Method and apparatus for calling digital information based on vehicle interaction

Abstract

Embodiments of the present disclosure disclose a method and device for calling digital information based on vehicle interaction. A specific embodiment of the method includes: calling a set of vehicle perception information corresponding to a driver's voice lane-changing instruction; filtering the set of vehicle perception information based on the driver's voice lane-changing instruction to obtain a target set of vehicle perception information; performing vehicle information recognition on each target vehicle perception information in the target set of vehicle perception information to generate a set of vehicle recognition information; generating lane-changing opportunity information; performing vehicle lateral trajectory planning on a current vehicle according to the lane-changing opportunity information to generate a current vehicle lateral planning trajectory; controlling the current vehicle to perform a lane-changing operation, and performing real-time monitoring on a lane-changing environment of the current vehicle to generate a lane-changing environment monitoring result; generating a fallback trajectory; and controlling the current vehicle to return to an original lane according to the fallback trajectory. The embodiment can improve the robustness and safety of an assisted driving system in a complex urban scenario.

Description

Technical Field

[0001] The embodiments disclosed herein relate to the fields of computer technology, vehicle control technology, and path planning technology, and specifically to a method and apparatus for retrieving digital information based on vehicle interaction. Background Technology

[0002] Digital information retrieval is a key technology for intelligent driving vehicles to acquire environmental data through perception systems to support various driver assistance functions. In current urban road driver assistance applications, especially in lane-changing scenarios, to ensure safety, driver assistance systems typically rely on accessing and processing massive amounts of perception information to make comprehensive decisions.

[0003] However, this "extensive" information processing approach has significant bottlenecks: First, it leads to excessive system computational load and delayed decision response, resulting in low information processing efficiency. Second, in complex and dynamic traffic flows, there is a large amount of redundant information that is difficult to filter, which can easily interfere with the core decision-making process. For example, introducing unnecessary computational disturbances when generating lateral trajectories for vehicles not only affects planning efficiency but may also jeopardize driving safety. Therefore, it severely restricts the decision-making efficiency and trajectory planning accuracy of assisted driving systems. Consequently, it reduces the robustness and safety of assisted driving systems in complex urban scenarios. Summary of the Invention

[0004] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0005] Some embodiments of this disclosure propose a method and apparatus for retrieving digital information based on vehicle interaction to solve the technical problems mentioned in the background section above.

[0006] In a first aspect, some embodiments of this disclosure provide a digital information retrieval method based on vehicle interaction. The method includes: in response to receiving a driver's voice lane-change instruction, retrieving a vehicle perception information set corresponding to the driver's voice lane-change instruction; filtering the vehicle perception information set based on the driver's voice lane-change instruction to obtain a target vehicle perception information set; performing vehicle information recognition on each target vehicle perception information in the target vehicle perception information set to generate a vehicle recognition information set; generating lane-change timing information based on the vehicle recognition information set; performing lateral trajectory planning for the current vehicle based on the lane-change timing information to generate a lateral planning trajectory for the current vehicle; controlling the current vehicle to perform a lane-change operation based on the lateral planning trajectory for the current vehicle, and performing real-time monitoring of the lane-change environment for the current vehicle to generate a lane-change environment monitoring result; in response to the lane-change environment monitoring result indicating an abnormal lane-change state, generating a reversal trajectory based on the lateral planning trajectory for the current vehicle; controlling the current vehicle to return to the original lane according to the reversal trajectory, and issuing a lane-change waiting prompt message to the driver.

[0007] Secondly, some embodiments of this disclosure provide a digital information retrieval device based on vehicle interaction. The device includes: a retrieval unit configured to, in response to receiving a driver's voice lane change command, retrieve a set of vehicle perception information corresponding to the driver's voice lane change command; a filtering unit configured to filter the set of vehicle perception information based on the driver's voice lane change command to obtain a target set of vehicle perception information; a vehicle information recognition unit configured to perform vehicle information recognition on each target vehicle perception information in the target set of vehicle perception information to generate a vehicle recognition information set; and a first generation unit configured to generate lane change timing information based on the vehicle recognition information set. The vehicle lateral trajectory planning unit is configured to plan the lateral trajectory of the current vehicle based on the lane change timing information to generate the current vehicle's lateral trajectory planning; the control and monitoring unit is configured to control the current vehicle to perform a lane change operation based on the current vehicle's lateral trajectory planning, and to perform real-time monitoring of the lane change environment to generate a lane change environment monitoring result; the second generation unit is configured to respond to the lane change environment monitoring result indicating an abnormal lane change state and generate a reversal trajectory based on the current vehicle's lateral trajectory planning; the control and prompting unit is configured to control the current vehicle to return to the original lane according to the reversal trajectory and to issue a lane change waiting prompt message to the driver.

[0008] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation of the first aspect above.

[0009] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method described in any of the implementations of the first aspect above.

[0010] The various embodiments of this disclosure have the following beneficial effects: the vehicle-interaction-based digital information retrieval method of some embodiments of this disclosure can improve the robustness and safety of assisted driving systems in complex urban scenarios. Specifically, the reason for the reduced robustness and safety of assisted driving systems in complex urban scenarios is that the "extensive" information processing method has obvious bottlenecks: First, it leads to excessive system computational load and delayed decision response, i.e., low information processing efficiency; second, in complex and dynamic traffic flows, there is more redundant information that is difficult to filter, which can easily interfere with the core decision-making process. For example, when generating lateral trajectories for vehicles, unnecessary computational disturbances are introduced, which not only affect planning efficiency but may also endanger driving safety. Therefore, it seriously restricts the decision-making efficiency and trajectory planning accuracy of assisted driving systems. Based on this, the vehicle-interaction-based digital information retrieval method of some embodiments of this disclosure first, in response to receiving a driver's voice lane change command, retrieves the vehicle perception information set corresponding to the driver's voice lane change command, and filters the vehicle perception information set based on the driver's voice lane change command to obtain the target vehicle perception information set. Here, by filtering the retrieved vehicle perception information set, the amount of data that the assisted driving system needs to process is greatly reduced, overcoming the problem of information overload. Simultaneously, by combining the driver's voice lane-change command with vehicle interaction, precise filtering of massive amounts of data can be achieved, thereby improving the system's processing efficiency. Then, vehicle information recognition is performed on each target vehicle perception information in the aforementioned target vehicle perception information set to generate a vehicle identification information set. Based on this vehicle identification information set, lane-change timing information is generated. Here, by recognizing vehicle information and generating lane-change timing information, precise control of lane-change timing is possible. Next, based on the aforementioned lane-change timing information, lateral trajectory planning is performed on the current vehicle to generate its lateral trajectory. Here, based on the lane-change timing condition, lateral trajectory planning generates a more accurate lateral trajectory for the current vehicle. This improves the decision-making efficiency and trajectory planning accuracy of the assisted driving system. Finally, based on the current lateral trajectory, the system controls the current vehicle to perform a lane-change operation and monitors the lane-change environment in real time to generate lane-change environment monitoring results. Here, real-time monitoring of the lane-changing environment allows for timely identification of situations that do not conform to lane-changing rules (e.g., a narrowing lane-changing interval or other vehicles entering the lane-changing interval). Finally, in response to the abnormal lane-changing status indicated by the aforementioned lane-changing environment monitoring results, a reversal trajectory is generated based on the current vehicle's lateral planning trajectory. The vehicle is then controlled to return to its original lane according to this reversal trajectory, and a lane-changing waiting prompt is issued to the driver. By generating the reversal trajectory and controlling the vehicle to return to its original lane, the vehicle can temporarily avoid risks, significantly improving the robustness and safety of the assisted driving system in complex urban scenarios. Attached Figure Description

[0011] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0012] Figure 1 This is a flowchart of some embodiments of the digital information retrieval method based on vehicle interaction according to the present disclosure; Figure 2 This is a schematic diagram of the current lateral planning trajectory of a vehicle according to some embodiments of the digital information retrieval method based on vehicle interaction disclosed herein; Figure 3 This is a schematic diagram of the elastic deformation of the current vehicle's lateral planned trajectory; Figure 4 These are schematic diagrams illustrating the structure of some embodiments of the vehicle-interaction-based digital information retrieval device according to this disclosure; Figure 5 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0013] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0014] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0015] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0016] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0017] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0018] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0019] Figure 1 A flow 100 of some embodiments of a vehicle-interaction-based digital information retrieval method according to the present disclosure is shown. This vehicle-interaction-based digital information retrieval method includes the following steps: Step 101: In response to receiving a driver's voice lane change command, the vehicle perception information set corresponding to the driver's voice lane change command is invoked.

[0020] In some embodiments, the executor of the vehicle-interaction-based digital information retrieval method (e.g., a computing device, i.e., a driver assistance system) responds to receiving a driver's voice lane change command by retrieving a set of vehicle perception information corresponding to the driver's voice lane change command via wired or wireless means. Each piece of vehicle perception information in the set corresponds to one vehicle. Here, the set of vehicle perception information corresponding to the driver's voice lane change command can be retrieved from a data pool. The data pool stores various data detected by the vehicle within a certain period of time.

[0021] It should be noted that the aforementioned wireless connection methods may include, but are not limited to, 3G / 4G / 5G connections, WiFi connections, Bluetooth connections, WiMAX connections, Zigbee connections, UWB (Ultra Wide Band) connections, and other currently known or future wireless connection methods.

[0022] It should be noted that the aforementioned computing devices can be either hardware or software. When the computing device is hardware, it can be implemented as a distributed cluster consisting of multiple servers or terminal devices, or as a single server or a single terminal device. When the computing device is software, it can be installed on the hardware devices listed above. It can be implemented as, for example, multiple software programs or software modules used to provide distributed services, or as a single software program or software module. No specific limitations are made here.

[0023] Step 102: Based on the driver's voice lane change command, filter the vehicle perception information set to obtain the target vehicle perception information set.

[0024] In some embodiments, the execution entity may filter the vehicle perception information set based on the driver's voice lane change command to obtain the target vehicle perception information set.

[0025] In some optional implementations of certain embodiments, the execution entity filters the vehicle perception information set based on the driver's voice lane change command to obtain a target vehicle perception information set, including: Step S1: Perform speech recognition on the driver's voice lane change command to obtain command keywords. These command keywords represent the direction of the lane change. The speech recognition module within the driver assistance system can be used to perform speech recognition on the driver's voice lane change command to obtain the command keywords. Here, the semantic recognition module can first convert the driver's voice lane change command into text, and then extract keywords from the text using a pre-set keyword extraction algorithm. For example, the keyword extraction algorithm may include: term frequency-inverse document frequency algorithm, bag-of-words model, etc.

[0026] In practice, the instruction keywords can be "change lanes to the left" or "change lanes to the right".

[0027] Step S2 involves determining the positional relationship between the vehicle corresponding to each piece of vehicle perception information in the aforementioned vehicle perception information set and the current vehicle, thus obtaining a positional relationship information set. The vehicle perception information may include a sequence of consecutive perception images, point cloud data, millimeter-wave radar data, etc. Furthermore, the vehicle perception information may be pre-processed information. Specifically, the perception image may be a pre-selected single-target image after target segmentation, corresponding to a single vehicle. The point cloud data may be the point cloud data corresponding to a single vehicle after removing interference points.

[0028] First, the point cloud data can be converted to the current vehicle's coordinate system. Then, the point cloud coordinates closest to the current vehicle can be selected from the converted point cloud data. Second, the positional relationship between the target vehicle and the current vehicle can be determined based on the point cloud coordinates. For example, if the converted point cloud coordinates are in the same lane as the current vehicle and their x-coordinate is greater than the x-coordinate of the current vehicle's position, then the target vehicle is determined to be in the same lane and ahead of the current vehicle. The corresponding marker can then be used to determine the positional relationship information.

[0029] Step S3: Based on the above instruction keywords and the above location relationship information set, filter the above vehicle perception information set to obtain the target vehicle perception information set.

[0030] In practice, when changing lanes, drivers typically focus their attention on the vehicle in front in their current lane, the vehicles in front and behind in the target lane, and several vehicles in the lane on the other side of the lane change area to complete the lane change safely and efficiently. Therefore, to avoid excessive sensory information affecting the driver assistance system, positional relationship information is generated by simulating the driver's observation habits. This facilitates the selection of vehicles requiring attention, thereby reducing the amount of information processed and improving data processing efficiency.

[0031] Specifically, vehicle perception information that matches the corresponding positional relationship information with the instruction keywords can be selected from the vehicle perception information set as the target vehicle perception information.

[0032] As an example, if the instruction keyword is "change lanes to the right," then the positional relationship information matching the instruction keyword can include: the relationship between the vehicle in front and behind the current lane, the relationship between the vehicle in the first lane to the right and the vehicle in the second lane to the right being less than a preset distance threshold, and the relationship between the vehicle in the second lane to the right and the vehicle in the second lane to the right being less than a preset distance threshold. Therefore, the following vehicles can be identified: the vehicle in front of the current vehicle, the n closest vehicles to the current vehicle in the first lane to the right, and the m closest vehicles to the current vehicle in the second lane to the right. Thus, the target vehicle perception information set can be obtained. Here, the numbers n and m are less than a preset upper limit.

[0033] Step 103: Perform vehicle information identification on each target vehicle perception information in the target vehicle perception information set to generate a vehicle identification information set.

[0034] In some embodiments, the aforementioned executing entity can perform vehicle information identification on the various target vehicle perception information in the aforementioned target vehicle perception information set to generate a vehicle identification information set. This can be achieved by communicating with the nearest roadside unit on the road to obtain the vehicle identification information set for each target vehicle. Here, the roadside unit can be a road facility in V2I (Vehicle-to-Infrastructure) capable of collecting data from vehicles on the road. For example, a traffic light with data acquisition and communication functions.

[0035] In some optional implementations of certain embodiments, the target vehicle perception information may include: a sequence of perception images, point cloud data, radar perception data, and vehicle identification. The aforementioned executing entity performs vehicle information identification on each piece of target vehicle perception information in the aforementioned target vehicle perception information set to generate a vehicle identification information set, including: For each target vehicle perception information in the above target vehicle perception information set, perform the following generation steps: Step S1 involves using the perceived image sequence and point cloud data included in the target vehicle perception information to identify the vehicle size and obtain vehicle size information. Specifically, this involves performing point cloud clustering on the point cloud data in the target vehicle perception information and determining the minimum bounding rectangle of the clustered point cloud. Next, a preset vehicle recognition algorithm (e.g., convolutional neural network, YOLO (You Only Look Once) object detection algorithm) can be used to identify the vehicle in the perceived image to obtain a vehicle identifier. Then, a lookup table is used to find the vehicle size corresponding to the vehicle identifier. Here, the size of the minimum bounding rectangle is adjusted based on the vehicle size to obtain an adjusted minimum bounding rectangle. Finally, the size of the adjusted minimum bounding rectangle is determined as the vehicle size information. Specifically, due to vehicle occlusion issues, incomplete point cloud data detection can easily occur, leading to potentially larger vehicle sizes. Therefore, the minimum bounding rectangle is adjusted through perceived image recognition.

[0036] Step S2: Based on the radar sensing data and point cloud data included in the target vehicle perception information, determine the vehicle driving information and vehicle distance value. The vehicle driving information includes: speed value, acceleration value, and angular velocity value. Next, a clustering algorithm can be used to determine the center coordinates of the point cloud region corresponding to the target vehicle at different times in the point cloud data. Therefore, the target vehicle's speed value can be determined using the center coordinates at two different times and the time interval between the two times. Here, the acceleration value can be generated using the collected center coordinates and time points at different times. Additionally, the radar sensing data may include the target vehicle's heading angle. Therefore, the target vehicle's speed value in the lane direction and acceleration value in the lane direction can be calculated based on the angle between the heading angle and the lane direction. Finally, the target vehicle's angular velocity value is calculated using the steering angular velocity formula and the heading angle. Here, the vehicle distance value can be the distance between the target vehicle and the current vehicle. The point cloud coordinates closest to the current vehicle can be selected from the point cloud data to determine the vehicle distance between the point cloud coordinates and the current vehicle.

[0037] Step S3: The vehicle size information, vehicle driving information, and vehicle distance value are determined as the vehicle identification information in the above vehicle identification information set.

[0038] Step 104: Generate lane change timing information based on the vehicle identification information set.

[0039] In some embodiments, the aforementioned executing entity may generate lane change timing information based on the aforementioned vehicle identification information set.

[0040] In some optional implementations of certain embodiments, the execution entity generates lane change timing information based on the vehicle identification information set, including: Step S1: In the current vehicle's coordinate system, based on the vehicle distance and size information included in the vehicle identification information set, determine the vehicle intervals between target vehicles in the target lane to obtain a vehicle interval sequence. The target lane is the lane adjacent to the current lane corresponding to the lane change direction, and each vehicle interval corresponds to the vehicle identifiers of two target vehicles. Here, the closest point cloud coordinates can be selected from the point cloud regions corresponding to the two target vehicles, and the distance between the two point cloud coordinates is determined as the vehicle interval value. Alternatively, a keypoint extraction algorithm can be used to extract keypoints from the target vehicles in the perceived image to obtain the keypoint coordinates of the target vehicles. Specifically, the keypoint coordinates of the target vehicles can be coordinates within the front or rear regions of the target vehicle. Therefore, the distance between the keypoint coordinates of the target vehicles can be combined with the vehicle distance value for further adjustment to determine the accuracy of the vehicle interval.

[0041] As an example, keypoint extraction algorithms may include, but are not limited to, at least one of the following: SIFT (Scale-Invariant Feature Transform), FAST (Features from Accelerated Segment Test), Harris corner detection algorithm, etc.

[0042] Step S2: Vehicle identification information that meets preset screening conditions within the vehicle identification information set is identified as target vehicle identification information, resulting in a target vehicle identification information group. Specifically, the vehicle interval in the vehicle interval sequence corresponding to the target vehicle identification information is greater than a preset interval threshold, and the included speed and acceleration values ​​meet preset lane-changing conditions. A vehicle interval greater than the preset interval threshold (e.g., 10 meters) indicates sufficient space to support a lane change. For the first target vehicle in the target lane (i.e., the lane indicated by the instruction keyword) located at either end of the lane-changing space, the preset lane-changing conditions can be that the speed of the first target vehicle is greater than the target speed value and its acceleration value is greater than the target acceleration value. Here, the target speed and target acceleration values ​​can be generated based on the current vehicle's speed and acceleration values. This can be used to determine that the first target vehicle is not in an emergency braking deceleration state, thus ensuring sufficient space in the lane-changing area. For the second target vehicle in the target lane (i.e., the adjacent lane indicated by the instruction keyword) located at either end of the lane-changing space, the preset lane-changing conditions can be that the speed of the second target vehicle is less than the target speed value and its acceleration value is less than the target acceleration value. Therefore, it can be determined that the target vehicle is not accelerating (or its acceleration and speed are both less than the current vehicle), thus ensuring sufficient space in the lane-changing area. This allows for further filtering of suitable vehicle intervals for lane changes and the corresponding target vehicle identification information.

[0043] Step S3: Perform continuous behavior detection on the target vehicles corresponding to each target vehicle identification information in the above target vehicle identification information group to generate a vehicle spacing feature value group. The vehicle spacing feature values ​​can be generated using the following formula: .

[0044] in, This represents the characteristic value of vehicle spacing. This indicates the weight of the speed term. This indicates the weight of the acceleration term. This represents the weight of the time component. Here, the sum of the three weights is 1. This indicates the speed of the target vehicle. This indicates the acceleration value of the target vehicle or the speed limit value of the road. This indicates the duration the vehicle's turn signal is currently on. This indicates the maximum speed of the traffic flow. This represents the absolute value of the maximum acceleration of the traffic flow. This represents the time decay constant.

[0045] In practice, firstly, it's considered that the lower the speed of the following vehicle, the stronger its intention to yield is likely to be. If the following vehicle is moving slowly, it's easier for it to slow down or maintain a distance, thus providing space for the current vehicle to change lanes. Simultaneously, relative speed (the ratio to the maximum speed of the traffic flow) is normalized to between 0 and 1. When the following vehicle's speed equals the maximum permissible speed, this item is 0, indicating no intention to yield; when the following vehicle's speed is 0, this item is the speed-weighted value, representing the maximum intention to yield. Similarly, considering the driving experience of "slowing down means yielding," acceleration is the most direct indicator of whether the following vehicle is yielding. If the following vehicle is decelerating (a is negative), it's very likely that it is yielding. We take the negative value of acceleration and divide it by the absolute value of the maximum deceleration for normalization, then take the maximum value between this and 0, ensuring that this item only has a positive value when decelerating. The greater the deceleration, the larger the value of this item, indicating a stronger intention to yield. If the following vehicle is accelerating, this item is 0, indicating no intention to yield. Furthermore, considering that if a following vehicle responds promptly (i.e., decelerates within a short time) after the first vehicle activates its turn signal, its yielding intention is stronger. Therefore, we use an exponential decay function to simulate this time sensitivity. The shorter the time t after the turn signal is activated, the larger this value is; as time progresses, this value decays, indicating that if the following vehicle does not respond for a long time, the yielding intention of the target vehicle weakens. Thus, a yielding intention quantification formula that conforms to physical laws and closely reflects actual driving scenarios is constructed, allowing for accurate determination of vehicle spacing characteristic values. Here, vehicle spacing characteristic values ​​characterize the intensity of lane change timing.

[0046] Step S4: The vehicle interval corresponding to the largest vehicle spacing feature value in the above vehicle spacing feature value group and the vehicle identification information of the two target vehicles are determined as lane change timing information, wherein the two target vehicles are the first vehicle and the second vehicle.

[0047] In practice, lane-change timing is often determined based on the current gap size and relative speed, requiring the classification of the target vehicle's behavioral intent. However, in urban road scenarios, vehicle intent changes frequently, leading not only to difficulties in intent classification but also to the potential for misclassification, thus impacting subsequent planning by the assisted driving system. Therefore, step 104 and related content in this application do not require classifying the target vehicle's behavioral intent. Instead, they utilize continuous observation data of the target vehicle's motion state that influences the lane-change area to determine the target vehicle's motion trend. This effectively solves the classification difficulties caused by frequent changes in vehicle intent in urban roads, thereby improving the accuracy of lane-change timing control.

[0048] Step 105: Based on the lane change timing information, perform lateral trajectory planning for the current vehicle to generate the lateral trajectory planning for the current vehicle.

[0049] In some embodiments, the aforementioned execution entity may perform lateral trajectory planning for the current vehicle based on the aforementioned lane change timing information to generate the current vehicle's lateral planning trajectory.

[0050] Optionally, the lateral trajectory of the current vehicle can be planned using the following methods to generate the lateral trajectory of the current vehicle: Step S1: Obtain the current vehicle wheelbase, current vehicle speed, current vehicle front wheel steering angle, and current vehicle heading angle.

[0051] Step S2: Based on the current vehicle speed and the aforementioned current front wheel steering angle, generate a first lateral planning trajectory for the vehicle. This first lateral planning trajectory may include path planning coordinates and path planning heading angles at corresponding consecutive time points, and can be the first segment of the lateral planning trajectory during the vehicle's lateral trajectory planning process. Here, consecutive time points can be any time point within a preset planning duration. Furthermore, the first lateral planning trajectory can be generated using the following formula: .

[0052] in, This represents the x-coordinate value of the path planning coordinates at a specific point in time. This represents the ordinate value of the path planning coordinates at a specific point in time. This represents the heading angle for the path planning at a specific point in time. This represents the current vehicle speed. Here, the current vehicle speed must satisfy the constraint that it is greater than or equal to a preset minimum speed threshold. This indicates the current wheelbase of the vehicle. This indicates the heading angle of the current vehicle. This indicates the current front wheel steering angle of the vehicle. It represents the pose vector of the vehicle in the vehicle body coordinate system at a certain moment.

[0053] In practice, the above formula can be used to generate vehicle pose vectors for each time point within a preset planning time. Furthermore, considering that the first lateral planning trajectory cannot be generated at excessively low vehicle speeds, a minimum speed limit can be set.

[0054] Optionally, generating the current vehicle's lateral trajectory may also include the following steps: Step S1: In response to determining that the path planning heading angle corresponding to the last time point in the vehicle's lateral planning trajectory is not parallel to the lane line, and obtaining map information corresponding to the current vehicle position coordinates, a lane centerline equation is generated based on the map information and the lane where the vehicle is currently located. The non-parallelism between the path planning heading angle and the lane line indicates that the vehicle's driving direction has not been aligned after moving along the lateral planning trajectory. Therefore, with the map information obtained, two lateral planning trajectories can be generated. Secondly, the map information can be lane information of the current vehicle's lane obtained from a high-precision map. For example, lane information can include a set of road line equations. Here, the lane centerline equation can be generated as follows: First, determine the centerline equation of the two road line equations corresponding to the lane where the first lateral planning trajectory is located in the road line equation set, obtaining the first centerline equation. Then, obtain the second centerline equation by combining the centerline equations of the two lane line equations corresponding to the lane where the first lateral planning trajectory is located in the lane information. Finally, the lane centerline equation can be obtained by fitting the first and second centerline equations using the least squares method.

[0055] Step S2: Starting from the path planning coordinates corresponding to the last time point in the vehicle's lateral planning trajectory, and aiming for the path planning heading angle corresponding to the last time point to be parallel to the lane line, trajectory planning is performed to obtain the second vehicle lateral planning trajectory. This second vehicle lateral planning trajectory can be the second segment of the lateral planning trajectory in the vehicle lateral trajectory planning process. During the trajectory planning process, the heading angles corresponding to each trajectory coordinate in the second vehicle lateral planning trajectory gradually decrease to 0. Specifically, the heading angles corresponding to each trajectory coordinate in the second vehicle lateral planning trajectory can be generated using the following formula: .

[0056] in, This represents the heading angle corresponding to the first of two adjacent trajectory coordinates in the lateral planning trajectory of the second vehicle. This represents the heading angle corresponding to the second trajectory coordinate among two adjacent trajectory coordinates in the lateral planning trajectory of the second vehicle. This indicates the preset heading angle attenuation rate.

[0057] Step S3: Starting from the vehicle coordinates corresponding to the last time point in the second vehicle lateral planning trajectory, generate a third vehicle lateral planning trajectory that moves along the lane centerline equation. This third vehicle lateral planning trajectory is the third segment of the lateral planning trajectory in the vehicle lateral trajectory planning process.

[0058] In practice, the first of the three trajectory segments is mandatory, while the second and third segments are optional depending on the system output. If the driver assistance system obtains map information, it can generate the second and third segments; otherwise, it can only generate the first trajectory segment. Therefore, this scheme can plan normally even without lane lines, thus ensuring the robustness of lateral planning and the consistency of longitudinal motion perception.

[0059] As an example, such as Figure 2 As shown, the right lane is the current vehicle's lane, and the left lane is the target lane the vehicle wants to change to. The dashed line represents the centerline equation of the target lane. The diagram illustrates a three-segment lateral planning trajectory for the current vehicle, including trajectory segment 201, trajectory segment 202, and trajectory segment 203. This facilitates guiding the current vehicle to change lanes from its current lane to the target lane.

[0060] In some optional implementations of certain embodiments, the execution entity performs lateral trajectory planning for the current vehicle based on the lane change timing information to generate the current vehicle's lateral planning trajectory, including: Step S1: Determine the endpoint coordinates of the path planning based on the vehicle spacing in the lane change timing information. Specifically, if the radial distance (i.e., in the lane direction) between the vehicle spacing and the current vehicle is greater than or equal to the base distance, extend the current vehicle's position coordinates forward along the lane direction by the lane change distance value to obtain the extended coordinates. Finally, determine the coordinates of the nearest point on the target lane's centerline that is the same as the extended coordinates as the endpoint coordinates of the path planning. Here, the base distance can be greater than the lane change distance value, and the difference between the base distance and the lane change distance value can be greater than the current vehicle's braking distance. The lane change distance value can be the optimal lane change distance pre-calibrated based on parameters such as the current vehicle's wheelbase, vehicle speed, and turning radius.

[0061] If the radial distance between the vehicles is less than the base distance, the position of the radial distance between the current vehicle and the first vehicle is determined as the abscissa of the path planning endpoint coordinates by a preset ratio, and the ordinate of the position on the center line of the target lane that is at the same position as the abscissa is determined as the ordinate of the path planning endpoint coordinates, thus obtaining the path planning endpoint coordinates.

[0062] Step S2: Based on the current vehicle's location coordinates, the coordinates of the path planning endpoint, and the lane change timing information, perform lateral trajectory planning for the current vehicle to generate its lateral trajectory. This can be achieved by using a fifth-order polynomial to plan the lateral trajectory between the current vehicle's location coordinates and the path planning endpoint coordinates, resulting in an initial trajectory sequence. Next, each initial trajectory is scored using a pre-set cost function, yielding an initial trajectory score sequence. This cost function can include the following formula terms: lateral acceleration, longitudinal acceleration, curvature, and speed deviation. Specifically, each formula term can have a corresponding dynamic weight. The lateral acceleration term can be the maximum lateral acceleration corresponding to the initial trajectory. The longitudinal acceleration term can be the maximum longitudinal acceleration corresponding to the initial trajectory. The curvature term can be the maximum curvature corresponding to the initial trajectory. The speed deviation term can be the integral of the absolute value of the deviation between the current speed and the target speed. Finally, the initial trajectory with the smallest initial trajectory score is determined as the current vehicle's lateral trajectory.

[0063] In practice, considering the comfort of lane-changing processes, lateral acceleration and longitudinal acceleration terms are introduced. A curvature term is introduced to prevent loss of vehicle control. Finally, a speed deviation term is introduced to maintain speed consistency.

[0064] Step 106: Based on the current vehicle's lateral planning trajectory, control the current vehicle to perform a lane change operation and monitor the lane change environment in real time to generate lane change environment monitoring results.

[0065] In some embodiments, the aforementioned execution entity can control the current vehicle to perform a lane-changing operation based on the current vehicle's lateral planning trajectory, and perform real-time monitoring of the lane-changing environment to generate lane-changing environment monitoring results. Specifically, it can control the current vehicle to move along the vehicle's lateral planning trajectory.

[0066] In some optional implementations of certain embodiments, the aforementioned execution entity performs real-time monitoring of the lane-changing environment of the current vehicle to generate lane-changing environment monitoring results, including: Step S1: Determine the lane-changing section between the first vehicle and the second vehicle. Specifically, the lane-changing section can be defined in the vehicle coordinate system of the current vehicle, specifically the ground area between the first and second vehicles in the target lane.

[0067] Step S2: Retrieve real-time vehicle perception information of target vehicles adjacent to the aforementioned lane-changing section to obtain a real-time vehicle perception information set. Here, real-time vehicle perception information detected by the current vehicle's sensors can be retrieved in real time. Sensors may include, but are not limited to, at least one of the following: vehicle-mounted camera, LiDAR, millimeter-wave radar, etc.

[0068] Step S3 involves state tracking of each real-time vehicle perception information in the aforementioned real-time vehicle perception information set to generate a real-time vehicle state information set. This real-time vehicle state information includes the vehicle's real-time position coordinates, real-time speed, real-time acceleration, and movement status identifier. Here, the Kalman filter method can be used to track the state of each real-time vehicle perception information to obtain the real-time vehicle state information set. State tracking may include tracking data such as the target vehicle's speed, acceleration, and minimum bounding box. Therefore, the real-time vehicle state information may include real-time updated data such as the target vehicle's speed, acceleration, and minimum bounding box.

[0069] Step S4: In response to determining that there are real-time vehicle status information in the real-time vehicle status information set that meets the lane change conflict conditions, a lane change environment monitoring result characterizing the lane change anomaly is generated. The lane change conflict condition can be that the coordinates of the smallest bounding rectangle included in the real-time vehicle status information are within the lane change interval, thus indicating that other target vehicles also have a lane change requirement at the same time. Therefore, a lane change environment monitoring result characterizing the lane change anomaly can be generated. Alternatively, the lane change conflict condition can also include the first vehicle's speed reduction exceeding a preset speed percentage and its acceleration being less than a preset acceleration threshold, or the second vehicle's speed increase exceeding a preset speed percentage and its acceleration being greater than a preset acceleration threshold.

[0070] Optionally, the aforementioned implementing entity performs real-time monitoring of the lane-changing environment of the current vehicle to generate lane-changing environment monitoring results, and also includes: Step S1: Sample the current vehicle's lateral planning trajectory to obtain a trajectory sampling point sequence. This can be achieved by uniformly sampling the current vehicle's lateral planning trajectory using preset lateral or longitudinal intervals to obtain the trajectory sampling point sequence.

[0071] Step S2 involves real-time vehicle state detection of the first and second vehicles to generate a first state detection feature vector sequence and a second state detection feature vector sequence. Specifically, the first vehicle can be detected using a Kalman filter to obtain a first predicted speed value, a first predicted acceleration value, and a first predicted minimum bounding rectangle; the second vehicle can be detected using a Kalman filter to obtain a second predicted speed value, a second predicted acceleration value, and a second predicted minimum bounding rectangle. Next, the minimum distance between the first predicted minimum bounding rectangle and each trajectory sampling point in the trajectory sampling point sequence is determined to obtain a first distance value sequence. The minimum distance between the second predicted minimum bounding rectangle and each trajectory sampling point in the trajectory sampling point sequence is then determined to obtain a second distance value sequence. Here, the units for speed, acceleration, and minimum distance are all metric units. Then, for each trajectory sampling point in the trajectory sampling point sequence corresponding to the first predicted speed value, first predicted acceleration value, first distance value, second predicted speed value, second predicted acceleration value, and second distance value, the following steps are performed: The first step is to construct the first state detection feature vector using the first predicted velocity value, the first predicted acceleration value, and the first distance value, through the following formula: .

[0072] in, Indicates the first The first state detection feature vector corresponding to each trajectory sampling point. Indicates the first predicted velocity value, This represents the first predicted acceleration value. Indicates the first The first distance value corresponding to each trajectory sampling point. This represents the first distance attenuation coefficient. This represents the normalized velocity baseline value. This represents the normalized acceleration reference value. This represents the normalized distance reference value. Here, the normalized velocity reference value, normalized acceleration reference value, and normalized distance reference value can be preset.

[0073] The second step involves constructing a second state detection feature vector using the second predicted velocity value, the second predicted acceleration value, and the second distance value, through the following formula: .

[0074] in, Indicates the first The second state detection feature vector corresponding to each trajectory sampling point. Indicates the second predicted velocity value, This represents the second predicted acceleration value. Indicates the first The second distance value corresponding to each trajectory sampling point. This represents the second distance attenuation coefficient. In practice, the second distance attenuation coefficient is greater than the first distance attenuation coefficient, making the motion state of the second vehicle more important.

[0075] Step S3: Using the pre-constructed trajectory elastic deformation model, the first state detection feature vector sequence, and the second state detection feature vector sequence, the elastic strength value of each trajectory sampling point in the trajectory sampling point sequence is determined, resulting in an elastic strength value sequence. Specifically, the elastic strength value corresponding to each trajectory sampling point can be determined using the trajectory elastic deformation model, thus obtaining the elastic strength value sequence.

[0076] Specifically, the trajectory elastic deformation model can be expressed by the following formula: .

[0077] in, Indicates the relationship with the first The first vehicle elasticity value corresponding to each trajectory sampling point. Indicates the relationship with the first The second vehicle elasticity value corresponding to each trajectory sampling point. This represents the first scaling factor, corresponding to the first vehicle. It is used to control the degree to which the state of the first vehicle affects the trajectory sampling points. This represents the second scaling factor, corresponding to the second vehicle. It controls the degree to which the state of the second vehicle affects the trajectory sampling points. Here, the first scaling factor is smaller than the second scaling factor. The first and second scaling factors belong to [0, 1], and their sum is 1. This represents the first weight vector, including the weights: . This represents the second weight vector, including the weights: .

[0078] Here, the sum of the first vehicle elasticity value and the corresponding second vehicle elasticity value can be determined as the elasticity intensity value applied by the first and second vehicles to the trajectory sampling points. Furthermore, since the positive velocity or acceleration of the first vehicle has a reduced impact on the lane change area, while the negative velocity or acceleration has a stronger impact, the first weight vector is a negative vector. Since the positive velocity or acceleration of the second vehicle has a stronger impact on the lane change area, while the negative velocity or acceleration has a weaker impact, the second weight vector is a positive vector. Therefore, the sign of the second vehicle elasticity value is the same as the sign of the second predicted acceleration value. Thus, when the first and second vehicle elasticity values ​​are added, the sign of the first vehicle elasticity value is opposite to the sign of the first predicted acceleration value.

[0079] Step S4: In response to the existence of an elastic strength value in the above elastic strength value sequence that satisfies a preset elastic abrupt change condition, a lane change environment monitoring result characterizing the lane change anomaly is generated. The preset elastic abrupt change condition can be that the elastic strength value applied to the trajectory sampling point is greater than a preset elastic threshold (i.e., resilience value), indicating a conflict between the lane change trajectories of the second vehicle and the current vehicle. In practice, if an elastic strength value that satisfies the preset elastic abrupt change condition exists, the lateral planning trajectory of the current vehicle is disconnected from the corresponding trajectory sampling point.

[0080] In addition, in response to the absence of an elastic strength value that satisfies the preset elastic change condition in the above elastic strength value sequence, the current vehicle lateral planning trajectory is elastically adjusted using the above elastic strength value sequence to obtain the adjusted planning trajectory.

[0081] The current lateral trajectory of the vehicle can be flexibly adjusted using the following formula: .

[0082] in, Indicates the relationship with the first The trajectory elastic adjustment value corresponding to each trajectory sampling point. This indicates the preset deformation sensitivity coefficient. This indicates the current location of the vehicle. This represents the deformation propagation attenuation constant. Indicates the first Each trajectory sampling point. Indicates the adjusted number Each trajectory sampling point. Indicates the first The elastic strength value corresponding to each trajectory sampling point.

[0083] In addition, after flexibly adjusting the trajectory sampling points, the current lateral planned trajectory of the vehicle can be fitted with the adjusted trajectory sampling points to obtain the adjusted planned trajectory. Here, flexibly adjusting the current vehicle's lateral trajectory is an iterative process. After adjustment, the lane change environment of the current vehicle continues to be monitored in real time to generate lane change environment monitoring results until a lane change environment monitoring result characterizing the lane change anomaly is generated or the lane change is completed.

[0084] In practice, considering the continuously changing traffic environment, traditional planning methods typically generate fixed trajectories. When the environment changes, replanning is necessary. Directly adjusting the trajectory based on vehicle data of surrounding target vehicles can easily introduce too many disturbances, leading to frequent trajectory changes and jitter. This results in discontinuous trajectories and abrupt vehicle behavior. This application addresses this by using a trajectory elastic deformation model to treat the trajectory as a continuously deformable boundary, allowing for smooth adjustments based on environmental changes (such as the movement of surrounding vehicles), avoiding the jumps caused by replanning. Secondly, in the lateral trajectory planning process, a trajectory elastic deformation model and its "resilience" concept are introduced for the first time, naturally defining a safety boundary. When environmental pressure (such as pressure from vehicles in front and behind) exceeds the trajectory's resilience limit, the trajectory breaks, triggering more conservative behavior (such as abandoning lane changes). Specifically, elastic deformation allows the trajectory to deform under external influences, similar to the fine-tuning behavior of a human driver, making the behavior of assisted driving vehicles easier for other traffic participants to understand and predict. Furthermore, an elastic coefficient is introduced to characterize the trajectory's resistance to deformation, and a damping coefficient is introduced to characterize the oscillation attenuation of deformation. This makes the debugging process more intuitive, easier to understand and adjust, and improves the robustness of vehicle lateral trajectory planning.

[0085] As an example, such as Figure 3 As shown, Figure 3 Due to the influence of the first state detection feature vector 301 and the second state detection feature vector 302 on the trajectory sampling point 304 on the current vehicle's lateral planning trajectory 303, a path adjustment was made, resulting in the adjusted planning trajectory 305.

[0086] Step 107: In response to the abnormal lane change status indicated by the lane change environment monitoring results, generate a reversal trajectory based on the current vehicle's lateral planning trajectory.

[0087] In some embodiments, the execution entity may, in response to the lane change environment monitoring results indicating an abnormal lane change state, generate a reversal trajectory based on the current vehicle lateral planning trajectory. Specifically, first, trajectory sampling points that are disconnected in the current vehicle lateral planning trajectory are identified. Then, these trajectory sampling points are moved a target distance towards the original lane to obtain a reversal trajectory reference point. Finally, using the reversal trajectory reference point as a waypoint, trajectory planning is performed again through the vehicle lateral trajectory planning steps to obtain the reversal trajectory. Here, setting a reversal trajectory reference point can be used to avoid path conflicts between the reversal trajectory and the vehicle corresponding to the abnormal lane change state (e.g., a second vehicle).

[0088] Step 108: Control the current vehicle to return to the original lane according to the reversing trajectory, and send a lane change waiting prompt message to the driver.

[0089] In some embodiments, the aforementioned executing entity can control the current vehicle to return to its original lane according to the reversing trajectory and issue a lane change waiting prompt to the driver. This lane change waiting prompt can be used to provide the user with a voice prompt indicating a failed lane change and requesting them to wait for another opportunity to change lanes.

[0090] In addition, if the lane change environment monitoring results indicate that the lane change is normal, the lane change operation will continue to be performed, and a lane change completion prompt message will be issued to the driver after the lane change is completed.

[0091] The various embodiments of this disclosure have the following beneficial effects: the vehicle-interaction-based digital information retrieval method of some embodiments of this disclosure can improve the robustness and safety of assisted driving systems in complex urban scenarios. Specifically, the reason for the reduced robustness and safety of assisted driving systems in complex urban scenarios is that the "extensive" information processing method has obvious bottlenecks: First, it leads to excessive system computational load and delayed decision response, i.e., low information processing efficiency; second, in complex and dynamic traffic flows, there is more redundant information that is difficult to filter, which can easily interfere with the core decision-making process. For example, when generating lateral trajectories for vehicles, unnecessary computational disturbances are introduced, which not only affect planning efficiency but may also endanger driving safety. Therefore, it seriously restricts the decision-making efficiency and trajectory planning accuracy of assisted driving systems. Based on this, the vehicle-interaction-based digital information retrieval method of some embodiments of this disclosure first, in response to receiving a driver's voice lane change command, retrieves the vehicle perception information set corresponding to the driver's voice lane change command, and filters the vehicle perception information set based on the driver's voice lane change command to obtain the target vehicle perception information set. Here, by filtering the retrieved vehicle perception information set, the amount of data that the assisted driving system needs to process is greatly reduced, overcoming the problem of information overload. Simultaneously, by combining the driver's voice lane-change command with vehicle interaction, precise filtering of massive amounts of data can be achieved, thereby improving the system's processing efficiency. Then, vehicle information recognition is performed on each target vehicle perception information in the aforementioned target vehicle perception information set to generate a vehicle identification information set. Based on this vehicle identification information set, lane-change timing information is generated. Here, by recognizing vehicle information and generating lane-change timing information, precise control of lane-change timing is possible. Next, based on the aforementioned lane-change timing information, lateral trajectory planning is performed on the current vehicle to generate its lateral trajectory. Here, based on the lane-change timing condition, lateral trajectory planning generates a more accurate lateral trajectory for the current vehicle. This improves the decision-making efficiency and trajectory planning accuracy of the assisted driving system. Finally, based on the current lateral trajectory, the system controls the current vehicle to perform a lane-change operation and monitors the lane-change environment in real time to generate lane-change environment monitoring results. Here, real-time monitoring of the lane-changing environment allows for timely identification of situations that do not conform to lane-changing rules (e.g., a narrowing lane-changing interval or other vehicles entering the lane-changing interval). Finally, in response to the abnormal lane-changing status indicated by the aforementioned lane-changing environment monitoring results, a reversal trajectory is generated based on the current vehicle's lateral planning trajectory. The vehicle is then controlled to return to its original lane according to this reversal trajectory, and a lane-changing waiting prompt is issued to the driver. By generating the reversal trajectory and controlling the vehicle to return to its original lane, the vehicle can temporarily avoid risks, significantly improving the robustness and safety of the assisted driving system in complex urban scenarios.

[0092] Further reference Figure 4 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of a digital information retrieval device based on vehicle interaction. These device embodiments are similar to... Figure 1 Corresponding to the method embodiments shown, this vehicle-interaction-based digital information retrieval device can be specifically applied to various electronic devices.

[0093] like Figure 4 As shown, a vehicle-interactive digital information retrieval device 400 according to some embodiments includes: a retrieval unit 401, a filtering unit 402, a vehicle information recognition unit 403, a first generation unit 404, a vehicle lateral trajectory planning unit 405, a control and monitoring unit 406, a second generation unit 407, and a control and prompting unit 408. The retrieval unit 401 is configured to, in response to receiving a driver's voice lane change command, retrieve a set of vehicle perception information corresponding to the driver's voice lane change command; the filtering unit 402 is configured to, based on the driver's voice lane change command, filter the set of vehicle perception information to obtain a target vehicle perception information set; the vehicle information recognition unit 403 is configured to, perform vehicle information recognition on each target vehicle perception information in the target vehicle perception information set to generate a vehicle recognition information set; the first generation unit 404 is configured to, based on the vehicle recognition information set, generate lane change timing information; and the vehicle lateral trajectory planning unit 405 is configured to, based on... Based on the aforementioned lane change timing information, the current vehicle's lateral trajectory is planned to generate the current vehicle's lateral planning trajectory; the control and monitoring unit 406 is configured to control the current vehicle to perform a lane change operation based on the aforementioned current vehicle's lateral planning trajectory, and to perform real-time monitoring of the lane change environment to generate a lane change environment monitoring result; the second generation unit 407 is configured to respond to the lane change environment monitoring result indicating an abnormal lane change state, and to generate a reversal trajectory based on the aforementioned current vehicle's lateral planning trajectory; the control and prompting unit 408 is configured to control the current vehicle to return to the original lane according to the aforementioned reversal trajectory, and to issue a lane change waiting prompt message to the driver.

[0094] It is understandable that the units described in the vehicle-interaction-based digital information retrieval device 400 are related to the reference... Figure 1 The steps in the described method correspond to each other. Therefore, the operations, features, and beneficial effects described above for the method also apply to the vehicle-interaction-based digital information retrieval device 400 and the units contained therein, and will not be repeated here.

[0095] The following is for reference. Figure 5 It illustrates a schematic diagram of the structure of an electronic device (such as a computing device) suitable for implementing some embodiments of the present disclosure. Figure 5The electronic device shown is merely an example and should not be construed as limiting the functionality or scope of the embodiments of this disclosure. Figure 5 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The memory may include a non-volatile storage medium and internal memory. The non-volatile storage medium may store an operating system and a computer program. The computer program includes program instructions that, when executed, cause the processor to perform any of the methods described above. The processor provides computational and control capabilities to support the operation of the entire computer device. The internal memory provides an environment for the execution of the computer program in the non-volatile storage medium; when executed by the processor, the computer program causes the processor to perform any of the methods described above. The network interface is used for network communication, such as sending assigned tasks. Those skilled in the art will understand that... Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present disclosure and does not constitute a limitation on the computer device to which the present disclosure is applied. A specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0096] It should be understood that the processor can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among these, a general-purpose processor can be a microprocessor or any conventional processor.

[0097] In one embodiment, the processor is configured to run a computer program stored in a memory to perform the following steps: in response to receiving a driver's voice lane change command, calling a vehicle perception information set corresponding to the driver's voice lane change command; filtering the vehicle perception information set based on the driver's voice lane change command to obtain a target vehicle perception information set; performing vehicle information recognition on each target vehicle perception information in the target vehicle perception information set to generate a vehicle recognition information set; generating lane change timing information based on the vehicle recognition information set; performing lateral trajectory planning for the current vehicle based on the lane change timing information to generate a lateral planning trajectory for the current vehicle; controlling the current vehicle to perform a lane change operation based on the lateral planning trajectory for the current vehicle, and performing real-time monitoring of the lane change environment for the current vehicle to generate a lane change environment monitoring result; in response to the lane change environment monitoring result indicating an abnormal lane change state, generating a reversal trajectory based on the lateral planning trajectory for the current vehicle; controlling the current vehicle to return to the original lane according to the reversal trajectory, and issuing a lane change waiting prompt message to the driver.

[0098] This disclosure also provides a computer-readable storage medium storing a computer program, the computer program including program instructions, and the method implemented when the program instructions are executed can be referred to the various embodiments of the methods described above.

[0099] The aforementioned computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiments, such as the hard disk or memory of the computer device. Alternatively, the aforementioned computer-readable storage medium may be an external storage device of the computer device, such as a plug-in hard disk, SmartMedia Card (SMC), Secure Digital (SD) card, or Flash Card equipped on the computer device.

[0100] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0101] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A method for retrieving digital information based on vehicle interaction, applied to an assisted driving system, characterized in that, include: In response to receiving a driver's voice lane change command, the vehicle perception information set corresponding to the driver's voice lane change command is invoked; Based on the driver's voice lane change command, the vehicle perception information set is filtered to obtain the target vehicle perception information set; Vehicle information identification is performed on each target vehicle perception information in the target vehicle perception information set to generate a vehicle identification information set; Based on the vehicle identification information set, lane change timing information is generated; Based on the lane change timing information, the vehicle's lateral trajectory is planned to generate the current vehicle's lateral trajectory. Based on the current vehicle's lateral planning trajectory, the system controls the current vehicle to perform a lane change operation and monitors the lane change environment in real time to generate lane change environment monitoring results. This includes: sampling the current vehicle's lateral planning trajectory to obtain a trajectory sampling point sequence; performing real-time vehicle state detection on the first and second vehicles to generate a first state detection feature vector sequence and a second state detection feature vector sequence, where the first and second vehicles are two target vehicles on the target lane, the target lane is the lane adjacent to the current lane corresponding to the lane change direction, and the lane change direction corresponds to the driver's voice lane change command; and using a pre-constructed trajectory elastic deformation model, the first state detection feature vector sequence, and the second state detection feature vector sequence, determining the elastic strength value of each trajectory sampling point in the trajectory sampling point sequence to obtain an elastic strength value sequence, where the trajectory elastic deformation model is expressed by the following formula: ； in, Indicates the relationship with the first The first vehicle elasticity value corresponding to each trajectory sampling point; Indicates the relationship with the first The second vehicle elasticity value corresponding to each trajectory sampling point; This represents the first scaling factor, corresponding to the first vehicle, which is used to control the degree of influence of the state of the first vehicle on the trajectory sampling points; The second scaling factor corresponds to the second vehicle and is used to control the degree of influence of the state of the second vehicle on the trajectory sampling points; the first scaling factor is less than the second scaling factor; the first scaling factor and the second scaling factor belong to [0, 1] and their sum is 1; This represents the first weight vector, including the weights: ; This represents the second weight vector, including the weights: ; In response to the absence of an elastic strength value satisfying a preset elastic mutation condition in the elastic strength value sequence, the current vehicle's lateral planning trajectory is elastically adjusted using the elastic strength value sequence to obtain an adjusted planning trajectory. After the elastic adjustment, the current vehicle's lateral planning trajectory is fitted with the adjusted trajectory sampling points to obtain the adjusted planning trajectory. Real-time monitoring of the lane-changing environment of the current vehicle continues to generate lane-changing environment monitoring results or lane-changing completion indicators that characterize lane-changing anomalies. In response to the lane change environment monitoring results indicating an abnormal lane change state, a reversal trajectory is generated based on the current vehicle's lateral planning trajectory. Control the current vehicle to return to the original lane according to the described reversal trajectory, and issue a lane change waiting prompt message to the driver.

2. The method according to claim 1, characterized in that, The step of filtering the vehicle perception information set based on the driver's voice lane change command to obtain a target vehicle perception information set includes: The driver's voice lane change command is subjected to voice recognition to obtain command keywords, wherein the command keywords are used to represent the direction of lane change; Determine the positional relationship between the vehicle corresponding to each vehicle perception information in the vehicle perception information set and the current vehicle to obtain the positional relationship information set; Based on the instruction keywords and the location relationship information set, the vehicle perception information set is filtered to obtain the target vehicle perception information set.

3. The method according to claim 2, characterized in that, in, The target vehicle perception information includes: perception image sequences, point cloud data, radar perception data, and vehicle identification. The step of performing vehicle information identification on each piece of target vehicle perception information in the target vehicle perception information set to generate a vehicle identification information set includes: For each target vehicle perception information in the target vehicle perception information set, the following generation steps are performed: Vehicle size information is obtained by using the perceived image sequence and point cloud data included in the target vehicle perception information; Based on the target vehicle perception information, including radar perception data and point cloud data, vehicle driving information and vehicle distance values ​​are determined, wherein the vehicle driving information includes: speed value, acceleration value and angular velocity value. The vehicle size information, vehicle driving information, and vehicle distance value are determined as the vehicle identification information in the vehicle identification information set.

4. The method according to claim 3, characterized in that, The step of generating lane change timing information based on the vehicle identification information set includes: In the current vehicle coordinate system, based on the vehicle distance value and vehicle size information included in the vehicle identification information set, the vehicle interval between each target vehicle in the target lane is determined to obtain a vehicle interval sequence. The target lane is the lane adjacent to the current lane corresponding to the lane change direction, and each vehicle interval corresponds to the vehicle identifiers of two target vehicles. Vehicle identification information that meets the preset screening conditions in the vehicle identification information set is identified as target vehicle identification information, and a target vehicle identification information group is obtained. The vehicle interval in the vehicle interval sequence corresponding to the target vehicle identification information is greater than the preset interval threshold, and the speed value and acceleration value included meet the preset lane change conditions. Continuous behavior detection is performed on the target vehicles corresponding to each target vehicle identification information in the target vehicle identification information group to generate a vehicle spacing feature value group; The vehicle interval corresponding to the largest vehicle spacing feature value in the vehicle spacing feature value group and the vehicle identification information of the two target vehicles are determined as lane change timing information, wherein the two target vehicles are the first vehicle and the second vehicle, respectively.

5. The method according to claim 4, characterized in that, The step of planning the lateral trajectory of the current vehicle based on the lane change timing information to generate the lateral trajectory of the current vehicle includes: Based on the vehicle intervals in the lane change timing information, determine the coordinates of the route planning endpoint; Based on the current vehicle location coordinates, the path planning endpoint coordinates, and the lane change timing information, the vehicle's lateral trajectory is planned to generate the current vehicle's lateral trajectory.

6. The method according to claim 5, characterized in that, The real-time monitoring of the lane-changing environment of the current vehicle to generate lane-changing environment monitoring results includes: Determine the lane change zone between the first vehicle and the second vehicle; Retrieve real-time vehicle perception information of target vehicles adjacent to the lane change section to obtain a real-time vehicle perception information set. The status of each real-time vehicle perception information in the real-time vehicle perception information set is tracked to generate a real-time vehicle status information set, wherein the real-time vehicle status information includes the real-time vehicle position coordinates, the real-time vehicle speed value, the real-time vehicle acceleration value, and the vehicle movement status identifier. In response to the determination that there is real-time vehicle status information in the real-time vehicle status information set that meets the conditions for lane change conflict, a lane change environment monitoring result characterizing the lane change anomaly is generated.

7. A digital information retrieval device based on vehicle interaction, applied to an assisted driving system, characterized in that, include: The invocation unit is configured to invoke the vehicle perception information set corresponding to the driver's voice lane change command in response to receiving the driver's voice lane change command. The filtering unit is configured to filter the vehicle perception information set based on the driver's voice lane change command to obtain a target vehicle perception information set. The vehicle information recognition unit is configured to perform vehicle information recognition on each target vehicle perception information in the target vehicle perception information set to generate a vehicle recognition information set. The first generation unit is configured to generate lane change timing information based on the vehicle identification information set; The vehicle lateral trajectory planning unit is configured to perform vehicle lateral trajectory planning on the current vehicle based on the lane change timing information, so as to generate the current vehicle lateral planning trajectory. The control and monitoring unit is configured to control the current vehicle to perform a lane-changing operation based on the current vehicle's lateral planning trajectory, and to perform real-time monitoring of the lane-changing environment to generate lane-changing environment monitoring results. This includes: sampling the current vehicle's lateral planning trajectory to obtain a trajectory sampling point sequence; performing real-time vehicle state detection on a first vehicle and a second vehicle to generate a first state detection feature vector sequence and a second state detection feature vector sequence, wherein the first vehicle and the second vehicle are two target vehicles on a target lane, the target lane is a lane adjacent to the current lane corresponding to the lane-changing direction, and the lane-changing direction corresponds to the driver's voice lane-changing command; and using a pre-constructed trajectory elastic deformation model, the first state detection feature vector sequence, and the second state detection feature vector sequence, determining the elastic strength value of each trajectory sampling point in the trajectory sampling point sequence to obtain an elastic strength value sequence, wherein the trajectory elastic deformation model is expressed by the following formula: ； in, Indicates the relationship with the first The first vehicle elasticity value corresponding to each trajectory sampling point; Indicates the relationship with the first The second vehicle elasticity value corresponding to each trajectory sampling point; This represents the first scaling factor, corresponding to the first vehicle, which is used to control the degree of influence of the state of the first vehicle on the trajectory sampling points; The second scaling factor corresponds to the second vehicle and is used to control the degree of influence of the state of the second vehicle on the trajectory sampling points; the first scaling factor is less than the second scaling factor; the first scaling factor and the second scaling factor belong to [0, 1] and their sum is 1; This represents the first weight vector, including the weights: ; This represents the second weight vector, including the weights: ; In response to the absence of an elastic strength value satisfying a preset elastic mutation condition in the elastic strength value sequence, the current vehicle's lateral planning trajectory is elastically adjusted using the elastic strength value sequence to obtain an adjusted planning trajectory. After the elastic adjustment, the current vehicle's lateral planning trajectory is fitted with the adjusted trajectory sampling points to obtain the adjusted planning trajectory. Real-time monitoring of the lane-changing environment of the current vehicle continues to generate lane-changing environment monitoring results or lane-changing completion indicators that characterize lane-changing anomalies. The second generation unit is configured to generate a reversal trajectory based on the current vehicle lateral planning trajectory in response to the lane change environment monitoring results characterizing an abnormal lane change state. The control and prompting unit is configured to control the current vehicle to return to the original lane according to the said reversal trajectory and to issue a lane change waiting prompt message to the driver.

8. An electronic device, characterized in that, include: One or more processors; Storage device, on which one or more programs are stored, When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1-6.

9. A computer-readable medium, characterized in that, It stores a computer program thereon, wherein the program, when executed by a processor, implements the method as described in any one of claims 1-6.

Images