicle environmental data unit collects environmental data around the vehicle through sensor units and camera units deployed on the vehicle body, such as lidar, ultrasonic radar, millimeter wave radar, camera, etc., through the combination of radar sensor units and camera units, and adjusts the installation position and angle of each radar sensor unit and camera unit according to the actual situation, so as to effectively capture image data and radar data on the road, and obtain environmental data around the target vehicle.

[0061] The traffic monitoring unit generates traffic flow data of the vehicle's current road by receiving real-time traffic data transmitted by traffic monitoring equipment. In this embodiment, the traffic monitoring unit receives real-time traffic data transmitted back by traffic monitoring cameras, sensors and other equipment through wireless transmission, such as traffic volume, vehicle speed, road conditions, etc., so as to obtain traffic flow data in real time.

[0062] The preprocessing module is used to preprocess the collected environmental data and traffic flow data to generate preprocessed environmental data and preprocessed traffic flow data respectively. In this embodiment, the environmental data is mainly preprocessing the captured image data, including color hole conversion, converting the RGB image into a grayscale image; it also includes using Gaussian filtering to remove noise in the image, and using histogram equalization to improve the contrast of the image.

[0063] The preprocessing of traffic flow data involves data cleaning, data supplementation and data conversion of the collected original real-time traffic data, so as to ensure the accuracy and completeness of the traffic flow data.

[0064] The pre-processed environmental data and traffic flow data are then transmitted to the target detection module, which performs obstacle detection and identification on the pre-processed environmental data according to a preset target detection algorithm and outputs obstacle detection results; and performs feature index detection and identification on the pre-processed traffic flow data and outputs traffic index identification results; specifically, for the target detection algorithm used in obstacle detection and identification of the pre-processed environmental data, this application adopts the FasterR-CNN model, and its operating principle is as follows:

[0065] First, collect image datasets containing different obstacles. The collected image datasets about obstacles should cover various scenes and conditions, such as obstacles under different weather and lighting conditions. The collected images are annotated to mark the location and category of the obstacles in each image.

[0066] Then, the image dataset is preprocessed, including resizing, cropping, color enhancement, and standardization. Data augmentation techniques are used, including random cropping, rotation, flipping, and color transformation, to generate a more diverse image training dataset, which helps improve the robustness and generalization ability of the FasterR-CNN model in the object detection algorithm.

[0067] Then the FasterR-CNN model initializes the weights and inputs the generated diverse image training data set into the FasterR-CNN model for training. During the training process, the weights are continuously adjusted through the back-propagation algorithm to optimize the loss function to improve the accuracy of obstacle recognition and classification, and the learning rate, batch size, number of iterations and other hyperparameters are adjusted to optimize the training effect of the model.

[0068] Finally, the reserved validation set is input into the trained model to evaluate the performance of the model, including accuracy, recall rate, F1-score, etc., and analyze the performance of the model on the validation set, identify and solve possible problems with the model, and adjust the model architecture, data processing flow or training strategy based on the evaluation results and feedback from error analysis to further improve model performance.

[0069] For the detection and identification of characteristic indicators of pre-processed traffic flow data, this application adopts a CNN-RNN hybrid model to analyze and model the pre-processed traffic flow data to obtain the traffic flow indicators and road condition indicators of the current traffic. The obtained current characteristic indicators of the traffic are helpful for predicting the subsequent traffic flow dynamics.

[0070] The traffic prediction module uses the prediction model to predict the probability of the current vehicle passing the obstacle based on the obstacle detection results. Specifically, the prediction process of the vehicle passing the obstacle probability is as follows:

[0071] First, actual data of the vehicle passing through obstacles is collected, including vehicle speed, obstacle type (such as pedestrians, vehicles, roadblocks, etc.), obstacle location, size, shape and other information;

[0072] Then, the corresponding features are extracted from the collected data, such as distance features (the distance between the vehicle and the obstacle), speed features (the speed and acceleration of the vehicle), obstacle attribute features (the type, size, shape, etc. of the obstacle), and environmental features (weather conditions, road conditions, etc.); and the feature data is preprocessed, including missing value processing and outlier processing in the feature data, and non-numeric features are converted into numerical features using one-hot encoding or label encoding, and finally the training set and test set are generated;

[0073] Then, a prediction model is selected, such as a linear regression model, a logistic regression model, or a random forest model. The training set is input into the model for training. The model parameters are tuned during the training process. After the training, the model is tested with a test set to evaluate whether the various indicators of the model meet the standards, such as mean square error, mean absolute error, accuracy, recall rate, F1-score, etc.

[0074] Finally, based on the training results of different models, the model with the best performance is selected to predict the obstacle detection results collected by the vehicle in real time to obtain the probability of the vehicle passing through the obstacle.

[0075] At the same time, the traffic prediction module also predicts future traffic flow dynamics based on the traffic indicator identification results. The CNN-RNN hybrid model is used in this application to predict future traffic flow dynamics, such as road congestion, smoothness, etc. The CNN-RNN hybrid model used in this application, the CNN part is responsible for processing the image or video data transmitted back by the traffic monitoring equipment and extracting high-level features from it; the RNN part receives the feature representation from the CNN or processes the time series data, and predicts the future traffic status based on these features and historical traffic data.

[0076] For the above-mentioned vehicle passing obstacle probability result, the pass threshold calculation module then calls the adaptive algorithm to calculate the dynamic pass threshold of the target vehicle according to the predicted current vehicle passing obstacle probability, obstacle detection result, and traffic index recognition result. Specifically, the pass threshold calculation module includes a threshold setting unit and a threshold adjustment unit. The threshold setting unit is used to use the pass threshold that meets the preset requirements during the training process of the prediction model in the pass prediction module as the preliminary preset pass threshold; for example, select a threshold that balances the accuracy and recall rate of the prediction model, and use the threshold to test in a simulation environment, evaluate the performance of the threshold in the simulation environment, and take the one with the least false alarms or missed alarms as the best performance to obtain the preliminary preset pass threshold;

[0077] However, the initially preset pass threshold is not applicable to independent target vehicles in actual use. In this regard, the present application uses a threshold adjustment unit to dynamically adjust the initially preset pass threshold according to the predicted probability of the current vehicle passing an obstacle, the obstacle detection result, and the traffic index recognition result to generate a dynamic pass threshold. In this embodiment, the threshold adjustment unit implements the following steps:

[0078] Step 1: Obtain the currently predicted probability of the current vehicle passing through an obstacle and the obstacle detection result, and assign risk scores to different types of obstacles. Specifically, first obtain the obstacle detection result and preset the risk level range of different types of obstacles. For example, the risk level range of obstacle A is set to 1-5, and the risk level range of obstacle B is set to 6-10. Then, assign the pass probability thresholds of different types of obstacles from the current probability of the vehicle passing through the obstacle. For example, the pass probability of obstacle A is considered high risk if it is lower than 0.3, medium risk if it is 0.3-0.7, and low risk if it is higher than 0.7. Finally, map the pass probability thresholds of different types of obstacles to the preset risk level range of different types of obstacles to generate risk scores of different types of obstacles. The expression is:

[0079] P barrier = L min +[G×(L max -L min )]

[0080] Among them, P obstacle represents the obstacle risk score, L min Indicates the lowest risk level of the preset risk level range of the obstacle, L max It represents the highest risk level in the preset risk level range of the obstacle, and G represents the obstacle passage probability threshold.

[0081] For example, if the probability of passing obstacle A is 0.4 and the risk level ranges from 1 to 5, then:

[0082] P barrier -A =1+[0.4*(5-1)]=2.6

[0083] Step 2: retrieve the environmental parameters in the traffic index identification results of the target vehicle passage area, quantify and weight the environmental parameters, generate the target vehicle environmental complexity value, and map the generated target vehicle environmental complexity value to the preset risk score range to generate the target vehicle environmental complexity risk score; specifically:

[0084] Obtaining environmental parameters in the traffic index recognition result within the target vehicle passage area, wherein the environmental parameters include traffic volume, road conditions (congestion level, road flatness) and route complexity (number of intersections, turning angles);

[0085] The traffic volume, road condition and route complexity are quantified, and corresponding weights are assigned to the quantified traffic volume, road condition and route complexity according to the preset parameter importance; wherein the traffic volume is quantified as the number of vehicles passing through a certain section of road per hour, the road condition is evaluated by the road maintenance status and weather, and the route complexity is evaluated by the number of intersections and the degree of tortuosity of the road; then, weights are assigned according to the importance of each parameter, and the sum of the weights of the three indicators in this embodiment is 1;

[0086] Preset environment scoring range. In this embodiment, the scoring range of environmental parameters is consistent. For example, in the vehicle flow scoring range, the grades are divided based on the number of vehicles per hour, with a vehicle flow of less than 100 vehicles as 1 point, 100-500 as 5 points, and more than 500 as 10 points; in the road condition scoring range, based on the road surface flatness and maintenance status, the high-quality road is 1 point, the ordinary is 5 points, and the poor is 10 points; in the route complexity scoring range, the route is scored according to the intersection and the degree of tortuosity, with a few intersections and a straight path as 1 point, and many intersections and tortuosity as 10 points;

[0087] Then the corresponding scores of the weighted traffic volume, road conditions and route complexity are calculated to generate the environmental complexity value. The calculation formula is:

[0088] H ring = w 1 ×P car+w 2 ×D road + w 3 ×L Road

[0089] Among them, H ring represents the environment complexity value, P car represents the traffic flow score, and w 1 represents the traffic flow parameter weight, Ddao represents the road condition score, and w 2 represents the weight of the road condition parameter, L represents the route complexity score, and w 3 Indicates the weight of route complexity parameter;

[0090] The risk score range is preset. In this embodiment, the risk score range is 1-10, 1 represents the lowest risk, and 10 represents the highest risk. The environment complexity value is mapped to the preset risk score range to generate the target vehicle environment complexity risk score, which is expressed as:

[0091] P ring = (H ring - H ring -min ) / (H ring -max -H ring -min )×(S max -S min )+S min

[0092] Among them, the P ring represents the environmental complexity risk score, the H ring represents the environmental complexity value, and the H ring -min Indicates the minimum value of the preset environmental rating range, H ring-max Indicates the maximum value of the preset environment score range, S max Indicates the maximum value of the preset risk score range, S min Indicates the minimum value of the preset risk score range.

[0093] For example, the minimum value of environment complexity is 0 and the maximum value is 100, the minimum value of risk score is 1 and the maximum value is 10. If the calculated environment complexity is 75, then:

[0094]

[0095] Combined with the obstacle risk score and the target vehicle environment complexity risk score, the initially preset pass threshold is dynamically adjusted according to the preset adaptive threshold algorithm. The expression is:

[0096] Z=Z 0 +w 4 ×P barrier + w 5 ×P Ring

[0097] Among them, Z represents the dynamic pass threshold, Z 0 represents the initial preset pass threshold, w 4 represents the weight of obstacle risk score, w 5 Indicates the weight of the environmental complexity risk score.

[0098] After calculating the dynamic pass threshold of the target vehicle, the pass judgment module is used to judge whether the dynamic pass threshold of the target vehicle exceeds the preset pass threshold. If so, a warning command is generated; if not, the dynamic behavior of the target vehicle is dynamically generated according to the predicted future traffic flow; wherein the pass judgment module includes a judgment unit, a warning trigger unit and a vehicle action adjustment unit, and the judgment unit is used to judge whether the dynamic pass threshold of the target vehicle exceeds the preset pass threshold, and output the judgment result;

[0099] The warning trigger unit is used to trigger the preset warning instruction to generate a warning command when the dynamic passage threshold of the target vehicle exceeds the preset passage threshold in the judgment result;

[0100] The vehicle action adjustment unit is used to retrieve the predicted future traffic flow dynamics and analyze and generate the dynamic behavior of the target vehicle when the dynamic passage threshold of the target vehicle in the judgment result does not exceed the preset passage threshold. In this embodiment, the warning level is set according to different ranges of the dynamic passage threshold, including:

[0101] Normal (low risk): the dynamic pass threshold is lower than the preset pass threshold, for example, 10;

[0102] Warning (medium risk): The dynamic pass threshold is between two preset pass thresholds, such as 10-15;

[0103] Danger (high risk): The dynamic pass threshold is higher than the preset pass threshold, such as 15;

[0104] Then determine at which warning level the specific dynamic threshold is located. For example, the preset initial threshold is 5, the obstacle risk score calculation result is 6, the environmental complexity risk score is 4, and the weight coefficient w 4 is 0.4, w 5 is 0.6, then the calculated adaptive threshold is:

[0105] Z=5+0.4*6+0.6*4=9.8

[0106] Therefore, the calculated dynamic threshold is 9.8, and the warning level is normal; specifically, in each warning level, when:

[0107] If the dynamic traffic threshold is ≤10, the target vehicle system continues to operate normally, and displays relevant information such as "normal status" or "no warning required". At this time, the target vehicle's driving path and driving status are generated based on the predicted future traffic flow dynamics;

[0108] If 10 < dynamic traffic threshold < 15, a warning will be triggered, displaying "Warning: Risk Increased", and a corresponding warning command will be generated; under this warning command, the backend staff will decide whether manual intervention is needed;

[0109] If the dynamic pass threshold is greater than 15, a high-risk warning is triggered, the message "Danger: Manual action required" is displayed, and a corresponding warning command is generated.

[0110] The execution control module generates control commands based on the dynamic behavior of the target vehicle and transmits them to the control system of the target vehicle, and transmits warning commands to the control background of the target vehicle. Under different warning commands, different decisions are made to achieve the purpose of dynamically adjusting the operating status of the target vehicle.

[0111] like Figure 2 As shown, in another embodiment of this embodiment, an intelligent traffic flow management method for unmanned vehicles is also included, which is applied to the above-mentioned intelligent traffic flow management system for unmanned vehicles, including:

[0112] S1: Collect environmental data around the vehicle and the current traffic flow data of the vehicle, and perform preprocessing to generate preprocessed environmental data and preprocessed traffic flow data;

[0113] S2: calling the preset target detection algorithm to perform obstacle detection and identification on the preprocessed environmental data, and to perform feature index detection and identification on the preprocessed traffic flow data, and outputting obstacle detection results and traffic index identification results respectively;

[0114] S3: Call the prediction model to predict the probability of the current vehicle passing the obstacle based on the obstacle detection results, and predict the future traffic flow dynamics based on the traffic indicator recognition results;

[0115] S4: calling the adaptive algorithm to calculate the dynamic passage threshold of the target vehicle according to the predicted probability of the current vehicle passing the obstacle, the obstacle detection result, and the traffic index recognition result;

[0116] S5: Determine whether the dynamic passing threshold of the target vehicle exceeds the preset passing threshold. If so, generate a warning command; if not, generate the dynamic behavior of the target vehicle according to the predicted future traffic flow dynamics;

[0117] S6: Transmitting the dynamic behavior generation control command of the target vehicle to the control system of the target vehicle for execution, and transmitting the warning command to the control background of the target vehicle.

[0118] The above are only embodiments of the present invention. Common knowledge such as the known specific structures and characteristics in the scheme is not described in detail here. Ordinary technicians in the relevant field know all the common technical knowledge in the technical field to which the invention belongs before the application date or priority date, can obtain all the existing technologies in the field, and have the ability to apply conventional experimental means before that date. Ordinary technicians in the relevant field can improve and implement this scheme in combination with their own abilities under the enlightenment given by this application. Some typical known structures or known methods should not become obstacles for ordinary technicians in the relevant field to implement this application. It should be pointed out that for those skilled in the art, without departing from the structure of the present invention, several deformations and improvements can be made, which should also be regarded as the scope of protection of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. An intelligent traffic flow management system for unmanned vehicles, characterized in that: include: A data acquisition module is used to collect environmental data around the vehicle and the vehicle's current traffic flow data; A preprocessing module, used to preprocess the collected environmental data and traffic flow data to generate preprocessed environmental data and preprocessed traffic flow data respectively; The target detection module is used to detect and identify obstacles on the pre-processed environmental data according to a preset target detection algorithm and output the obstacle detection results; and detecting and identifying characteristic indicators of the pre-processed traffic flow data, and outputting traffic indicator identification results; The traffic prediction module is used to use the prediction model to predict the probability of the current vehicle passing the obstacle based on the obstacle detection results, and to predict the future traffic flow dynamics based on the traffic indicator recognition results; The pass threshold calculation module calls the adaptive algorithm to calculate the dynamic pass threshold of the target vehicle based on the predicted probability of the current vehicle passing the obstacle, the obstacle detection result, and the traffic index recognition result; A traffic judgment module is used to judge whether the dynamic traffic threshold of the target vehicle exceeds the preset traffic threshold, and if so, generate a warning command; If not, the dynamic behavior of the target vehicle is generated based on the predicted future traffic flow dynamics; The execution control module is used to generate control commands according to the dynamic behavior of the target vehicle and transmit them to the control system of the target vehicle, and transmit warning commands to the control background of the target vehicle.

2. The intelligent traffic flow management system for unmanned vehicles according to claim 1, characterized in that: The data acquisition module includes a vehicle environment data acquisition unit and a traffic monitoring unit, wherein the vehicle environment data acquisition unit is used to collect environmental data around the vehicle through a sensor unit and a camera unit deployed on the vehicle body; The traffic monitoring unit is used to generate traffic flow data of the vehicle's current road by receiving real-time traffic data transmitted by the traffic monitoring equipment.

3. The intelligent traffic flow management system for unmanned vehicles according to claim 2, characterized in that: The traffic threshold calculation module includes a threshold setting unit and a threshold adjustment unit, wherein the threshold setting unit is used to use the traffic threshold that meets the preset requirements during the prediction model training process in the traffic prediction module as the preliminary preset traffic threshold; The threshold adjustment unit is used to dynamically adjust the initially preset pass threshold according to the predicted probability of the current vehicle passing an obstacle, the obstacle detection result, and the traffic index recognition result to generate a dynamic pass threshold.

4. The intelligent traffic flow management system for unmanned vehicles according to claim 3, characterized in that: The threshold adjustment unit dynamically adjusts the initially preset pass threshold according to the predicted probability of the current vehicle passing an obstacle, the obstacle detection result, and the traffic index recognition result, and generates a dynamic pass threshold specifically as follows: Obtain the currently predicted probability of the current vehicle passing through an obstacle, obstacle detection results, and assign risk scores to different types of obstacles; Retrieving environmental parameters from the traffic index identification results of the target vehicle's passage area, quantifying and weighting the environmental parameters, generating a target vehicle environmental complexity value, and mapping the generated target vehicle environmental complexity value to a preset risk score range to generate a target vehicle environmental complexity risk score; Based on the obstacle risk score and the target vehicle environment complexity risk score, the initially preset passage threshold is dynamically adjusted according to the preset adaptive threshold algorithm.

5. The intelligent traffic flow management system for unmanned vehicles according to claim 4, characterized in that: The method of obtaining the currently predicted probability of the current vehicle passing an obstacle and the obstacle detection result, and assigning risk scores to different types of obstacles is specifically as follows: Obtain obstacle detection results and preset risk level ranges for different types of obstacles; Assigning a passing probability threshold of different types of obstacles from the current probability of the vehicle passing through the obstacle; The pass probability thresholds of different types of obstacles are mapped to the preset risk level ranges of different types of obstacles to generate risk scores of different types of obstacles. The expression is: P 障 =L min +[G×(L max -L min )] Among them, P 障 represents the obstacle risk score, L min Indicates the lowest risk level of the preset risk level range of the obstacle, L max It represents the highest risk level in the preset risk level range of the obstacle, and G represents the obstacle passage probability threshold.

6. The intelligent traffic flow management system for unmanned vehicles according to claim 5, characterized in that: The environmental parameters in the traffic index identification result of the target vehicle passage area are retrieved, the environmental parameters are quantified and weighted, a target vehicle environmental complexity value is generated, and the generated target vehicle environmental complexity value is mapped to a preset risk score range to generate a target vehicle environmental complexity risk score: Obtaining environmental parameters in the traffic index recognition result within the target vehicle passage area, wherein the environmental parameters include vehicle flow, road conditions, and route complexity; Quantify the traffic volume, road conditions and route complexity, and assign corresponding weights to the quantified traffic volume, road conditions and route complexity according to the preset parameter importance; The environment score range is preset, and the corresponding scores of the weighted traffic volume, road conditions, and route complexity are calculated to generate the environment complexity value. The calculation formula is: H 环 =w1×P 车 +w2×D 道 +w3×L 路 Among them, H 环 Represents the environmental complexity value, P 车 represents the traffic flow score, w1 represents the traffic flow parameter weight, D 道 represents the road condition score, w2 represents the road condition parameter weight, L 路 represents the route complexity score, w3 represents the route complexity parameter weight; The risk score range is preset, and the environment complexity value is mapped into the preset risk score range to generate the target vehicle environment complexity risk score. The expression is: P 环 =(H 环 -H 环-min ) / (H 环-max -H 环-min )×(S max -S min )+S min Among them, P 环 represents the environmental complexity risk score, H 环 Represents the environmental complexity value, H 环-min Indicates the minimum value of the preset environmental rating range, H 环-max Indicates the maximum value of the preset environment score range, S max Indicates the maximum value of the preset risk score range, S min Indicates the minimum value of the preset risk score range.

7. The intelligent traffic flow management system for unmanned vehicles according to claim 6, characterized in that: The expression for dynamically adjusting the initially preset pass threshold according to the preset adaptive threshold algorithm by combining the obstacle risk score and the target vehicle environment complexity risk score is: Z=Z0+w4×P 障 +w5×P 环 Among them, Z represents the dynamic pass threshold, Z0 represents the preliminary preset pass threshold, w4 represents the weight of the obstacle risk score, and w5 represents the weight of the environmental complexity risk score.

8. The intelligent traffic flow management system for unmanned vehicles according to claim 7, characterized in that: The passage judgment module includes a judgment unit, a warning trigger unit and a vehicle action adjustment unit. The judgment unit is used to judge whether the dynamic passage threshold of the target vehicle exceeds the preset passage threshold and output the judgment result; The warning trigger unit is used to trigger the preset warning instruction to generate a warning command when the dynamic passage threshold of the target vehicle exceeds the preset passage threshold in the judgment result; The vehicle action adjustment unit is used to retrieve the predicted future traffic flow dynamics and analyze and generate the dynamic behavior of the target vehicle when the dynamic passage threshold of the target vehicle in the judgment result does not exceed the preset passage threshold.

9. An intelligent traffic flow management method for unmanned vehicles, applied to an intelligent traffic flow management system for unmanned vehicles as described in any one of claims 1 to 8 above, characterized in that: include: S1: Collect environmental data around the vehicle and the current traffic flow data of the vehicle, and perform preprocessing to generate preprocessed environmental data and preprocessed traffic flow data; S2: calling the preset target detection algorithm to perform obstacle detection and identification on the preprocessed environmental data, and to perform feature index detection and identification on the preprocessed traffic flow data, and outputting obstacle detection results and traffic index identification results respectively; S3: Call the prediction model to predict the probability of the current vehicle passing the obstacle based on the obstacle detection results, and predict the future traffic flow dynamics based on the traffic indicator recognition results; S4: calling the adaptive algorithm to calculate the dynamic passage threshold of the target vehicle according to the predicted probability of the current vehicle passing the obstacle, the obstacle detection result, and the traffic index recognition result; S5: determining whether the dynamic passage threshold of the target vehicle exceeds the preset passage threshold, and if so, generating a warning command; If not, the dynamic behavior of the target vehicle is generated based on the predicted future traffic flow dynamics; S6: Transmitting the dynamic behavior generation control command of the target vehicle to the control system of the target vehicle for execution, and transmitting the warning command to the control background of the target vehicle.

Images