High-speed vehicle cut-in intention prediction method and system, storage medium and vehicle

Abstract

The application discloses a high-speed vehicle cut-in intention prediction method and system, a storage medium and a vehicle, relates to the technical field of vehicle behavior prediction, and comprises the following steps: dividing a target road in front of a vehicle into regions to obtain a straight-ahead area, a lane-changing area and a lane-changing buffer area; collecting vehicle dynamic data of at least one target vehicle located in the straight-ahead area, the lane-changing area and the lane-changing buffer area in real time; constructing a vehicle cut-in intention prediction model based on a preset long short-term memory network, and training and verifying the vehicle cut-in intention prediction model; and inputting the vehicle dynamic data of the target vehicle into the vehicle cut-in intention prediction model to output vehicle intention data. The method disclosed by the application is simple in sensor equipment, can realize the prediction of the cut-in intention of a high-speed vehicle by analyzing relevant data obtained through existing road section speed detection and a self-vehicle sensor, and can effectively reduce vehicle manufacturing cost.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vehicle behavior prediction, in particular to a high-speed driving vehicle cut-in intention prediction method, system, storage medium and vehicle. BACKGROUND

[0002] With the continuous development of automatic driving technology, it is particularly important to judge the cut-in intention and behavior of vehicles during high-speed driving. Most existing prediction methods are based on parameters such as vehicle motion state, speed, acceleration, etc., but in complex traffic environment, the prediction accuracy and timeliness need to be improved.

[0003] The existing intelligent driving solution relies on various sensors when facing the vehicle cut-in problem, but the sensors have the problem of too late detection time when facing high-speed cut-in scenes, which may not be enough to deal with sudden cut-in vehicles, thereby increasing the risk of accidents. At the same time, due to the lack of intelligent processing mechanism, the decision system cannot accurately analyze and predict the scene of sudden high-speed cut-in of vehicles. This makes vehicles often only take simple avoidance or braking measures when dealing with sudden situations such as high-speed cut-in, and cannot realize higher-level coordination and decision-making. Modern traffic technology through V2V (vehicle-to-vehicle communication) and V2I (vehicle-to-infrastructure cooperation) technologies, vehicles can exchange information and cooperate with other vehicles and infrastructure around them, but there are still problems such as high equipment prices, huge upfront investment, and the need to maintain software and hardware systems and network security.

[0004] Therefore, the current scheme for detecting the cut-in intention of high-speed driving vehicles still has the problem of high cost, which significantly increases the vehicle manufacturing cost. SUMMARY

[0005] In view of the deficiencies of the prior art, the purpose of the present application is to provide a high-speed driving vehicle cut-in intention prediction method, system, storage medium and vehicle, which aims to solve the problem that the current scheme for detecting the cut-in intention of high-speed driving vehicles still has high cost, which significantly increases the vehicle manufacturing cost.

[0006] The first aspect of the present application is to provide a high-speed driving vehicle cut-in intention prediction method, which comprises:

[0007] During the driving of the vehicle, the target road in front of the vehicle is regionally divided to obtain a straight-ahead area, a lane-changing area and a lane-changing buffer area;

[0008] The vehicle dynamic data of at least one target vehicle located in the straight-ahead area, the lane-changing area and the lane-changing buffer area are collected in real time by the speed sensor pre-set on the target road and the camera and radar device pre-set on the vehicle;

[0009] A vehicle cut-in intention prediction model is constructed based on a preset long short-term memory network, and the vehicle cut-in intention prediction model is trained and verified.

[0010] Vehicle dynamic data of the target vehicle is input into the vehicle cut-in intention prediction model, and vehicle intention data is output.

[0011] According to an aspect of the above technical solution, in the process of vehicle driving, the target road in front of the vehicle is regionally divided to obtain the straight-ahead area, the lane-changing area, and the lane-changing buffer area, and the steps include:

[0012] The road identification of the target road on which the vehicle is driving is obtained, and the road grade condition of the target road is determined according to the road identification;

[0013] The current traffic flow data of the target road and the current speed of the vehicle are obtained.

[0014] The target road in front of the vehicle is divided into a straight-ahead area located directly in front of the vehicle, and a lane-changing area and a lane-changing buffer area connected in turn with the straight-ahead area.

[0015] According to an aspect of the above technical solution, the vehicle dynamic data of at least one target vehicle located in the straight-ahead area, the lane-changing area, and the lane-changing buffer area is collected in real time through the speed sensor preset on the target road and the camera and radar device preset on the vehicle, and the steps include:

[0016] The driving position, driving speed, and acceleration of at least one target vehicle located in the straight-ahead area, the lane-changing area, and the lane-changing buffer area are collected in real time through the speed sensor preset on both sides of the vehicle road or above the road and the camera and radar device preset on the vehicle;

[0017] The driving position, driving speed, and acceleration are marked according to a unified timestamp, and data cleaning and filtering are performed to obtain the vehicle dynamic data of the target vehicle;

[0018] The acceleration includes longitudinal acceleration and transverse acceleration.

[0019] According to an aspect of the above technical solution, the vehicle dynamic data of the target vehicle is input into the vehicle cut-in intention prediction model, and vehicle intention data is output, and the steps include:

[0020] The driving position, driving speed, and acceleration in the vehicle dynamic data are sequentially input into the vehicle cut-in intention prediction model;

[0021] The vehicle dynamic data is analyzed by the vehicle cut-in intention prediction model, and vehicle intention data is output.

[0022] According to an aspect of the above technical solution, the step of analyzing the vehicle dynamic data by the vehicle cut-in intention prediction model and outputting vehicle intention data comprises:

[0023] The driving position, driving speed and acceleration of the target vehicle collected in real time are sequentially introduced into the pre-trained vehicle cut-in intention prediction model;

[0024] The probability distribution of various driving intentions of the target vehicle is calculated and output, and vehicle intention data is obtained.

[0025] According to an aspect of the above technical solution, the probability distribution of various driving intentions of the target vehicle includes cutting into the lane, maintaining straight driving, accelerating and decelerating.

[0026] The second aspect of the application is to provide a high-speed vehicle cut-in intention prediction system, which is applied to the method in the above technical solution, and the system comprises:

[0027] The region division module is used for dividing the target road in front of the vehicle into a straight driving area, a lane changing area and a lane changing buffer area during vehicle driving;

[0028] The data acquisition module is used for acquiring vehicle dynamic data of at least one target vehicle located in the straight driving area, the lane changing area and the lane changing buffer area in real time through the speed sensor preset on the target road and the camera and radar device preset on the vehicle;

[0029] The model construction module is used for constructing a vehicle cut-in intention prediction model based on a preset long short-term memory network, and training and verifying the vehicle cut-in intention prediction model;

[0030] The intention output module is used for introducing the vehicle dynamic data of the target vehicle into the vehicle cut-in intention prediction model and outputting vehicle intention data.

[0031] According to an aspect of the above technical solution, the region division module is specifically used for:

[0032] Obtaining the road identification of the target road on which the vehicle is driving, and determining the road grade condition of the target road according to the road identification;

[0033] Obtaining the current traffic flow data of the target road and the current speed of the vehicle;

[0034] The target road in front of the vehicle is divided into a straight driving area located in front of the vehicle, and a lane changing area and a lane changing buffer area connected with the straight driving area in sequence.

[0035] The third aspect of the present application provides a readable storage medium, which stores computer instructions, and the instructions are executed by a processor to implement the method described in the above technical solutions.

[0036] The fourth aspect of the present application provides a vehicle, which comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the method described in the above technical solutions when executing the program.

[0037] Compared with the prior art, the high-speed driving vehicle cut-in intention prediction method, system, storage medium and vehicle provided by the present application have the beneficial effects that:

[0038] The present application can timely detect and track the high-speed cut-in vehicle in multiple regions by dividing the target road in front of the vehicle into straight-ahead region, lane-changing region and lane-changing buffer region during the driving of the vehicle, and then constructing a prediction model of the driving region of the surrounding vehicle through a long short-term memory network (LSTM), which can understand the interactive behavior between vehicles and respond to changes in the traffic environment, so that the driving region of the vehicle can be predicted in real time in the actual scene, and the vehicle intention data of the target vehicle can be output to the decision module of the automatic driving system to provide real-time and accurate driving region prediction information for the vehicle, and finally the cut-in intention of the target vehicle can be timely judged. The sensor equipment used in the method of the present application is simple, and the cut-in intention of the high-speed vehicle can be predicted by analyzing the relevant data obtained by the existing road speed detection and the self-vehicle sensor, which can effectively reduce the cost of the vehicle. BRIEF DESCRIPTION OF DRAWINGS

[0039] The above and / or additional aspects and advantages of the present application will become apparent and more readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings, in which:

[0040] Figure 1 A flowchart of a high-speed driving vehicle cut-in intention prediction method in an embodiment of the present application is shown in the figure.

[0041] Figure 2 A structural block diagram of a high-speed driving vehicle cut-in intention prediction system in an embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0042] In order to make the objectives, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the accompanying drawings. However, the present application can be realized in many different forms, and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present application more thorough and comprehensive.

[0043] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in the description herein is for describing particular embodiments only and is not intended to be limiting of the application. As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items.

[0044] Embodiment one

[0045] Referring to Figure 1 The first embodiment of the present application provides a high-speed vehicle cut-in intention prediction method, the method comprising steps S10-S40:

[0046] Step S10, in the process of vehicle driving, the target road in front of the vehicle is regionally divided to obtain a straight area, a lane-changing area and a lane-changing buffer area.

[0047] In this embodiment, the vehicle driving process is a high-speed driving process, that is, the driving process on urban expressways and highways, and then the target road currently driven is detected by the intelligent driving sensor carried on the vehicle to obtain video information, radar information, etc., and then the target road in front of the vehicle is regionally divided based on the video information, radar information, etc., specifically, the target road in front of the vehicle is divided into a straight area of the vehicle and a lane-changing area and a lane-changing buffer area of the target vehicle.

[0048] It should be noted that the lane-changing area and the lane-changing buffer area are road areas that are sequentially connected with the straight area, the lane-changing area is the road area in front of the vehicle lane and the adjacent lane, which usually refers to the area on both sides of the lane line, and the lane-changing buffer area mainly refers to the ramp area of the merging injection.

[0049] In this embodiment, the step of regionally dividing the target road in front of the vehicle to obtain the straight area, the lane-changing area and the lane-changing buffer area comprises:

[0050] Obtaining the road identification of the target road driven by the vehicle, and determining the road grade condition of the target road according to the road identification;

[0051] Obtaining the current traffic flow data of the target road and the current speed of the vehicle;

[0052] Dividing the target road in front of the vehicle into a straight area located in front of the vehicle, and a lane-changing area and a lane-changing buffer area sequentially connected with the straight area.

[0053] The road signs on the target road include road guide signs and speed limit signs, and the road level of the target road is determined through the road guide signs and the speed limit signs, such as a highway or an urban expressway, so that the road level condition of the target road can be determined to finally determine the required vehicle spacing and speed limit of the target road.

[0054] In step S20, the vehicle dynamic data of at least one target vehicle located in the straight area, the lane changing area and the lane changing buffer area is collected in real time through the speed sensor preset on the target road and the camera and radar device preset on the vehicle.

[0055] In this embodiment, the step of collecting the vehicle dynamic data of at least one target vehicle located in the straight area, the lane changing area and the lane changing buffer area in real time through the speed sensor preset on the target road and the camera and radar device preset on the vehicle includes:

[0056] The driving position, driving speed and acceleration of at least one target vehicle located in the straight area, the lane changing area and the lane changing buffer area are collected in real time through the speed sensor preset on the two sides of the vehicle road or above the road and the camera and radar device preset on the vehicle;

[0057] The driving position, driving speed and acceleration are marked according to a unified timestamp, and data cleaning and filtering are performed to obtain the vehicle dynamic data of the target vehicle;

[0058] The acceleration includes longitudinal acceleration and transverse acceleration.

[0059] Specifically, the driving speed and acceleration of at least one target vehicle located in the straight area or the lane changing buffer area or the lane changing area are obtained through the speed sensor preset on the two sides of the vehicle road or above the road, the acceleration includes longitudinal acceleration and transverse acceleration, and the driving position of the target vehicle located in the straight area or the lane changing buffer area or the lane changing area is obtained in real time through the camera and radar device preset on the vehicle, so that the relative position and distance between the target vehicle and the ego vehicle are determined based on the driving position of the target vehicle and the driving position of the ego vehicle.

[0060] After obtaining the driving position, driving speed and acceleration of the target vehicle, the driving position, driving speed and acceleration are marked according to a unified timestamp, which is convenient for subsequent time series analysis. The data format of the driving position, driving speed and acceleration can be a structured table or a JSON object with timestamp+position(x, y, z), speed and acceleration.

[0061] In step S30, a vehicle cut-in intention prediction model is constructed based on a preset long short-term memory network, and the vehicle cut-in intention prediction model is trained and verified.

[0062] In this embodiment, the steps of building a vehicle cut-in intention prediction model and training and verifying include:

[0063] 1. Data Preparation

[0064] Step 1.1: Data Collection

[0065] Collect dynamic data of numerous vehicles from vehicle sensors, V2X systems, or historical trajectory data, including but not limited to driving location (GPS coordinates), speed, acceleration, etc.

[0066] Labeling data: For the training set, it is necessary to determine which data points represent vehicle cut-in events through artificial labeling or rule-based algorithms.

[0067] Step 1.2: Data Preprocessing

[0068] Clean data: remove outliers, fill in missing values.

[0069] Feature engineering: extract and construct features useful for prediction, such as speed change rate, acceleration change rate, lane information, surrounding vehicle information, etc.

[0070] Data normalization: scale feature values to an appropriate range to speed up model training and improve performance.

[0071] Time series division: divide continuous data points into time series segments, each containing a fixed number of time steps, for input to the LSTM model.

[0072] 2. LSTM Model Construction

[0073] Step 2.1: Select Framework

[0074] Use deep learning frameworks such as TensorFlow, PyTorch, or Keras to build the LSTM model.

[0075] Step 2.2: Define Model Structure

[0076] Input layer: accepts preprocessed time series data.

[0077] LSTM layer: one or more LSTM layers to capture long-term dependencies in time series.

[0078] Fully connected layer (Dense layer): converts the output of the LSTM layer to the final intention prediction.

[0079] Output layer: usually uses a softmax activation function to output the probability distribution of each intention.

[0080] Step 2.3: Set Hyperparameters

[0081] Number of units, number of layers in LSTM layer.

[0082] Optimizer (e.g., Adam, RMSprop).

[0083] Loss function (e.g., cross-entropy loss).

[0084] Learning rate, batch size, number of iterations, etc.

[0085] 3. Model Training

[0086] Step 3.1: Splitting Dataset

[0087] Divide the dataset into training, validation, and test sets (common ratio: 70%, 15%, 15%).

[0088] Step 3.2: Training Model

[0089] Train the LSTM model using the training set data, optimizing model parameters through the backpropagation algorithm.

[0090] Monitor loss values and accuracy during training to assess the learning progress of the model.

[0091] Step 3.3: Early Stopping (Optional)

[0092] If performance on the validation set does not improve after consecutive iterations, stop training to avoid overfitting.

[0093] 4. Model Validation and Evaluation

[0094] Step 4.1: Validating Model

[0095] Evaluate the model's performance using the validation set data, calculating accuracy, recall, F1 score, and other metrics.

[0096] Confusion Matrix: Analyze the prediction accuracy of the model on different intents.

[0097] Step 4.2: Adjusting Model

[0098] Adjust model structure, hyperparameters, or feature selection based on validation results to improve performance.

[0099] Repeat the training and validation steps until the model performance reaches a satisfactory level.

[0100] 5. Model Testing and Deployment

[0101] Step 5.1: Testing Model

[0102] Evaluate the final performance of the model using the test set data, ensuring that the model performs well on unseen data.

[0103] Step 5.2: Model Deployment

[0104] Integrate the trained model into the vehicle control system or vehicle networking platform for real-time vehicle cut-in intention prediction.

[0105] Monitor the performance of the model in actual application and fine-tune or update as needed.

[0106] 6. Continuous Optimization

[0107] Collect more data, especially data containing more cut-in events and complex scenarios, to further train and optimize the model.

[0108] Explore other deep learning techniques such as attention mechanism, Transformer, etc. to improve the prediction ability of the model.

[0109] Evaluate the computational efficiency and resource requirements of the model and optimize the model to adapt to different hardware platforms.

[0110] Step S40, import the vehicle dynamic data of the target vehicle into the vehicle cut-in intention prediction model, and output vehicle intention data.

[0111] In this embodiment, the step of importing the vehicle dynamic data of the target vehicle into the vehicle cut-in intention prediction model and outputting vehicle intention data includes:

[0112] Import the driving position, driving speed and acceleration in the vehicle dynamic data into the vehicle cut-in intention prediction model in sequence;

[0113] Analyze the vehicle dynamic data through the vehicle cut-in intention prediction model and output vehicle intention data.

[0114] The step of analyzing the vehicle dynamic data through the vehicle cut-in intention prediction model and outputting vehicle intention data includes:

[0115] Import the driving position, driving speed and acceleration of the target vehicle collected in real time into the pre-trained vehicle cut-in intention prediction model in sequence;

[0116] Calculate and output the probability distribution of various driving intentions of the target vehicle, and obtain vehicle intention data.

[0117] The probability distribution of various driving intentions of the target vehicle includes cut-in the lane, straight driving, acceleration and deceleration.

[0118] Specifically, the vehicle dynamic data (the driving position, speed and acceleration of the target vehicle) of the target vehicle collected in real time is introduced into the trained vehicle cut-in intention prediction model, the model will perform logical reasoning and analysis according to the input data, and output vehicle intention data. Among them, the vehicle intention data usually includes whether the target vehicle intends to cut into the adjacent lane, the timing and speed of cutting in, etc.

[0119] Then, according to the vehicle intention data output by the model, the vehicle control system can make corresponding responses, such as decelerating in advance, adjusting the lane, etc.

[0120] At the same time, the prediction result is fed back to the driver or the Internet of Vehicles system to improve the driving safety and efficiency.

[0121] In summary, compared with the prior art, the high-speed driving vehicle cut-in intention prediction method shown in the embodiment has the beneficial effects that:

[0122] The embodiment can timely detect and track the high-speed cut-in vehicles in multiple regions by dividing the target road in front of the vehicle into straight-ahead region, lane-changing region and lane-changing buffer region during vehicle driving, and then constructing a prediction model of the driving region of surrounding vehicles through a long short-term memory network (LSTM). The LSTM model can understand the interactive behavior between vehicles, respond to changes in the traffic environment, so that the driving region of the vehicle can be predicted in real time in the actual scene, and the vehicle intention data of the target vehicle can be output to the decision module of the autonomous driving system to provide real-time and accurate driving region prediction information for the vehicle, and finally the cut-in intention of the target vehicle can be timely judged. The sensor equipment used in the method shown in the embodiment is simple, and only the existing road speed detection and self-vehicle sensor can be used to analyze the relevant data to predict the cut-in intention of the high-speed vehicle, which can effectively reduce the cost of the vehicle.

[0123] Embodiment two

[0124] Please refer to Figure 2 The second embodiment of the present application provides a high-speed driving vehicle cut-in intention prediction system applied to the method described in the above embodiment, the system comprises:

[0125] The region division module 10 is used for dividing the target road in front of the vehicle into straight-ahead region, lane-changing region and lane-changing buffer region during vehicle driving;

[0126] The data acquisition module 20 is used for acquiring the vehicle dynamic data of at least one target vehicle in the straight-ahead region, lane-changing region and lane-changing buffer region through the speed sensor preset on the target road and the camera and radar device preset on the vehicle in real time;

[0127] The model construction module 30 is configured to construct a vehicle cut-in intention prediction model based on a preset long short-term memory network, and train and verify the vehicle cut-in intention prediction model.

[0128] The intention output module 40 is configured to input the vehicle dynamic data of the target vehicle into the vehicle cut-in intention prediction model, and output vehicle intention data.

[0129] The area division module 10 is specifically configured to:

[0130] The road identification of a target road on which the vehicle travels is acquired, and the road grade condition of the target road is determined according to the road identification;

[0131] The current traffic flow data of the target road and the current speed of the vehicle are acquired;

[0132] The target road in front of the vehicle is divided into a straight-ahead area located in front of the vehicle, and a lane-changing area and a lane-changing buffer area connected in sequence with the straight-ahead area.

[0133] Compared with the prior art, the high-speed vehicle cut-in intention prediction system has the beneficial effects that:

[0134] The embodiment can timely detect and track the high-speed cut-in vehicles in multiple areas by dividing the target road in front of the vehicle into a straight-ahead area, a lane-changing area and a lane-changing buffer area during the vehicle travel, and then constructing a prediction model of the vehicle travel area around the vehicle by using a long short-term memory network (LSTM). The LSTM model can understand the interactive behavior between vehicles and respond to changes in the traffic environment, so that the vehicle travel area can be predicted in real time in the actual scene, and the vehicle intention data of the target vehicle can be output to the decision module of the automatic driving system to provide real-time and accurate travel area prediction information for the vehicle, and finally the cut-in intention of the target vehicle can be timely judged. The system has simple sensor equipment, and can realize the prediction of the cut-in intention of the high-speed vehicle by analyzing the relevant data obtained by the existing road speed detection and the self-vehicle sensor, thereby effectively reducing the cost of the vehicle.

[0135] Embodiment Three

[0136] The third embodiment of the present application provides a readable storage medium having computer instructions stored thereon, and the instructions are executed by a processor to implement the method described in the above embodiments.

[0137] Embodiment Four

[0138] A fourth aspect of the application provides a vehicle comprising a memory, a processor, and a computer program stored on the memory and loadable on the processor, the processor implementing the method described in the above embodiments when executing the program.

[0139] In the description of the specification, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the application. In the specification, the illustrative description of the above terms does not necessarily mean the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0140] The above-described embodiments only express several implementation manners of the application, which are described in a more specific and detailed manner, but cannot be understood as the limitation of the patent scope of the application. It should be noted that, for those skilled in the art, several modifications and improvements can be made without departing from the concept of the application, which are all within the protection scope of the application. Therefore, the protection scope of the patent of the application should be subject to the appended claims.

Claims

1. A method for predicting the cutting intention of a high-speed vehicle, characterized in that, The method includes: During the vehicle's journey, the target road ahead is divided into zones, resulting in a straight-ahead zone, a lane-changing zone, and a lane-changing buffer zone. By using speed sensors pre-installed on the target road, as well as cameras and radar devices pre-installed on the vehicle, the vehicle dynamic data of at least one target vehicle located in the straight-ahead area, lane-changing area, and lane-changing buffer zone is collected in real time. A vehicle entry intention prediction model is constructed based on a pre-defined long short-term memory network, and the vehicle entry intention prediction model is trained and validated. The vehicle dynamic data of the target vehicle is imported into the vehicle entry intention prediction model, and the vehicle intention data is output. The steps involved in dividing the target road ahead of the vehicle into zones, including a straight-ahead zone, a lane-changing zone, and a lane-changing buffer zone, during vehicle operation include: Obtain the road signs of the target road the vehicle is traveling on, and determine the road grade conditions of the target road based on the road signs; Obtain the current traffic flow data of the target road, and the current speed of the vehicles; The target road in front of the vehicle is divided into a straight-ahead area directly in front of the vehicle, and a lane-changing area and a lane-changing buffer zone that are sequentially connected to the straight-ahead area. The step of collecting real-time vehicle dynamic data of at least one target vehicle located in the straight-ahead area, lane-changing area, and lane-changing buffer zone using speed sensors pre-installed on the target road and cameras and radar devices pre-installed on the vehicle includes: By using speed sensors pre-installed on both sides or above the road, as well as cameras and radar devices pre-installed on the vehicle, the driving position, speed and acceleration of at least one target vehicle located in the straight-ahead area, lane-changing area and lane-changing buffer zone are collected in real time. The driving position, driving speed and acceleration are marked with a unified timestamp, and the data is cleaned and filtered to obtain the vehicle dynamic data of the target vehicle. The acceleration includes longitudinal acceleration and lateral acceleration.

2. The method for predicting the cutting intention of a high-speed vehicle according to claim 1, characterized in that, The steps of importing the vehicle dynamic data of the target vehicle into the vehicle entry intent prediction model and outputting vehicle intent data include: The vehicle's dynamic data, including its driving position, speed, and acceleration, are sequentially imported into the vehicle's cutting intention prediction model. The vehicle dynamic data is analyzed by the vehicle entry intention prediction model to output vehicle intention data.

3. The method for predicting the cutting intention of a high-speed vehicle according to claim 2, characterized in that, The steps of analyzing the vehicle dynamic data using the vehicle entry intention prediction model and outputting vehicle intention data include: The real-time collected driving position, driving speed and acceleration of the target vehicle are sequentially imported into the pre-trained vehicle entry intention prediction model; Calculate and output the probability distribution of various driving intentions of the target vehicle to obtain vehicle intention data.

4. The method for predicting the cutting intention of a high-speed vehicle according to claim 3, characterized in that, The probability distribution of various driving intentions of the target vehicle, including cutting into the current lane, maintaining straight driving, accelerating, and decelerating.

5. A system for predicting the cutting intention of a high-speed vehicle, characterized in that, The system, applicable to the method of any one of claims 1-4, comprises: The area division module is used to divide the target road in front of the vehicle into areas during the vehicle's movement, resulting in a straight-through area, a lane-changing area, and a lane-changing buffer zone. The data acquisition module is used to collect real-time vehicle dynamic data of at least one target vehicle located in the straight-ahead area, lane-changing area and lane-changing buffer zone by using speed sensors pre-installed on the target road and cameras and radar devices pre-installed on the vehicle. The model building module is used to build a vehicle entry intention prediction model based on a preset long short-term memory network, and to train and verify the vehicle entry intention prediction model. The intent output module is used to import the vehicle dynamic data of the target vehicle into the vehicle entry intent prediction model and output vehicle intent data.

6. The high-speed vehicle cutting-in intention prediction system according to claim 5, characterized in that, The region division module is specifically used for: Obtain the road signs of the target road the vehicle is traveling on, and determine the road grade conditions of the target road based on the road signs; Obtain the current traffic flow data of the target road, and the current speed of the vehicles; The target road in front of the vehicle is divided into a straight-ahead area directly in front of the vehicle, and a lane-changing area and a lane-changing buffer zone that are sequentially connected to the straight-ahead area.

7. A readable storage medium having computer instructions stored thereon, characterized in that, When executed by the processor, this instruction implements the method as described in any one of claims 1-4.

8. A vehicle comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1-4.

Images