A method, device and storage medium for extracting and processing driving interaction intention fragments

Abstract

The present application relates to the field of electronic digital processing technology, and specifically, to a method, device and storage medium for extracting and processing driving interaction intention fragments. The method for extracting and processing driving interaction intention fragments includes: during the driving process of the own vehicle, collecting the driving behavior data of the own vehicle and the interaction behavior data between the own vehicle and surrounding vehicles; the surrounding vehicle is any vehicle around the own vehicle; dividing the driving behavior data to obtain a plurality of own vehicle behavior fragments; dividing the interaction behavior data according to the division time point of the own vehicle behavior fragments to obtain a plurality of interaction intention fragments; each of the interaction intention fragments includes the driving behavior data of the own vehicle and the interaction behavior data between the own vehicle and surrounding vehicles; clustering the multiple interaction intention fragments, and performing intent recognition on each type of interaction intention fragment set. The present application can improve the accuracy of the division and recognition of interaction intention fragments.

Description

Technical Field

[0001] The present application relates to the field of electronic digital processing technology, and in particular to a method, device and storage medium for extracting and processing driving interaction intention fragments. Background Art

[0002] Driving in complex traffic environments is a game of negotiation. The driver and surrounding vehicles interact through this game to achieve a dynamic equilibrium and achieve their respective driving goals. By breaking down the time-series data of vehicle-to-vehicle interactions into multiple, physically meaningful, minimal interaction behaviors or intentions, we can analyze and define the patterns of these interaction segments to identify basic types of vehicle-to-vehicle interactions, such as lane change negotiation, intersection conflict, and car-following adjustments.

[0003] In the existing technology, interaction segments are generally divided based on simple acceleration and deceleration changes and the amplitude of vehicle heading angle changes. This division method has low accuracy. For example, when a vehicle is driving straight normally, it may deviate from the lane slightly. If the heading angle change is used to divide the segments, it will be classified as turning or changing lanes.

[0004] In view of this, this application is filed. Summary of the Invention

[0005] The purpose of this application is to provide a method, device and storage medium for extracting and processing driving interaction intention segments to improve the accuracy of segmentation and recognition of interaction intention segments.

[0006] In order to achieve the above objectives, this application adopts the following technical solutions:

[0007] In a first aspect, the present application provides a method for extracting and processing driving interaction intention fragments, comprising:

[0008] During the driving process of the vehicle, the driving behavior data of the vehicle and the interaction behavior data between the vehicle and surrounding vehicles are collected; the surrounding vehicles are any vehicles around the vehicle;

[0009] Dividing the driving behavior data to obtain a plurality of vehicle behavior segments;

[0010] Divide the interaction behavior data according to the division time points of the vehicle behavior segments to obtain multiple interaction intention segments; each interaction intention segment includes the driving behavior data of the vehicle and the interaction behavior data between the vehicle and surrounding vehicles;

[0011] The multiple interaction intention fragments are clustered, and intent recognition is performed on each type of interaction intention fragment set.

[0012] The driving behavior data is divided into multiple vehicle behavior segments, including:

[0013] Performing a first-level division on the driving behavior data according to a lateral behavior characterization variable of the vehicle to obtain a plurality of preliminary vehicle behavior segments;

[0014] Each preliminary ego-vehicle behavior segment is divided into two levels according to the longitudinal behavior representation variables of the ego-vehicle to obtain multiple ego-vehicle behavior segments.

[0015] Optionally, a Bayesian agglomerative data segmentation algorithm is used to perform primary and secondary segmentation.

[0016] Optionally, before performing the first-level division on the driving behavior data according to the lateral behavior characterization variable of the vehicle, the method further includes:

[0017] Perform correlation analysis on the lateral behavior characterization variables of the vehicle, and retain the lateral behavior characterization variables with correlations below the set threshold;

[0018] A correlation analysis is performed on the longitudinal behavior characterization variables of the vehicle, and the longitudinal behavior characterization variables with correlations lower than the set threshold are retained.

[0019] Optionally, before clustering the multiple interaction intent fragments to obtain interaction intent categories, the method further includes:

[0020] From multiple interaction intent segments, select a target segment whose duration is less than a set threshold;

[0021] The maximum a posteriori estimation is used to solve the regression model coefficients of each variable in the target segment and its adjacent segments before and after;

[0022] According to the similarity of the regression model coefficients of the target segment and its preceding and following adjacent segments, the target segment is merged with the adjacent interaction intention segments.

[0023] Optionally, perform intent recognition on each set of interaction intent fragments, including:

[0024] Perform numerical statistics on the variables in each type of interaction intention fragment set;

[0025] Intent recognition is performed based on numerical statistical results.

[0026] Optionally, the lateral behavior characterization variables include lateral acceleration and vehicle body heading angle, and the longitudinal behavior characterization variables include longitudinal velocity and longitudinal acceleration.

[0027] In a second aspect, the present application provides an electronic device, comprising:

[0028] at least one processor, and a memory communicatively coupled to the at least one processor;

[0029] In which, the memory stores instructions that can be executed by at least one of the processors, and the instructions are executed by at least one of the processors so that at least one of the processors can execute the above-mentioned driving interaction intention fragment extraction and processing method.

[0030] In a third aspect, the present application provides a computer-readable storage medium having computer instructions stored thereon, wherein the computer instructions are used to enable a computer to execute the above-mentioned method for extracting and processing driving interaction intention fragments.

[0031] Compared with the prior art, the present invention has the following advantages:

[0032] This application leverages the fact that the vehicle's behavioral decisions are the result of the driving interaction process. It segments the interaction behavior data based on the time points of the segmentation, effectively improving segmentation accuracy. Furthermore, multiple interaction intent segments are clustered, and intent recognition is performed on each set of interaction intent segments. Clustering is then used to aggregate interaction intent segments with the same intent but not adjacent in time, enabling the extraction and identification of vehicle-to-vehicle interaction intent. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the specific implementation methods of the present application or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the specific implementation methods or the description of the prior art. Obviously, the drawings described below are some implementation methods of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0034] Figure 1 This is a flowchart of a method for extracting and processing driving interaction intention fragments provided in an embodiment of the present application;

[0035] Figure 2 1 is a schematic diagram of the division results of the driving behavior data of the vehicle provided in an embodiment of the present application;

[0036] Figure 3 It is a structural diagram of the electronic device provided in this application. DETAILED DESCRIPTION

[0037] The following description of exemplary embodiments of the present application is made in conjunction with the accompanying drawings, including various details of the embodiments of the present application to facilitate understanding. These details should be considered as merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present application. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0038] Figure 1This is a flowchart of a method for extracting and processing driving interaction intention segments provided by an embodiment of the present application. The method can be executed by an electronic device. The method provided in this embodiment is applicable to extracting and identifying driving intention segments from driving data generated during vehicle driving and interaction behavior data with surrounding vehicles. Figure 1 , the method provided in this embodiment includes the following operations:

[0039] S110 . During the driving process of the own vehicle, collect driving behavior data of the own vehicle and interaction behavior data between the own vehicle and surrounding vehicles.

[0040] The ego vehicle is the subject of interaction intent analysis, and this embodiment does not limit the type of ego vehicle. Surrounding vehicles are any vehicles around the ego vehicle, such as the preceding or following adjacent vehicles in the same lane, or lateral adjacent vehicles in different lanes. The ego vehicle's driving strategy is influenced by surrounding vehicles. For example, if the preceding adjacent vehicle slows down, the ego vehicle will also slow down to avoid a rear-end collision.

[0041] Both the ego vehicle and surrounding vehicles are equipped with positioning devices, inertial measurement units, gyroscopes, and other sensors to collect driving behavior data of the ego vehicle and the interaction behavior data between the ego vehicle and the surrounding vehicles. The ego vehicle's driving behavior data includes, but is not limited to, lateral behavior characterization variables and longitudinal behavior characterization variables. The lateral behavior characterization variables include lateral acceleration and body yaw angle, and the longitudinal behavior characterization variables include longitudinal velocity and longitudinal acceleration. Similarly, the interaction behavior data includes, but is not limited to, the relative position and relative driving behavior data of the ego vehicle and the surrounding vehicles. The relative driving behavior data includes relative lateral behavior characterization variables and relative longitudinal behavior characterization variables between the ego vehicle and the surrounding vehicles. The relative lateral behavior characterization variables include relative lateral acceleration and relative body yaw angle, and the relative longitudinal behavior characterization variables include relative longitudinal velocity and relative longitudinal acceleration.

[0042] Both driving behavior data and interaction behavior data are time series data.

[0043] S120: Divide the driving behavior data to obtain a plurality of vehicle behavior segments.

[0044] During the driving interaction between the ego vehicle and surrounding vehicles, the ego vehicle makes its own behavioral decisions based on its interaction situation with the surrounding vehicles. Therefore, the ego vehicle's behavioral decisions can be considered the corresponding results of the driving interaction process. To this end, this application uses the segmentation of the ego vehicle's driving behavior data as the basis for segmenting the interaction behavior data between the ego vehicle and surrounding vehicles, and achieves the segmentation of interaction intent segments through the segmentation of the ego vehicle's driving behavior.

[0045] Optionally, the driving behavior data is divided into a first level according to the lateral behavior characterization variable of the own vehicle to obtain multiple preliminary own vehicle behavior segments; and each preliminary own vehicle behavior segment is divided into a second level according to the longitudinal behavior characterization variable of the own vehicle to obtain multiple own vehicle behavior segments.

[0046] Initial ego-vehicle behavior segments are divided based on the variation characteristics of the variables representing lateral behavior. For example, driving behavior data with significantly different vehicle heading angles should be divided into different initial ego-vehicle behavior segments, assuming there are m such segments. For each initial ego-vehicle behavior segment, further segmentation is performed based on the variation characteristics of its longitudinal behavior variables. For example, driving behavior data with significantly different longitudinal accelerations should be divided into different ego-vehicle driving segments, assuming that n ego-vehicle driving segments are ultimately obtained.

[0047] In a preferred embodiment, before the driving behavior data is divided into the first level according to the lateral behavior characterization variables of the vehicle, it also includes: performing a correlation analysis on the lateral behavior characterization variables of the vehicle, and retaining the lateral behavior characterization variables whose correlation is lower than a set threshold; performing a correlation analysis on the longitudinal behavior characterization variables of the vehicle, and retaining the longitudinal behavior characterization variables whose correlation is lower than a set threshold.

[0048] Specifically, assuming that there are multiple horizontal behavior characterization variables, correlation analysis is performed between each of the multiple horizontal behavior characterization variables, such as calculating the Pearson correlation coefficient. If the correlation coefficient / correlation of two horizontal behavior characterization variables is higher than the set threshold, it means that the two horizontal behavior characterization variables are correlated, and only one of them is retained. If the correlation coefficient / correlation of two horizontal behavior characterization variables is lower than the set threshold, both are retained, and the horizontal behavior characterization variables that are finally retained are all variables with correlation coefficients / correlations higher than the set threshold. Similarly, the vertical behavior characterization variables whose correlation is lower than the set threshold are retained, and no further details are given here.

[0049] In a preferred embodiment, a Bayesian agglomerative data segmentation algorithm is used to perform primary and secondary segmentation. The Bayesian model-based agglomerative sequences segmentation algorithm (BMASS) is an unsupervised learning method based on hierarchical clustering and Bayesian inference, which is used to divide time series data or sequence data into statistically significant continuous segments. Its core idea is to use a bottom-up agglomerative merging strategy combined with the Bayesian criterion to evaluate the rationality of the segmentation points and ultimately generate the optimal data partitioning. First, the lateral behavior characterization variables of the vehicle are input into BMASS. BMASS adopts a bottom-up agglomerative strategy. By calculating the marginal likelihood of all lateral behavior characterization variables, it gradually merges adjacent segmentation segments to obtain multiple segments obtained by dividing the lateral behavior characterization variables of the vehicle. The lateral behavior representation variable of the ego vehicle is part of the driving behavior data and has a temporal relationship with other driving behavior data (i.e., longitudinal behavior representation variables). Therefore, when the lateral behavior representation variable is partitioned, the other driving behavior data can be partitioned accordingly based on the time point of the partition, thereby partitioning the complete driving behavior data of the ego vehicle, resulting in multiple preliminary ego vehicle behavior segments. All longitudinal behavior representation variables in each preliminary ego vehicle behavior segment are then input into BMASS. BMASS employs a bottom-up agglomerative strategy, calculating the marginal likelihood of all longitudinal behavior representation variables and gradually merging adjacent segments to obtain multiple segments of the longitudinal behavior representation variable within each preliminary ego vehicle behavior segment. The complete data of each preliminary ego vehicle behavior segment is then partitioned based on the time points of the segmentation, resulting in multiple final ego vehicle behavior segments, each of which includes both lateral and longitudinal behavior representation variables.

[0050] Figure 2 : is a schematic diagram of the division result of the driving behavior data of the vehicle provided in the embodiment of the present application, Figure 2 The horizontal axis is time, and the vertical axis is driving behavior data, including speed, longitudinal acceleration, lateral acceleration and yaw angle. Figure 2 The middle vertical dotted line is the time point at which the driving behavior data is divided. The driving behavior data between two adjacent vertical dotted lines constitute a vehicle behavior segment.

[0051] S130 , dividing the interaction behavior data according to the division time points of the vehicle behavior segments to obtain a plurality of interaction intention segments.

[0052] Assuming the ego vehicle behavior segments are segmented at time points t1, t2, t3, and t4, the interaction behavior data is similarly segmented at t1, t2, t3, and t4. It should be noted that the interaction intention segments here include both the ego vehicle's driving behavior data and the interaction behavior data between the ego vehicle and surrounding vehicles. In other words, the segments generated by segmenting the ego vehicle behavior segments and the interaction behavior data are combined.

[0053] S140: Cluster multiple interaction intention fragments, and perform intent recognition on each set of interaction intention fragments.

[0054] In a preferred embodiment, some interaction intent segments are short-lived, insufficient to reflect driving intent, and therefore need to be merged with preceding or following interaction intent segments. Specifically, a target segment whose duration is less than a set threshold is selected from multiple interaction intent segments; this threshold can be customized, for example, 3 seconds. Maximum a posteriori estimation (MAP) is used to calculate the regression model coefficients for each variable in the target segment and its preceding and following adjacent segments. The target segment is merged with adjacent interaction intent segments based on the similarity of the regression model coefficients between the target segment and its preceding and following adjacent segments.

[0055] MAP is a Bayesian estimation method that assumes that data follows a linear regression model and treats the model parameters as random variables. Parameters are estimated by fusing prior probabilities with likelihood functions. In this embodiment, MAP is used to calculate the Bayesian linear regression model coefficients for each behavior representation variable relative to time in each interaction intent segment (including the target segment and both preceding and following adjacent segments). This ultimately yields a regression model coefficient matrix for each interaction intent segment. Similarity analysis is then performed between the regression model coefficient matrix of the target segment and the regression model coefficient matrices of adjacent segments, for example, by calculating the Euclidean distance between the coefficients. Similar segments are then merged.

[0056] Optionally, clustering methods such as LDA, DTW, and fuzzy k-means can be used to cluster the multiple interaction intent segments. Fuzzy k-means is an extension of k-means that allows data points to belong to multiple clusters based on membership (between 0 and 1), making it suitable for datasets with unclear boundaries. LDA (Latent Dirichlet Allocation) is a latent Dirichlet allocation model, a bag-of-words model that can be used for document clustering. Interaction intent segments consist of multiple multidimensional data points, so they can be clustered as documents. DTW (Dynamic Time Warping) is a dynamic time warping model that calculates the similarity between two or more time series by stretching and compressing time series data. If the similarity between time series exceeds a set threshold, they are clustered into one cluster.

[0057] A clustering algorithm generates multiple sets of interaction intention segments. Each set includes at least one interaction intention segment, each belonging to a specific interaction intention. This embodiment performs numerical statistics on the variables in each set of interaction intention segments, and identifies the intent type based on the statistical results. The statistical method can be mean or variance calculation. Optionally, the statistical method includes calculating the mean, maximum, standard deviation, and entropy of the ego vehicle's driving behavior data (speed, longitudinal acceleration, lateral acceleration, etc.) and the interaction behavior data between the ego vehicle and surrounding vehicles (relative speed, relative longitudinal distance, relative lateral distance, etc.) within each set of interaction intention segments (i.e., each cluster obtained through clustering). Intention refers to the interaction intention between the ego vehicle and surrounding vehicles, including: competitive game between the ego vehicle accelerating to overtake, competitive game between the surrounding vehicle accelerating to approach, negotiated game between the ego vehicle slowing down to yield, and mixed game between the two vehicles. The statistical values ​​of the ego vehicle's driving behavior data and interaction behavior data can be used to analyze the interaction intention between the ego vehicle and surrounding vehicles. For example, if the average longitudinal acceleration of the ego vehicle is high and the relative longitudinal distance is reduced, the ego vehicle will accelerate to overtake the opponent in the competitive game between the two vehicles.

[0058] Through the above steps, this embodiment divides the interaction behaviors between the ego vehicle and surrounding vehicles during driving. Each cluster represents an independent interaction intention type, and the interaction pattern between the ego vehicle and surrounding vehicles in the cluster remains constant.

[0059] This application leverages the fact that the vehicle's behavioral decisions are the result of the driving interaction process and segments the interaction behavior data based on the time points of the segmentation, effectively improving segmentation accuracy. Furthermore, multiple interaction intent segments are clustered, and intent recognition is performed on each set of interaction intent segments. Clustering is then used to aggregate interaction intent segments with the same intent but not adjacent in time, enabling the extraction and identification of vehicle-to-vehicle interaction intent.

[0060] like Figure 3 As shown, this embodiment provides an electronic device, including:

[0061] at least one processor; and

[0062] a memory communicatively connected to at least one of the processors; wherein,

[0063] The memory stores instructions executable by at least one processor, and the instructions are executed by at least one processor to enable at least one processor to perform the above method. The at least one processor in the electronic device is capable of performing the above method, thereby having at least the same advantages as the above method.

[0064] Optionally, the electronic device also includes interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. The various components are connected to each other using different buses and can be installed on a common motherboard or installed in other ways as needed. The processor can process instructions executed in the electronic device, including instructions stored in or on the memory to display graphical information of a GUI (Graphical User Interface) on an external input / output device (such as a display device coupled to the interface). In other embodiments, if necessary, multiple processors can be used with multiple memories, and / or multiple buses can be used with multiple memories. Similarly, multiple electronic devices can be connected (for example, as a server array, a group of blade servers, or a multi-processor system), with each device providing part of the necessary operations. Figure 3 A processor 301 is taken as an example.

[0065] Memory 302, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as the program instructions / modules corresponding to the method for extracting and processing driving interaction intention segments in the embodiments of this application. Processor 301 executes the software programs, instructions, and modules stored in memory 302 to execute various functional applications and data processing of the device, thereby implementing the aforementioned method for extracting and processing driving interaction intention segments.

[0066] The memory 302 may primarily include a program storage area and a data storage area. The program storage area may store an operating system and at least one application required for a function; the data storage area may store data created based on the use of the terminal, etc. Furthermore, the memory 302 may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some instances, the memory 302 may further include a memory remotely located relative to the processor 301, and these remote memories may be connected to the device via a network. Examples of the aforementioned networks include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0067] The electronic device may further include: an input device 303 and an output device 304. The processor 301, the memory 302, the input device 303 and the output device 304 may be connected via a bus or other means. Figure 3 The bus connection is taken as an example.

[0068] The input device 303 can receive input digital or character information, and the output device 304 can include a display device, an auxiliary lighting device (e.g., an LED), a tactile feedback device (e.g., a vibration motor), etc. The display device may include, but is not limited to, a liquid crystal display (LCD), a light-emitting diode (LED) display, and a plasma display. In some embodiments, the display device may be a touch screen.

[0069] This embodiment provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to cause a computer to execute the above method. The computer instructions on the computer-readable storage medium are used to cause a computer to execute the above method, thereby having at least the same advantages as the above method.

[0070] The medium in this application may be any combination of one or more computer-readable media. The medium may be a computer-readable signal medium or a computer-readable storage medium. The medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or component, or any combination thereof. More specific examples of media (a non-exhaustive list) include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this document, a medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device.

[0071] A computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0072] The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wire, optical cable, RF (Radio Frequency), etc., or any suitable combination of the foregoing.

[0073] Computer program code for performing the operations of the present application can be written in one or more programming languages, or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0074] In the above embodiments, all or part of the embodiments can be implemented using software, hardware, firmware, or any combination thereof. When implemented using software, all or part of the embodiments can be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired connection, such as a coaxial cable, optical fiber, digital subscriber line (DSL), or wireless connection, such as infrared, wireless, or microwave. The computer-readable storage medium can be any available medium that can be accessed by a computer, or a data storage device, such as a server or data center, that includes one or more available media. The available medium can be a magnetic medium (e.g., a floppy disk, hard disk, or magnetic tape), an optical medium, or a semiconductor medium. It is worth noting that the computer-readable storage medium mentioned in the embodiments of the present application may be a non-volatile storage medium, in other words, a non-transitory storage medium.

[0075] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this application can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this application can be achieved. This is not a limitation herein.

[0076] The above specific embodiments do not constitute a limitation on the scope of protection of this application. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application shall be included within the scope of protection of this application.

Claims

1. A method for extracting and processing driving interaction intention fragments, characterized in that: include: During the driving process of the ego vehicle, the driving behavior data of the ego vehicle and the interaction behavior data between the ego vehicle and surrounding vehicles are collected; The surrounding vehicle is any vehicle around the vehicle; Dividing the driving behavior data to obtain a plurality of vehicle behavior segments; Divide the interaction behavior data according to the division time points of the vehicle behavior segments to obtain multiple interaction intention segments; each interaction intention segment includes the driving behavior data of the vehicle and the interaction behavior data between the vehicle and surrounding vehicles; Clustering the multiple interaction intention fragments, and performing intent recognition on each set of interaction intention fragments; The driving behavior data is divided into multiple vehicle behavior segments, including: Performing a first-level division on the driving behavior data according to a lateral behavior characterization variable of the vehicle to obtain a plurality of preliminary vehicle behavior segments; Perform a secondary division on each preliminary ego-vehicle behavior segment based on the longitudinal behavior representation variables of the ego-vehicle to obtain multiple ego-vehicle behavior segments; Among them, the lateral behavior characterization variables include lateral acceleration and vehicle heading angle, and the longitudinal behavior characterization variables include longitudinal velocity and longitudinal acceleration.

2. The method for extracting and processing driving interaction intention fragments according to claim 1, characterized in that: The Bayesian agglomerative data segmentation algorithm is used for primary and secondary segmentation.

3. The method for extracting and processing driving interaction intention fragments according to claim 1, characterized in that: Before performing the first-level division of the driving behavior data according to the lateral behavior characterization variable of the vehicle, the method further includes: Perform correlation analysis on the lateral behavior characterization variables of the vehicle, and retain the lateral behavior characterization variables with correlations below the set threshold; A correlation analysis is performed on the longitudinal behavior characterization variables of the vehicle, and the longitudinal behavior characterization variables with correlations lower than the set threshold are retained.

4. The method for extracting and processing driving interaction intention fragments according to any one of claims 1 to 3, characterized in that: Before clustering the plurality of interaction intent fragments to obtain the category of the interaction intent, the method further includes: From multiple interaction intent segments, select a target segment whose duration is less than a set threshold; The maximum a posteriori estimation is used to solve the regression model coefficients of each variable in the target segment and its adjacent segments before and after; According to the similarity of the regression model coefficients of the target segment and its preceding and following adjacent segments, the target segment is merged with the adjacent interaction intention segments.

5. The method for extracting and processing driving interaction intention fragments according to any one of claims 1 to 3, characterized in that: Intent recognition is performed on each set of interaction intent fragments, including: Perform numerical statistics on the variables in each type of interaction intention fragment set; Intent recognition is performed based on numerical statistical results.

6. An electronic device, characterized in that: include: at least one processor, and a memory communicatively coupled to the at least one processor; Wherein, the memory stores instructions that can be executed by at least one of the processors, and the instructions are executed by at least one of the processors so that at least one of the processors can execute the driving interaction intention fragment extraction and processing method described in any one of claims 1-5.

7. A computer-readable storage medium, characterized in that The medium stores computer instructions, which are used to enable a computer to execute the driving interaction intention segment extraction and processing method according to any one of claims 1 to 5.

Images