Interaction method and interaction device

Abstract

Embodiments of the present application disclose an interaction method and an interaction device. After a trajectory sequence of a target task is acquired, if it is determined that the trajectory sequence deviates, associated information of a navigation path of the target task is presented in a navigation page according to a target path planning result of the target task. In the embodiments of the present application, the target navigation path is determined according to a deviation probability of the target task, and the deviation probability is used to represent a probability that the trajectory sequence deviates from a main road segment of a predetermined road to an auxiliary road segment of the predetermined road, or a probability that the trajectory sequence deviates from the auxiliary road segment of the predetermined road to the main road segment of the predetermined road. Therefore, the embodiments of the present application can accurately identify the possibility of the vehicle deviating, and update the navigation path in time according to the deviation probability, thereby improving the navigation accuracy.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and more specifically to an interaction method and an interaction device. Background Technology

[0002] In daily life, users can use navigation apps installed on their devices to travel to different locations. Taking driving with the device as an example, the roads a vehicle can travel on include main roads (i.e., main road segments) and auxiliary roads (i.e., auxiliary road segments). However, due to factors such as low positioning accuracy of the device's positioning system and positioning drift (i.e., a mismatch between the user's actual location and the positioning information), the positioning system may not be able to accurately distinguish whether the road the vehicle is traveling on is actually a main road or an auxiliary road. Therefore, when the vehicle veers off course (i.e., veers from a main road to an auxiliary road or vice versa), the device may not be able to recognize this, making it impossible for the navigation app to provide accurate navigation information and increasing the possibility of the user taking a detour. Summary of the Invention

[0003] In view of this, the purpose of this invention is to provide an interaction method and an interaction device for accurately identifying the possibility of a vehicle veerging, thereby improving navigation accuracy.

[0004] According to a first aspect of the present invention, an interaction method is provided, the method comprising:

[0005] Obtain the trajectory sequence of the target task, wherein the trajectory sequence is a sequence of multiple coordinate points;

[0006] In response to determining that the trajectory sequence has deviated, the navigation path association information of the target task is presented on the navigation page according to the target path planning result of the target task. The deviating behavior is determined according to the deviating probability of the target task. The deviating probability is used to characterize the probability that the trajectory sequence deviates from the target road segment to a non-target road segment. The target road segment is the main road segment or auxiliary road segment of the predetermined road, and the non-target road segment is the auxiliary road segment or main road segment of the predetermined road.

[0007] According to a second aspect of the present invention, an interaction method is provided, the method comprising:

[0008] In response to receiving a trajectory sequence of a target task, the target road segment corresponding to the trajectory sequence is determined based on the current path planning result corresponding to the target task. The trajectory sequence is a sequence of multiple coordinate points, and the target road segment is a main road segment or an auxiliary road segment of a predetermined road.

[0009] Based on the road segment characteristics of the target road segment and the coordinate point characteristics of each coordinate point, and based on a pre-trained yaw probability prediction model, the yaw probability corresponding to the trajectory sequence is determined. The yaw probability is used to characterize the probability that the trajectory sequence deviates from the target road segment to a non-target road segment. The non-target road segment is the auxiliary road segment or the main road segment of the predetermined road.

[0010] In response to the yaw probability satisfying the first probability condition, path planning is performed based on the road segment identifier of the non-target road segment and the target location to determine the target path planning result, wherein the target location is the termination location of the target task;

[0011] Send the target path planning results.

[0012] According to a third aspect of the present invention, an interactive device is provided, the device comprising:

[0013] A sequence acquisition unit is used to acquire the trajectory sequence of the target task, wherein the trajectory sequence is a sequence of multiple coordinate points;

[0014] A path update unit is configured to, in response to determining that the trajectory sequence has deviated, present the associated information of the navigation path of the target task in the navigation page according to the target path planning result of the target task. The deviating behavior is determined according to the deviating probability of the target task. The deviating probability is used to characterize the probability that the trajectory sequence deviates from the target road segment to a non-target road segment. The target road segment is the main road segment or auxiliary road segment of a predetermined road, and the non-target road segment is the auxiliary road segment or main road segment of the predetermined road.

[0015] According to a fourth aspect of the present invention, an interactive device is provided, the device comprising:

[0016] A road segment determination unit is used to respond to receiving a trajectory sequence of a target task and determine the target road segment corresponding to the trajectory sequence based on the current path planning result corresponding to the target task. The trajectory sequence is a sequence of multiple coordinate points, and the target road segment is a main road segment or an auxiliary road segment of a predetermined road.

[0017] The probability prediction unit is used to determine the yaw probability corresponding to the trajectory sequence based on the road segment characteristics of the target road segment and the coordinate point characteristics of each coordinate point, and on a pre-trained yaw probability prediction model. The yaw probability is used to characterize the probability that the trajectory sequence deviates from the target road segment to a non-target road segment, where the non-target road segment is an auxiliary road segment or a main road segment of the predetermined road.

[0018] The planning result determination unit is used to respond to the yaw probability satisfying the first probability condition, perform path planning based on the road segment identifier of the non-target road segment and the target location, and determine the target path planning result, wherein the target location is the termination location of the target task;

[0019] The planning result sending unit is used to send the target path planning result.

[0020] According to a fifth aspect of the present invention, a computer-readable storage medium is provided that stores computer program instructions thereon, wherein the computer program instructions, when executed by a processor, implement the method as described in the first aspect.

[0021] According to a sixth aspect of the present invention, an electronic device is provided, including a memory and a processor, wherein the memory is configured to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method as described in the first aspect.

[0022] According to a seventh aspect of the present invention, a computer program product is provided, comprising a computer program / instructions, wherein the computer program / instructions are executed by a processor to implement the method as described in the first aspect.

[0023] After obtaining the trajectory sequence of the target task in this embodiment of the invention, if it is determined that the trajectory sequence has deviated, the associated information of the navigation path of the target task is presented on the navigation page according to the target path planning result of the target task. In this embodiment of the invention, the target navigation path is determined based on the deviation probability of the target task, and the deviation probability is used to characterize the probability that the trajectory sequence deviates from the main road segment of the predetermined road to the auxiliary road segment of the predetermined road, or from the auxiliary road segment of the predetermined road to the main road segment of the predetermined road. Therefore, this embodiment of the invention can accurately identify the possibility of the vehicle deviating, and update the navigation path in a timely manner according to the deviation probability, thereby improving navigation accuracy. Attached Figure Description

[0024] The above and other objects, features and advantages of the present invention will become clearer from the following description of embodiments of the invention with reference to the accompanying drawings, in which:

[0025] Figure 1 This is a schematic diagram of the hardware system architecture according to an embodiment of the present invention;

[0026] Figure 2 This is a flowchart of the interaction method of the first embodiment of the present invention;

[0027] Figure 3 This is a flowchart of the interaction method according to the second embodiment of the present invention;

[0028] Figure 4This is a flowchart of determining the yaw tag in an optional implementation of the second embodiment of the present invention;

[0029] Figure 5 This is a schematic diagram of an existing road segment distribution;

[0030] Figure 6 This is a flowchart of determining the yaw tag in another optional implementation of the second embodiment of the present invention;

[0031] Figure 7 This is a flowchart of determining the yaw tag in another optional implementation of the second embodiment of the present invention;

[0032] Figure 8 This is a schematic diagram of an interface according to an embodiment of the present invention;

[0033] Figure 9 This is another interface diagram of an embodiment of the present invention;

[0034] Figure 10 This is a flowchart of the interaction method of the second embodiment of the present invention on the server side;

[0035] Figure 11 This is a flowchart of the interaction method of the second embodiment of the present invention on the terminal side;

[0036] Figure 12 This is a flowchart of the interaction method according to the third embodiment of the present invention;

[0037] Figure 13 This is another interface diagram of an embodiment of the present invention;

[0038] Figure 14 This is a flowchart of the interaction method on the server side according to the third embodiment of the present invention;

[0039] Figure 15 This is a flowchart of the interaction method of the third embodiment of the present invention on the terminal side;

[0040] Figure 16 This is a schematic diagram of the interactive system according to the fourth embodiment of the present invention;

[0041] Figure 17 This is a schematic diagram of an electronic device according to the fifth embodiment of the present invention. Detailed Implementation

[0042] The present invention is described below based on embodiments, but the invention is not limited to these embodiments. In the detailed description of the invention below, certain specific details are described in detail. Those skilled in the art will fully understand the invention even without these details. To avoid obscuring the essence of the invention, well-known methods, processes, flows, elements, and circuits are not described in detail.

[0043] Furthermore, those skilled in the art should understand that the accompanying drawings provided herein are for illustrative purposes only and are not necessarily drawn to scale.

[0044] Unless the context explicitly requires it, words such as "including" or "contains" in the instruction manual should be interpreted as including rather than exclusive or exhaustive; that is, meaning "including but not limited to".

[0045] In the description of this invention, it should be understood that the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Furthermore, in the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0046] This invention is applicable to user-driven vehicle scenarios. In this embodiment, navigation is used as the target task for illustration. However, those skilled in the art will readily understand that the method of this invention is equally applicable when the target task is other tasks, such as ride-hailing or order delivery.

[0047] Due to factors such as low positioning accuracy of the terminal's positioning system and positioning drift (i.e., a mismatch between the user's actual location and the positioning information), the positioning system may not be able to accurately distinguish whether the road the vehicle is traveling on is actually a main road or a secondary road when driving on any road. Therefore, when the vehicle veers off course (i.e., veers from a main road to a secondary road or vice versa), the terminal may not be able to recognize it, thus navigation clients may not be able to provide accurate navigation information to the user. Main roads usually do not have connecting road segments, while secondary roads have at least one connecting road segment, allowing vehicles to navigate to at least one connecting road segment. Therefore, if the vehicle veers off course, it greatly increases the likelihood that the user will take a detour.

[0048] Figure 1 This is a schematic diagram of the hardware system architecture of an embodiment of the present invention. Figure 1 The hardware system architecture shown may include at least one terminal 11 and at least one server 12. Figure 1The following description uses a terminal 11 and a server 12 as an example. Terminal 11 and server 12 can establish a communication connection via a network. In this embodiment, terminal 11 can be a device fixedly installed inside a vehicle (not shown in the figure), such as a navigation system, or other existing devices such as mobile phones, tablets, personal computers, etc. While the vehicle is moving, terminal 11 can obtain its own positioning information as the vehicle's location information through a positioning system (e.g., GPS (Global Positioning System), BeiDou Navigation Satellite System, etc.), and provide navigation services to users through a navigation client installed on terminal 11.

[0049] In this embodiment of the invention, after the user sets the starting position (i.e., departure point) and target position (i.e., destination) through the terminal 11, the terminal 11 can generate a target task based on the user-set starting and target positions, and use at least one location information (i.e., coordinate points) collected according to a predetermined period as the vehicle's trajectory sequence. Then, the terminal 11 can upload the trajectory sequence of the target task to the server 12. After receiving the trajectory sequence of the target task, the server 12 can determine the main road segment or auxiliary road segment (i.e., target road segment) of the predetermined road corresponding to the trajectory sequence based on the current path planning result corresponding to the target task, and determine the yaw probability corresponding to the trajectory sequence based on the road segment characteristics of the target road segment and the coordinate point characteristics of each coordinate point in the trajectory sequence, based on a pre-trained yaw probability prediction model, that is, the probability that the trajectory sequence deviates from the target road segment to the auxiliary road segment or main road segment of the predetermined road (i.e., non-target road segment). Then, when the yaw probability meets the first probability condition, the server performs path planning based on the road segment identifier of the non-target road segment and the target position, determines the target path planning result, and sends the target path planning result to the terminal 11. After receiving the target path planning result, terminal 11 updates the navigation path of the target task on the navigation page according to the target path planning result.

[0050] In an optional implementation of this invention, server 12 may send the yaw probability of the target task to terminal 11. If the yaw probability meets a first probability condition, terminal 11 may display a pre-defined pop-up window to indicate that the navigation path has changed, and upon receiving a path restoration instruction, restore the navigation path to the previous path planning result.

[0051] In another optional implementation of this invention, if the yaw probability satisfies the second probability condition, terminal 11 can render and display a location confirmation control to confirm whether the current road segment is the target road segment. If a route update instruction is received, terminal 11 can generate and send a route update request to server 12 based on the road segment identifier and target location of the current road segment. Server 12 performs route planning based on the road segment identifier and target location in the route update request, determines the target route planning result, and sends the target route planning result to terminal 11. After receiving the target route planning result, terminal 11 updates the navigation path of the target task on the navigation page according to the target route planning result.

[0052] In another optional implementation of the present invention, if the yaw probability satisfies the second probability condition and the terminal 11 receives a path-keeping instruction, or does not receive a path-update instruction within a predetermined time period, the navigation path can be kept unchanged.

[0053] The method of this invention can be implemented independently by a terminal, a server, or by interaction between the terminal and the server. The embodiments of this invention will be described in detail below through method examples. Figure 2 This is a flowchart of the interaction method of the first embodiment of the present invention, which is illustrated using a terminal implementation as an example. Figure 2 As shown, the method in this embodiment includes the following steps:

[0054] Step S100: Obtain the trajectory sequence of the target task.

[0055] The terminal can obtain its own coordinates as location information through a positioning system. Therefore, with user authorization, a navigation client can obtain a trajectory sequence composed of multiple coordinate points collected by the terminal at a predetermined period. After generating a navigation task determined based on the user-set starting and target positions, the client sends the trajectory sequence of the navigation task to the server. In this embodiment, the terminal can send the trajectory sequence collected within the current time period to the server every certain period of time (e.g., 15 seconds).

[0056] In one optional implementation, after obtaining the trajectory sequence and the current path planning result of the target task, the terminal can determine the current road segment (i.e., the current road segment) based on the trajectory sequence, and determine the target road segment corresponding to the trajectory sequence based on the current path planning result. If the current road segment and the target road segment do not match, the terminal can send the trajectory sequence to the server so that the server can determine the probability of the vehicle deviating from its course. Here, the current road segment is either a secondary or primary road segment of the predetermined road; the target road segment is either a primary or secondary road segment of the predetermined road; the current path planning result can be represented as starting position -> road segment 1 -> ... road segment n -> target position, where n is a predetermined integer greater than or equal to 1.

[0057] Specifically, the terminal can determine the current road segment in the trajectory sequence when a predetermined number of consecutive coordinate points in the trajectory sequence correspond to the same road segment, or when the ratio of the number of coordinate points corresponding to the same road segment in the trajectory sequence to the total number of coordinate points in the trajectory sequence is higher than (or not lower than) a first threshold. This reduces the negative impact of positioning drift on determining the current road segment. The terminal can perform road segment binding processing on coordinate points using various existing road segment binding methods to determine the road segment corresponding to the coordinate points. For example, a model can be trained based on the coordinate points, positioning accuracy, movement speed, azimuth angle of historical candidate road segments, road segment level (i.e., national highway, provincial highway, municipal highway, etc.), road segment type (i.e., main road or auxiliary road) of different terminals in historical data, as well as matching labels (used to characterize whether historical candidate road segments are road segments actually traversed by the terminal). Based on the trained model, the road segment corresponding to each coordinate point can be determined according to each coordinate point in the trajectory sequence, the terminal's positioning accuracy, movement speed, and the azimuth angle, road segment level, and road segment type of the candidate road segments. Among them, the candidate road segment is the road segment whose shortest distance to the coordinate point is less than (or does not exceed) the second threshold.

[0058] For example, if trajectory sequence A1 includes 10 coordinate points, and 5 consecutive coordinate points (i.e., a predetermined number) correspond to road segment L1, the terminal can determine that the current road segment corresponding to trajectory sequence A1 is road segment L1. As another example, if trajectory sequence A1 includes 10 coordinate points, and 7 of them correspond to road segment L1, the ratio of these 7 to the total number of coordinate points is higher than 60% (i.e., the first threshold), the terminal can determine that the current road segment corresponding to trajectory sequence A1 is road segment L1.

[0059] Simultaneously, the terminal can determine the target road segment of the trajectory sequence based on the position of at least one coordinate point in the trajectory sequence and the positions of each road segment in the current path planning result. It is easy to understand that in this embodiment, one trajectory sequence can correspond to multiple current road segments and multiple target road segments.

[0060] In step S200, in response to the determination that the trajectory sequence has deviated, the navigation path association information of the target task is presented on the navigation page according to the target path planning result of the target task.

[0061] In this embodiment, the terminal can determine whether the target trajectory sequence has deviated based on the deviance probability of the target task. If it is determined that the trajectory sequence has deviated, the terminal can present the associated information of the navigation path of the target task on the navigation page based on the path planning result of the target task. The target path planning result is obtained by path planning based on the segment identifiers of non-target road segments and the termination position of the target task (i.e., the target position).

[0062] (i) In the first possible case, if the yaw probability meets the first probability condition, the terminal can determine that the trajectory sequence has yawed and obtain the target path planning result, and update the target path on the navigation page according to the target path planning result.

[0063] To reduce the possibility of inaccurate predictions from probabilistic yaw models, the terminal may optionally display a pre-defined pop-up window to indicate a change in the navigation path.

[0064] Optionally, the terminal may display a predetermined pop-up window at a first position on the navigation page, wherein the first position is any position on the navigation page that does not obstruct the position indicator. Specifically, the display method of the predetermined pop-up window can be determined by the method described in other embodiments of the present invention, and will not be described in detail in this embodiment.

[0065] Optionally, the terminal can respond to receiving a route restoration command by restoring the navigation path to the previous route planning result. The route restoration command is triggered by the user. When the terminal receives the route restoration command, it indicates that the current road segment is the target road segment, therefore the terminal can restore the navigation path to the previous route planning result. Specifically, the terminal can render and display a route restoration control, and determine that a route restoration command has been received when the route restoration control is triggered. Alternatively, the terminal can also receive a first voice sequence sent by the user and perform intent recognition on the first voice sequence. If the recognition result of the first voice sequence is used to represent restoring the navigation path, it can be determined that a route restoration command has been received. Specifically, the display method of the route restoration control and the recognition method of the first voice sequence can be determined by the methods described in other embodiments of the present invention, and will not be described in detail in this embodiment.

[0066] (ii) In the second possible scenario, if the deviation probability meets the second probability condition, it means that the terminal cannot directly determine whether the trajectory sequence has deviated. The terminal can render and display a location confirmation control to confirm whether the current road segment is the target road segment. If a route update instruction is received, the terminal can determine that the trajectory sequence has deviated, obtain the target route planning result, and update the target route on the navigation page according to the target route planning result.

[0067] Optionally, the terminal can receive a second voice sequence sent by the user and perform intent recognition on the second voice sequence. If the recognition result of the second voice sequence is used to indicate that the current road segment does not match the target road segment, the terminal can determine that a path update instruction has been received.

[0068] Optionally, the location confirmation control may include a location update control, which is used to update the current road segment from a secondary road segment to a primary road segment of a predetermined road, or to update the current road segment from a primary road segment to a secondary road segment of a predetermined road. When the route update control is triggered, the terminal can also determine that a route update instruction has been received.

[0069] Specifically, the display method of the location confirmation control and the recognition method of the second voice sequence can be determined by the methods described in other embodiments of the present invention, and will not be described in detail in this embodiment.

[0070] In this embodiment, the terminal can predict the yaw probability using various existing methods. For example, the terminal can calculate the difference between the trajectory sequence and the current path planning result, which is the ratio of the difference between the length of the trajectory sequence (i.e., the first length) and the repetition length of the trajectory sequence and the current path planning result (i.e., the second length) to the length of the trajectory sequence (i.e., the first length), and use the difference as the yaw probability; or, based on the coordinate features of each coordinate point in the trajectory sequence, the terminal can input a pre-trained probability prediction model to predict the yaw probability, and the probability prediction model can be obtained through various methods.

[0071] In one optional implementation of this embodiment, the terminal can obtain the road segment features of the target road segment and the coordinate point features of each coordinate point, and predict the yaw probability corresponding to the trajectory sequence based on the road segment features of the target road segment and the coordinate point features of each coordinate point.

[0072] Optionally, the terminal can also determine the relative characteristics between the target road segment and each coordinate point based on the road segment characteristics and the coordinate point characteristics of each coordinate point, and / or determine the sequence characteristics of the trajectory sequence based on the coordinate point characteristics between each coordinate point. Thus, based on the road segment characteristics, the coordinate point characteristics, the relative characteristics between the target road segment and each coordinate point, and the sequence characteristics of the trajectory sequence, the terminal determines the yaw probability corresponding to the trajectory sequence using a yaw probability prediction model. It is easy to understand that if the trajectory sequence corresponds to multiple target road segments, the terminal can segment the trajectory sequence into trajectory sequence segments corresponding to each target road segment, and determine the yaw probability of the trajectory sequence based on the road segment characteristics of each target road segment, the coordinate point characteristics of the corresponding coordinate points, the relative characteristics between the target road segment and the corresponding coordinate points, and the sequence segment characteristics of each trajectory sequence segment.

[0073] The coordinate point features can include at least one of the following: the position of each coordinate point (specifically, the latitude and longitude values), the corresponding velocity, acceleration, and velocity direction. The velocity, acceleration, and direction can be acquired by sensors (including velocity sensors, acceleration sensors, etc.) configured in the terminal. The road segment features can include at least one of the following: the road segment level, the number of lanes, and the road segment direction. The relative features between the target road segment and its corresponding coordinate points can include at least one of the following: the directional angle difference between the target road segment and each coordinate point, and the distance between each coordinate point and the target road segment. The distance between a coordinate point and the target road segment can be the shortest distance between the coordinate point and the target road segment. The sequence features of the trajectory sequence can include at least one of the following: the directional angle difference between adjacent coordinate points, the difference in distance between adjacent coordinate points and the target road segment, and the sum of the absolute values ​​of the directional angle differences between adjacent coordinate points. It is easy to understand that, depending on actual needs, coordinate point features, road segment features, relative features, and sequence features can all include other features; this embodiment does not impose specific limitations.

[0074] The directional angle difference between the target road segment and the coordinate point can be calculated as follows. For example, if the target road segment is due east and the velocity direction of coordinate point p1 is 60° east of north, then the directional angle difference between the target road segment and coordinate point p1 is 30°.

[0075] In this embodiment, the yaw probability prediction model can be an existing model, such as xgboost (eXtremeGradient Boosting), Wide&Deep, RNN (Recurrent Neural Network), etc. Taking xgboost as an example, xgboost is a type of boosting tree model, which is lightweight, scalable, and distributed, and can perform regression predictions with high accuracy.

[0076] The yaw probability prediction model is trained based on a training sample set. In this embodiment, the training sample set is determined based on multiple historical trajectory sequences, preset movement segments corresponding to each historical trajectory sequence, and yaw tags. The preset movement segment is either a secondary or primary road segment of a specific road within the historical navigation path corresponding to the historical trajectory sequence, and the determination method for the preset movement segment is similar to that for the target road segment. The yaw tag is used to characterize whether the actual movement segment deviates from the preset movement segment. The actual movement segment is either a secondary or primary road segment of a specific road, and the determination method for the actual movement segment is similar to that for the current road segment.

[0077] In this embodiment, the yaw tag can be determined based on at least one of the road network connectivity relationship corresponding to the historical trajectory sequence and the image recognition result. The image recognition result is obtained by identifying the target image sequence, which is an image sequence synchronously acquired along with the corresponding historical trajectory sequence. Optionally, the target image sequence can be acquired by a terminal that reports the historical trajectory sequence, or by an image acquisition device (e.g., a dashcam) installed in a vehicle equipped with a terminal that reports the historical trajectory sequence. Determining the yaw tag based on at least one of the road network connectivity relationship and the image recognition result can effectively improve the accuracy of the yaw tag and reduce the cost of manual annotation, thereby effectively reducing the training cost and accuracy of the yaw probability prediction model.

[0078] Specifically, the method for determining the yaw tag can be determined by the method described in other embodiments of the present invention, which will not be described in detail in this embodiment.

[0079] Optionally, this embodiment may further include the following steps:

[0080] Step S300: In response to the determination that no deviation has occurred in the trajectory sequence, the navigation path remains unchanged.

[0081] In the second possible scenario, if the terminal receives a path-keeping instruction, or does not receive a path update instruction within a predetermined time period, the terminal can determine that the trajectory sequence has not deviated and maintain the navigation path unchanged.

[0082] Optionally, the terminal can receive a third voice sequence sent by the user and perform intent recognition on the third voice sequence. If the recognition result of the third voice sequence is used to represent that the current road segment matches the target road segment, the terminal can determine that a path-keeping instruction has been received.

[0083] Optionally, the location confirmation control may also include a location hold control, which is used to keep the current road segment as a main road segment of the predetermined road, or to keep the current road segment as a secondary road segment of the predetermined road. When the route hold control is triggered, the terminal can also determine that a route hold instruction has been received.

[0084] Specifically, the display method of the location confirmation control and the recognition method of the third voice sequence can be determined by the methods described in other embodiments of the present invention, which will not be described in detail in this embodiment.

[0085] In the third possible scenario, if the yaw probability meets the third probability condition, the terminal can determine that the trajectory sequence has not deviated and maintain the navigation path unchanged. The third probability condition can be that the yaw probability is not higher than the fourth threshold or is lower than the fourth threshold.

[0086] In this embodiment, after obtaining the trajectory sequence of the target task, if it is determined that the trajectory sequence has deviated, the navigation path association information of the target task is presented on the navigation page based on the target path planning result of the target task. In this embodiment, the target navigation path is determined based on the deviation probability of the target task, and the deviation probability is used to characterize the probability that the trajectory sequence deviates from the main road segment of the predetermined road to the auxiliary road segment of the predetermined road, or from the auxiliary road segment of the predetermined road to the main road segment of the predetermined road. Therefore, this embodiment can accurately identify the possibility of the vehicle deviating and update the navigation path in a timely manner according to the deviation probability, thereby improving navigation accuracy.

[0087] Figure 3 This is a flowchart of the interaction method according to the second embodiment of the present invention. Figure 3 As shown, the method in this embodiment is implemented through interaction between the terminal and the server, and includes the following steps:

[0088] Step S100': Obtain and send the trajectory sequence of the target task.

[0089] The terminal can obtain its own coordinates as location information through a positioning system. Therefore, with user authorization, a navigation client can obtain a trajectory sequence composed of multiple coordinate points collected by the terminal at a predetermined period. After generating a navigation task determined based on the user-set starting and target positions, the client sends the trajectory sequence of the navigation task to the server. In this embodiment, the terminal can send the trajectory sequence collected within the current time period to the server every certain period of time (e.g., 15 seconds).

[0090] In one optional implementation, after obtaining the trajectory sequence and the current path planning result of the target task, the terminal can determine the current road segment (i.e., the current road segment) based on the trajectory sequence, and determine the target road segment corresponding to the trajectory sequence based on the current path planning result. If the current road segment and the target road segment do not match, the terminal can send the trajectory sequence to the server so that the server can determine the probability of the vehicle deviating from its course. Here, the current road segment is either a secondary or primary road segment of the predetermined road; the target road segment is either a primary or secondary road segment of the predetermined road; the current path planning result can be represented as starting position -> road segment 1 -> ... road segment n -> target position, where n is a predetermined integer greater than or equal to 1.

[0091] Specifically, the terminal can determine the current road segment in the trajectory sequence when a predetermined number of consecutive coordinate points in the trajectory sequence correspond to the same road segment, or when the ratio of the number of coordinate points corresponding to the same road segment in the trajectory sequence to the total number of coordinate points in the trajectory sequence is higher than (or not lower than) a first threshold. This reduces the negative impact of positioning drift on determining the current road segment. The terminal can perform road segment binding processing on coordinate points using various existing road segment binding methods to determine the road segment corresponding to the coordinate points. For example, a model can be trained based on the coordinate points, positioning accuracy, movement speed, azimuth angle of historical candidate road segments, road segment level (i.e., national highway, provincial highway, municipal highway, etc.), road segment type (i.e., main road or auxiliary road) of different terminals in historical data, as well as matching labels (used to characterize whether historical candidate road segments are road segments actually traversed by the terminal). Based on the trained model, the road segment corresponding to each coordinate point can be determined according to each coordinate point in the trajectory sequence, the terminal's positioning accuracy, movement speed, and the azimuth angle, road segment level, and road segment type of the candidate road segments. Among them, the candidate road segment is the road segment whose shortest distance to the coordinate point is less than (or does not exceed) the second threshold.

[0092] For example, if trajectory sequence A1 includes 10 coordinate points, and 5 consecutive coordinate points (i.e., a predetermined number) correspond to road segment L1, the terminal can determine that the current road segment corresponding to trajectory sequence A1 is road segment L1. As another example, if trajectory sequence A1 includes 10 coordinate points, and 7 of them correspond to road segment L1, the ratio of these 7 to the total number of coordinate points is higher than 60% (i.e., the first threshold), the terminal can determine that the current road segment corresponding to trajectory sequence A1 is road segment L1.

[0093] Simultaneously, the terminal can determine the target road segment of the trajectory sequence based on the position of at least one coordinate point in the trajectory sequence and the positions of each road segment in the current path planning result. It is easy to understand that in this embodiment, one trajectory sequence can correspond to multiple current road segments and multiple target road segments.

[0094] Step S200': In response to receiving the trajectory sequence of the target task, determine the target road segment corresponding to the trajectory sequence based on the current path planning result corresponding to the target task.

[0095] In existing technologies, navigation tasks are typically generated on the server side. Therefore, the server can obtain the corresponding current path planning result based on the task identifier of the target task, and determine the target road segment corresponding to the trajectory sequence based on the current path planning result. In this step, the method for determining the target road segment can be the same as in step S100'.

[0096] Step S300': Based on the road segment characteristics of the target road segment and the coordinate point characteristics of each coordinate point, determine the yaw probability corresponding to the trajectory sequence based on the pre-trained yaw probability prediction model.

[0097] After determining the target road segment, the server can predict the yaw probability corresponding to the trajectory sequence based on the road segment characteristics and the coordinate point characteristics of each coordinate point. In this embodiment, the yaw probability is used to characterize the probability that the trajectory sequence deviates from the target road segment to a non-target road segment, where the non-target road segment is the auxiliary road or main road of the predetermined road.

[0098] Optionally, the server can also determine the relative characteristics between the target road segment and each coordinate point based on the road segment characteristics and the coordinate point characteristics of each coordinate point, and / or determine the sequence characteristics of the trajectory sequence based on the coordinate point characteristics between each coordinate point. Thus, based on the road segment characteristics, the coordinate point characteristics, the relative characteristics between the target road segment and each coordinate point, and the sequence characteristics of the trajectory sequence, the server determines the yaw probability corresponding to the trajectory sequence using a yaw probability prediction model. It is easy to understand that if the trajectory sequence corresponds to multiple target road segments, the server can segment the trajectory sequence into trajectory sequence segments corresponding to each target road segment, and determine the yaw probability of the trajectory sequence based on the road segment characteristics of each target road segment, the coordinate point characteristics of the corresponding coordinate points, the relative characteristics between the target road segment and the corresponding coordinate points, and the sequence segment characteristics of each trajectory sequence segment.

[0099] The coordinate point features can include at least one of the following: the position of each coordinate point (specifically, the latitude and longitude values), the corresponding velocity, acceleration, and velocity direction. The velocity, acceleration, and direction can be acquired by sensors (including velocity sensors, acceleration sensors, etc.) configured in the terminal. The road segment features can include at least one of the following: the road segment level, the number of lanes, and the road segment direction. The relative features between the target road segment and its corresponding coordinate points can include at least one of the following: the directional angle difference between the target road segment and each coordinate point, and the distance between each coordinate point and the target road segment. The distance between a coordinate point and the target road segment can be the shortest distance between the coordinate point and the target road segment. The sequence features of the trajectory sequence can include at least one of the following: the directional angle difference between adjacent coordinate points, the difference in distance between adjacent coordinate points and the target road segment, and the sum of the absolute values ​​of the directional angle differences between adjacent coordinate points. It is easy to understand that, depending on actual needs, coordinate point features, road segment features, relative features, and sequence features can all include other features; this embodiment does not impose specific limitations.

[0100] The directional angle difference between the target road segment and the coordinate point can be calculated as follows. For example, if the target road segment is due east and the velocity direction of coordinate point p1 is 60° east of north, then the directional angle difference between the target road segment and coordinate point p1 is 30°.

[0101] In this embodiment, the yaw probability prediction model can be an existing model, such as xgboost (eXtremeGradient Boosting), Wide&Deep, RNN (Recurrent Neural Network), etc. Taking xgboost as an example, xgboost is a type of boosting tree model, which is lightweight, scalable, and distributed, and can perform regression predictions with high accuracy.

[0102] The yaw probability prediction model is trained based on a training sample set. In this embodiment, the training sample set is determined based on multiple historical trajectory sequences, preset movement segments corresponding to each historical trajectory sequence, and yaw tags. The preset movement segment is either a secondary or primary road segment of a specific road within the historical navigation path corresponding to the historical trajectory sequence, and the determination method for the preset movement segment is similar to that for the target road segment. The yaw tag is used to characterize whether the actual movement segment deviates from the preset movement segment. The actual movement segment is either a secondary or primary road segment of a specific road, and the determination method for the actual movement segment is similar to that for the current road segment.

[0103] In this embodiment, the yaw tag can be determined based on at least one of the road network connectivity relationship corresponding to the historical trajectory sequence and the image recognition result. The image recognition result is obtained by identifying the target image sequence, which is an image sequence synchronously acquired along with the corresponding historical trajectory sequence. Optionally, the target image sequence can be acquired by a terminal that reports the historical trajectory sequence, or by an image acquisition device (e.g., a dashcam) installed in a vehicle equipped with a terminal that reports the historical trajectory sequence. Determining the yaw tag based on at least one of the road network connectivity relationship and the image recognition result can effectively improve the accuracy of the yaw tag and reduce the cost of manual annotation, thereby effectively reducing the training cost and accuracy of the yaw probability prediction model.

[0104] If the server can directly determine the actual movement segment corresponding to the historical trajectory sequence, the server can determine the deviation tag in the following way. Figure 4 This is a flowchart illustrating the determination of the yaw tag in an optional implementation of the second embodiment of the present invention. For example... Figure 4 As shown, in one optional implementation of this embodiment, the yaw tag can be determined in the following way:

[0105] Step S310'A: For each historical trajectory sequence, determine the first type of road segment connected to the actual moving road segment.

[0106] Figure 5 This is a schematic diagram of an existing road segment distribution. For example... Figure 5 As shown, the actual moving road segment is segment L1 (that is, Figure 5 The road segment extending longitudinally in the middle), road segment L2 and road segment L3 (that is, Figure 5 The section extending laterally along the middle is the section connected to section L1, which is also the first section.

[0107] In this step, the server can determine the first type of road segment based on road network connectivity, or it can determine the first type of road segment by performing image recognition on the target image sequence and combining it with road network connectivity. The road network connectivity corresponding to the historical trajectory sequence is the connectivity of the road network covered within the range of the historical trajectory sequence. The range of the historical trajectory sequence can be determined in various ways, such as using a circular range with the midpoint of the historical trajectory sequence as the center and a predetermined length as the radius as the range of the historical trajectory sequence, where the predetermined length is not less than the maximum value of the distance from the midpoint of the historical trajectory sequence to each coordinate point in the historical trajectory sequence.

[0108] Specifically, the road network connectivity relationship is the pre-determined connectivity relationship between each road segment. Therefore, when determining the first type of road segment based on the road network connectivity relationship, the server can directly determine at least one first type of road segment connected to the actual moving road segment based on the road segment identifier of the actual moving road segment and the road network connectivity relationship corresponding to the historical trajectory sequence.

[0109] When determining the first type of road segment by performing image recognition on the target image sequence and combining it with road network connectivity, the server can perform image recognition (specifically, target recognition) on the target image sequence to determine whether there is a road segment connected to the actual moving road segment. If a road segment connected to the actual moving road segment exists, the server can directly determine at least one first type of road segment connected to the actual moving road segment based on the road segment identifier of the actual moving road segment and the road network connectivity corresponding to the historical trajectory sequence. The server can perform image recognition on the target image sequence using various existing models; this embodiment does not impose specific limitations.

[0110] Step S3'A: Determine the connectivity between the preset mobile road segment and the first type of road segment based on the road network connectivity.

[0111] If the connectivity is not connected, the server can execute step S320'A; if the connectivity is connected, the server can execute step S330'A.

[0112] Step S320'A: Determine the yaw label to characterize the yaw that occurred in the historical trajectory sequence.

[0113] If the preset moving segment is not connected to at least one first-class segment, it means that it is impossible to travel directly from the preset moving segment to the unconnected first-class segment. In other words, the preset moving segment does not match the actual moving segment. Therefore, the server can determine that the yaw tag is used to characterize the deviation of the historical trajectory sequence. Optionally, the yaw tag can be set to 1.

[0114] Step S330'A: Determine the yaw tag to indicate that no yaw has occurred in the historical trajectory sequence.

[0115] If the preset moving route segment and each of the first-class road segments are not connected, it means that the preset moving route segment and the actual moving route segment are matched. Therefore, the server can determine that the yaw tag is used to represent that the historical trajectory sequence has not deviated. Optionally, the yaw tag can be set to 0.

[0116] If the server cannot directly determine the actual movement segment corresponding to the historical trajectory sequence, the server can determine the deviation tag in the following way. Figure 6 This is a flowchart illustrating the determination of the yaw tag in another optional implementation of the second embodiment of the present invention. For example... Figure 6 As shown, in another optional implementation of this embodiment, the yaw tag can be determined in the following way:

[0117] Step S310'B: Perform image recognition on the target image sequence to determine the first main and auxiliary road label corresponding to the actual moving road segment.

[0118] In one optional implementation of this embodiment, the target image sequence is an image sequence acquired within a predetermined range, wherein the predetermined range can be at least one of the range where the road segment intersection is located and the range where the road segment entrance and exit are located.

[0119] In this step, the server can use various existing image recognition models to perform image recognition on the target image sequence to determine the first main and auxiliary road label corresponding to the actual moving road segment. The first main and auxiliary road label is used to characterize whether the actual moving road segment is a main road segment or an auxiliary road segment. The image recognition model can be CNN (Convolutional Neural Networks), RNN, etc. In real life, the main and auxiliary road segments of the same road can usually be distinguished by features such as the number of lanes and the location of road segment entrances and exits. For example, the main road segment usually has more lanes than the auxiliary road segment, and the entrances and exits of the main road segment are usually on the right side of the main road segment while the entrances and exits of the auxiliary road segment are usually on the left side. Therefore, the main and auxiliary road labels of each sample image sequence used as training samples for the image recognition model can be determined based on the above features. The image recognition model is then trained based on each sample image sequence and its corresponding main and auxiliary road labels. Finally, the trained image recognition model is used to perform image recognition on the target image sequence to determine the first main and auxiliary road label corresponding to the actual moving road segment.

[0120] Step S320'B: Determine the second main and auxiliary road label corresponding to the preset moving road segment.

[0121] In this step, the server can determine the second primary / secondary road label corresponding to the preset mobile road segment based on the road segment identifier. The second primary / secondary road label is used to identify whether the preset mobile road segment is a primary or secondary road segment.

[0122] Step S3'B: Determine whether the first main and auxiliary road labels match the second main and auxiliary road labels.

[0123] If the first primary / secondary road label does not match the second primary / secondary road label, the server can execute step S330'; if the first primary / secondary road label matches the second primary / secondary road label, the server can execute step S340'.

[0124] Step S330'B: Determine the yaw tag to characterize the yaw that occurred in the historical trajectory sequence.

[0125] If the first main and auxiliary road label does not match the second main and auxiliary road label, it means that the actual moving road segment is not the preset moving road segment. The server can determine that the yaw label is used to characterize the deviation of the historical trajectory sequence. Optionally, the yaw label can be set to 1.

[0126] Step S340'B: Determine the yaw tag to indicate that no yaw has occurred in the historical trajectory sequence.

[0127] Typically, the terminal's location information will not deviate too far from the actual location. Therefore, if the first main and auxiliary road label matches the second main and auxiliary road label, the actual moving road segment can be considered as the preset moving road segment. The server can determine that the yaw label is used to represent that the historical trajectory sequence has not deviated. Optionally, the yaw label can be set to 0.

[0128] Figure 7 This is a flowchart illustrating the determination of the yaw tag in another optional implementation of the second embodiment of the present invention. For example... Figure 7 As shown, in one optional implementation of this embodiment, the yaw tag can be determined in the following way:

[0129] Step S310'C: Determine the yaw label of the corresponding first-type trajectory sequence based on the road network connectivity relationship corresponding to each first-type trajectory sequence.

[0130] In this implementation, the first type of trajectory sequence refers to the historical trajectory sequence whose yaw tag can be directly determined based on the road network connectivity. In this step, the server can determine the yaw tag corresponding to the first type of trajectory sequence using the methods described in steps S310A-S330A, or the methods described in steps S310B-S340B, or other methods. This implementation does not impose specific limitations on these methods.

[0131] Step S320'C: Train the image recognition model based on the target image sequence corresponding to each first type of trajectory sequence and the yaw label to obtain the trained image recognition model.

[0132] In this implementation, the server can use the target image sequence corresponding to the first type of trajectory sequence as input to the image recognition model, and use the corresponding yaw label as output to train the image recognition model, thereby obtaining the trained image recognition model. The image recognition model can be any existing image recognition model, such as CNN, RNN, etc., and the image recognition model is a classification model.

[0133] Step S330'C: Based on the target image sequence corresponding to each second-type trajectory sequence, determine the yaw label of the corresponding second-type trajectory sequence based on the trained image recognition model.

[0134] In this implementation, the second type of trajectory sequence refers to the historical trajectory sequence whose yaw label cannot be directly determined based on road network connectivity. The server can input the target image sequence corresponding to the second type of trajectory sequence into the trained image recognition model to determine the yaw label corresponding to the second trajectory sequence.

[0135] In step S400', in response to the yaw probability satisfying the first probability condition, path planning is performed based on the road segment identifier of the non-target road segment and the target location to determine the target path planning result.

[0136] In this embodiment, the first probability condition can be that the yaw probability is greater than or not lower than the third threshold. When the yaw probability meets the first probability condition, it indicates that the vehicle has a high probability of yaw behavior. Therefore, the terminal needs to update the navigation path to improve the accuracy of the path navigation and reduce the possibility of the vehicle taking a detour. Therefore, the server can call the navigation service to perform path planning based on the road segment identifier of the non-target road segment and the target location to determine the target path planning result corresponding to the target task.

[0137] Step S500': Send the target path planning results.

[0138] After determining the target path planning result for the navigation task, the server can send the target path planning result to the corresponding terminal based on the terminal identifier.

[0139] Step S600': In response to receiving the target path planning result of the target task, update the navigation path of the target task in the navigation page according to the target path planning result.

[0140] In this step, the terminal can update the navigation path of the navigation task on the navigation page so that the navigation path matches the target path planning result.

[0141] In an alternative implementation, the method of this embodiment may further include the following steps:

[0142] Step S700': Send yaw probability.

[0143] In this embodiment, the server can also send the yaw probability of the target task to the corresponding terminal based on the terminal identifier, so that the corresponding terminal can interact with the user based on the yaw probability to improve the accuracy of the navigation path.

[0144] It is easy to understand that in this embodiment, steps S500' and S700' can be executed simultaneously or sequentially, and this embodiment does not impose any limitations.

[0145] In step S800', in response to receiving the yaw probability and the yaw probability satisfying the first probability condition, a predetermined pop-up window is displayed.

[0146] To reduce the possibility of inaccurate predictions from the probabilistic yaw model, in one optional implementation, if the yaw probability meets the first probability condition, the terminal can also display a pre-defined pop-up window to indicate that the navigation path has changed.

[0147] Optionally, the terminal may display a pre-defined pop-up window at a first position on the navigation page, where the first position is any position on the navigation page that does not obstruct the location indicator. Furthermore, to avoid the pre-defined pop-up window obscuring the navigation path for an excessively long period, the terminal may also stop displaying the pre-defined pop-up window after a first time period (e.g., 10 seconds). Optionally, if the terminal is currently in a locked screen state, the terminal may also display the pre-defined pop-up window on the lock screen page.

[0148] Figure 8 This is a schematic diagram of an interface according to an embodiment of the present invention. For example... Figure 8 As shown, page 81 is the navigation page. When the yaw probability meets the first probability condition, the terminal can display a predetermined pop-up window, i.e., pop-up window 82, at the first position on page 81 (e.g., the center of page 81). Pop-up window 82 does not obscure the position indicator, i.e., indicator 83. Pop-up window 82 is used to indicate that the navigation route has changed, and the prompt text can be "Switched to the main road, you can switch to the auxiliary road by voice".

[0149] Step S900': In response to receiving the path restoration command, the navigation path is restored to the previous path planning result.

[0150] The route restoration command is triggered by the user. When the terminal receives the route restoration command, it indicates that the current road segment is the target road segment. Therefore, the terminal can restore the navigation route to the previous route planning result.

[0151] Optionally, to reduce the possibility of inconvenience for users operating the terminal while driving, the terminal can receive a first voice sequence sent by the user and perform intent recognition on the first voice sequence. If the recognition result of the first voice sequence is used to represent restoring the navigation path, such as the recognition result being "do not update navigation path" or "I am driving on the main road" (that is, the yaw probability predicted by the yaw probability prediction model is used to represent the trajectory sequence deviating from the main road of the predetermined road to the auxiliary road), the terminal can determine that a path restoration command has been received. In this embodiment, the terminal can perform intent recognition using various existing voice recognition models, and this embodiment does not impose specific limitations. Furthermore, the voice recognition process can also be performed by a server.

[0152] Optionally, the terminal may also render and display a path restoration control. In this embodiment, the path restoration control may be part of a predetermined window or an independent control; this embodiment is not limited to this. When the path restoration control is triggered, the terminal may also determine that a path restoration instruction has been received. In this embodiment, the path restoration control may be triggered in various existing ways, such as a single click, a long press, or a swipe in a predetermined direction; this embodiment is not limited to this.

[0153] Figure 9 This is another interface diagram of an embodiment of the present invention. For example... Figure 9As shown, the terminal can render and display the path restoration control, i.e., control 91, on page 81. When control 91 is triggered, or when the terminal determines that the recognition result of the received first speech sequence is used to represent the restoration of the navigation path, the terminal can restore the navigation path of the navigation task to the previous path planning result.

[0154] Step S1000': Send the path restoration command.

[0155] After receiving the path restoration command, the terminal can also send the path restoration command to the server.

[0156] Step S1100': In response to receiving the path restoration instruction, the navigation path is restored to the previous path planning result.

[0157] After receiving the path restoration command from the terminal, the server can restore the navigation path of the navigation task to the previous path planning result.

[0158] Figure 10 This is a flowchart of the interaction method on the server side according to the second embodiment of the present invention. Figure 10 As shown, the method in this embodiment includes the following steps on the server side:

[0159] Step S200': In response to receiving the trajectory sequence of the target task, determine the target road segment corresponding to the trajectory sequence based on the current path planning result corresponding to the target task.

[0160] Step S300': Based on the road segment characteristics of the target road segment and the coordinate point characteristics of each coordinate point, determine the yaw probability corresponding to the trajectory sequence based on the pre-trained yaw probability prediction model.

[0161] In step S400', in response to the yaw probability satisfying the first probability condition, path planning is performed based on the road segment identifier of the non-target road segment and the target location to determine the target path planning result.

[0162] Step S500': Send the target path planning results.

[0163] In an optional implementation, the method of this embodiment may further include the following steps on the server side:

[0164] Step S700': Send yaw probability.

[0165] Step S1100': In response to receiving the path restoration instruction, the navigation path is restored to the previous path planning result.

[0166] Figure 11 This is a flowchart of the interaction method of the second embodiment of the present invention on the terminal side. For example... Figure 11 As shown, the method in this embodiment includes the following steps on the terminal side:

[0167] Step S100': Obtain and send the trajectory sequence of the target task.

[0168] Step S600': In response to receiving the target path planning result of the target task, update the navigation path of the target task in the navigation page according to the target path planning result.

[0169] In an optional implementation, the method of this embodiment may further include the following steps on the terminal side:

[0170] In step S800', in response to receiving the yaw probability and the yaw probability satisfying the first probability condition, a predetermined pop-up window is displayed.

[0171] Step S900': In response to receiving the path restoration command, the navigation path is restored to the previous path planning result.

[0172] Step S1000': Send the path restoration command.

[0173] In this embodiment, after receiving the trajectory sequence of the target task sent by the terminal, the server determines the main or auxiliary road segment of the predetermined road corresponding to the trajectory sequence based on the current path planning result corresponding to the target task. It then determines the yaw probability corresponding to the trajectory sequence based on the road segment characteristics and the coordinate point characteristics of each coordinate point in the trajectory sequence. If the yaw probability meets a first probability condition, the server determines and sends the target path planning result to the terminal based on the road segment identifier of the non-target road segment and the target location, enabling the terminal to update the navigation path according to the target path planning result. This embodiment can accurately identify the possibility of vehicle yaw behavior and update the navigation path in a timely manner when the yaw probability is high, thereby improving navigation accuracy.

[0174] Figure 12 This is a flowchart of the interaction method according to the third embodiment of the present invention. Figure 12 As shown, this embodiment is implemented through interaction between the terminal and the server. The method of this embodiment includes the following steps:

[0175] Step S100”: Obtain and send the trajectory sequence of the target task.

[0176] In this embodiment, the implementation of step S100” is similar to that of step S100’, and will not be described again here.

[0177] Step S200”: In response to receiving the trajectory sequence of the target task, the target road segment corresponding to the trajectory sequence is determined according to the current path planning result corresponding to the target task.

[0178] In this embodiment, the implementation of step S200” is similar to that of step S200’, and will not be described again here.

[0179] Step S300”: Based on the road segment characteristics of the target road segment and the coordinate point characteristics of each coordinate point, the yaw probability corresponding to the trajectory sequence is determined based on the pre-trained yaw probability prediction model.

[0180] In this embodiment, the implementation of step S300” is similar to that of step S300’, and will not be described again here.

[0181] In step S400, in response to the yaw probability satisfying the first probability condition, path planning is performed based on the road segment identifier of the non-target road segment and the target location to determine the target path planning result.

[0182] In this embodiment, the implementation of step S400” is similar to that of step S400’, and will not be described again here.

[0183] Step S500”: Send the target path planning results.

[0184] In this embodiment, the implementation of step S500” is similar to that of step S500’, and will not be described again here.

[0185] Step S600”: In response to receiving the target path planning result of the target task, update the navigation path of the target task in the navigation page according to the target path planning result.

[0186] In this embodiment, the implementation of step S600” is similar to that of step S600’, and will not be described again here.

[0187] In an alternative implementation, the method of this embodiment may further include the following steps:

[0188] Step S700”, send yaw probability.

[0189] In this embodiment, the implementation of step S700” is similar to that of step S700, and will not be described again here. In this embodiment, steps S500” and S700” can be executed simultaneously or sequentially, and this embodiment does not impose any restrictions.

[0190] In step S800, in response to receiving the yaw probability and the yaw probability satisfying the first probability condition, a pre-defined pop-up window is displayed.

[0191] In this embodiment, the implementation of step S800” is similar to that of step S800’, and will not be described again here.

[0192] Step S900”: In response to receiving the path restoration instruction, the navigation path is restored to the previous path planning result.

[0193] In this embodiment, the implementation of step S900” is similar to that of step S900’, and will not be described again here.

[0194] Step S1000”, send the path restoration command.

[0195] In this embodiment, the implementation of step S1000” is similar to that of step S1000’, and will not be described again here.

[0196] Step S1100”: In response to receiving the path restoration instruction, the navigation path is restored to the previous path planning result.

[0197] In this embodiment, the implementation of step S1100” is similar to that of step S1100’, and will not be described again here.

[0198] In step S1200, in response to receiving the yaw probability and the yaw probability satisfying the second probability condition, the position confirmation control is rendered and displayed.

[0199] In this embodiment, the second probability condition can be a yaw probability greater than the fourth threshold and not higher than the third threshold, not lower than the fourth threshold and not higher than the third threshold, or not lower than the fourth threshold and lower than the third threshold. When the yaw probability meets the second probability condition, it means that the server cannot determine whether the vehicle has yawed, so the terminal can confirm with the user through the location confirmation control whether the current road segment is the target road segment.

[0200] Optionally, the terminal may display a location confirmation control in a second position on the navigation page, where the second position is any location on the navigation page that does not obscure the location indicator. Furthermore, to avoid the predetermined pop-up window obscuring the navigation path for too long, the terminal may also stop rendering and displaying the location confirmation control for a second time period (e.g., 10 seconds). Optionally, if the terminal is currently in a locked screen state, the terminal may also render and display the location confirmation control on the lock screen page.

[0201] In step S1300, in response to receiving the route update instruction, a route update request is generated and sent based on the road segment identifier and the target location.

[0202] If a path update command is received, it means that the current road is not the target road. Therefore, the terminal can generate and send a path update request to the server based on the road segment identifier of the non-target road segment and the target location, so as to update the navigation path of the navigation task in a timely manner.

[0203] Optionally, to reduce the inconvenience for users operating the terminal while driving, the terminal can receive a second voice sequence sent by the user and perform intent recognition on the second voice sequence. If the recognition result of the second voice sequence indicates a mismatch between the current road segment and the target road segment, such as the recognition result being "I am driving on the main road" (i.e., the target road is a secondary road segment of the predetermined road, but the current road is the main road segment of the predetermined road) or "Update navigation route," the terminal can determine that a route update instruction has been received. In this embodiment, the terminal can perform intent recognition using various existing voice recognition models, and this embodiment does not impose specific limitations. Furthermore, the voice recognition process can also be performed by a server.

[0204] Optionally, the location confirmation control may include a location update control, which is used to update the current road segment from a secondary road segment to a primary road segment of a predetermined road, or to update the current road segment from a primary road segment to a secondary road segment of a predetermined road. When the route update control is triggered, the terminal can also determine that it has received a route update instruction. In this embodiment, the route update control can be triggered in various existing ways, such as a single click, a long press, or a swipe in a predetermined direction, etc., and this embodiment is not limited to these methods.

[0205] Step S1400”: In response to receiving a route update request, route planning is performed based on the road segment identifier and target location in the route update request, and the target route planning result is determined.

[0206] After receiving a route update request from the terminal, the server can invoke navigation services to perform route planning based on the road segment identifiers of non-target road segments and the target location in the route update request, thereby determining the target route planning result.

[0207] Step S1500”: Send the target path planning results.

[0208] After determining the target path planning result for the navigation task, the server can send the target path planning result to the corresponding terminal based on the terminal identifier.

[0209] Step S1600”: Receive the target path planning result and update the navigation path of the target task in the navigation page according to the target path planning result.

[0210] After receiving the target path planning results, the terminal can update the navigation path on the navigation page.

[0211] In an optional implementation of this embodiment, when the yaw probability satisfies the second probability condition, the method of this embodiment may further include the following steps:

[0212] Step S1700”: In response to receiving a path-keeping instruction or not receiving a path-update instruction within a predetermined time period, the navigation path remains unchanged.

[0213] If a path-keeping instruction is received or no path-update instruction is received within a predetermined time period, it indicates that the current road is the target road, and therefore the terminal can maintain the navigation path of the navigation task unchanged.

[0214] Optionally, to reduce the inconvenience for users operating the terminal while driving, the terminal can receive a third voice sequence sent by the user and perform intent recognition on the third voice sequence. If the recognition result of the third voice sequence indicates a match between the current road segment and the target road segment, such as "I am driving on the main road" (i.e., both the target road and the current road are auxiliary road segments of the predetermined road) or "Do not update the navigation route," the terminal can determine that it has received a route-keeping instruction. In this embodiment, the terminal can perform intent recognition using various existing voice recognition models, and this embodiment does not impose specific limitations. Furthermore, the voice recognition process can also be performed by a server.

[0215] Optionally, the location confirmation control may further include a location hold control, which is used to keep the current road segment as a main road segment of the predetermined road, or to keep the current road segment as a secondary road segment of the predetermined road. When the route hold control is triggered, the terminal can also determine that it has received a route hold instruction. In this embodiment, the route hold control can be triggered in various existing ways, such as a single click, a long press, or a swipe in a predetermined direction, etc., and this embodiment is not limited to these methods.

[0216] Figure 13 This is another schematic diagram of an embodiment of the present invention. The explanation will take the main road segment of road S1 as an example, where the target road segment corresponding to the trajectory sequence of the navigation task is the target road segment. Figure 13 Page 131 shown is the terminal's lock screen. When the yaw probability meets the second probability condition, the terminal can render and display control 132, i.e., the location confirmation control, on page 131. Control 132 includes control 133 and control 134, where control 133 is a path update control and control 134 is a path hold control. If control 133 is triggered, the terminal can confirm that it has received a path update instruction. If control 133 is not triggered within a predetermined time period, or control 134 is triggered, the terminal can confirm that it has received a path hold instruction. It is easy to understand that if the target road segment is an auxiliary road segment of S1, then control 133 is a path hold control and control 134 is a path update control.

[0217] Figure 14 This is a flowchart of the interaction method on the server side according to the third embodiment of the present invention. Figure 14 As shown, the method in this embodiment includes the following steps on the server side:

[0218] Step S200”: In response to receiving the trajectory sequence of the target task, the target road segment corresponding to the trajectory sequence is determined according to the current path planning result corresponding to the target task.

[0219] Step S300”: Based on the road segment characteristics of the target road segment and the coordinate point characteristics of each coordinate point, the yaw probability corresponding to the trajectory sequence is determined based on the pre-trained yaw probability prediction model.

[0220] In step S400, in response to the yaw probability satisfying the first probability condition, path planning is performed based on the road segment identifier of the non-target road segment and the target location to determine the target path planning result.

[0221] Step S500”: Send the target path planning results.

[0222] Step S700”, send yaw probability.

[0223] Step S1400”: In response to receiving a route update request, route planning is performed based on the road segment identifier and target location in the route update request, and the target route planning result is determined.

[0224] Step S1500”: Send the target path planning results.

[0225] In an optional implementation, the method of this embodiment may further include the following steps on the server side:

[0226] Step S1100”: In response to receiving the path restoration command, the navigation path is restored to the previous path planning result.

[0227] Figure 15 This is a flowchart of the interaction method of the third embodiment of the present invention on the terminal side. Figure 15 As shown, the method in this embodiment includes the following steps on the terminal side:

[0228] Step S100”: Obtain and send the trajectory sequence of the target task.

[0229] Step S600”: In response to receiving the target path planning result of the target task, update the navigation path of the target task in the navigation page according to the target path planning result.

[0230] In step S1200, in response to receiving the yaw probability and the yaw probability satisfying the second probability condition, the position confirmation control is rendered and displayed.

[0231] In step S1300, in response to receiving the route update instruction, a route update request is generated and sent based on the road segment identifier and the target location.

[0232] Step S1600”: Receive the target path planning result and update the navigation path of the target task in the navigation page according to the target path planning result.

[0233] In an optional implementation, the method of this embodiment may further include the following steps on the terminal side:

[0234] In step S800, in response to receiving the yaw probability and the yaw probability satisfying the first probability condition, a pre-defined pop-up window is displayed.

[0235] Step S800”: In response to receiving the path restoration instruction, the navigation path is restored to the previous path planning result.

[0236] Step S900”, send the path restoration command.

[0237] Step S1700”: In response to receiving a path-keeping instruction or not receiving a path-update instruction within a predetermined time period, the navigation path remains unchanged.

[0238] In this embodiment, after receiving the trajectory sequence of the target task sent by the terminal, the server determines the main or auxiliary road segment of the predetermined road corresponding to the trajectory sequence based on the current path planning result corresponding to the target task, and determines the yaw probability corresponding to the trajectory sequence based on the road segment characteristics and the coordinate point characteristics of each coordinate point in the trajectory sequence. If the yaw probability meets the first probability condition, the server determines and sends the target path planning result to the terminal based on the road segment identifier and target location of the non-target road segment, so that the terminal updates the navigation path according to the target path planning result. Furthermore, if the yaw probability meets the first probability condition, the terminal can display a predetermined pop-up window to prompt the user that the navigation path has changed. If the yaw probability meets the second probability condition, the terminal renders and displays a location confirmation control to confirm whether the current road segment is the target road segment. If a path update instruction is received, the terminal generates and sends a path update request to the server based on the road segment identifier and target location of the current road segment. The server determines and sends the target path planning result to the terminal based on the road segment identifier and target location in the path update request, so that the terminal updates the navigation path according to the received target path planning result; if a path hold instruction is received, or if no path update instruction is received within a predetermined time period, the terminal keeps the navigation path unchanged. The embodiments of the present invention can accurately identify the possibility of a vehicle veerging off course, and further determine whether the vehicle has veered off course through interactive means, thereby effectively improving navigation accuracy.

[0239] Figure 16 This is a schematic diagram of the interactive system according to the fourth embodiment of the present invention. Figure 16 As shown, the interactive system in this embodiment includes interactive device 16A and interactive device 16B.

[0240] The interactive device 16A includes a sequence acquisition unit 1601 and a path update unit 1602.

[0241] The sequence acquisition unit 1601 is used to acquire the trajectory sequence of the target task, which is a sequence of multiple coordinate points. The path update unit 1602, in response to determining that the trajectory sequence has deviated, presents the associated information of the navigation path of the target task on the navigation page according to the target path planning result of the target task. The deviation behavior is determined based on the deviation probability of the target task. The deviation probability characterizes the probability that the trajectory sequence deviates from the target road segment to a non-target road segment. The target road segment is the main road segment or auxiliary road segment of a predetermined road, and the non-target road segment is the auxiliary road segment or main road segment of the predetermined road.

[0242] The interactive device 16B includes a road segment determination unit 1603, a probability prediction unit 1604, a planning result determination unit 1605, and a planning result transmission unit 1606.

[0243] The route segment determination unit 1603, in response to receiving a trajectory sequence of a target task, determines the target route segment corresponding to the trajectory sequence based on the current path planning result corresponding to the target task. The trajectory sequence is a sequence of multiple coordinate points, and the target route segment is a main route segment or an auxiliary route segment of a predetermined road. The probability prediction unit 1604, based on the route characteristics of the target route segment and the coordinate characteristics of each coordinate point, determines the yaw probability corresponding to the trajectory sequence based on a pre-trained yaw probability prediction model. The yaw probability characterizes the probability that the trajectory sequence deviates from the target route segment to a non-target route segment, where the non-target route segment is an auxiliary route segment or a main route segment of the predetermined road. The planning result determination unit 1605, in response to the yaw probability satisfying a first probability condition, performs path planning based on the route segment identifier and target position of the non-target route segment to determine the target path planning result. The target position is the termination position of the target task. The planning result sending unit 1606 is used to send the target path planning result.

[0244] In this embodiment, after obtaining the trajectory sequence of the target task, if it is determined that the trajectory sequence has deviated, the navigation path association information of the target task is presented on the navigation page based on the target path planning result of the target task. In this embodiment, the target navigation path is determined based on the deviation probability of the target task, and the deviation probability is used to characterize the probability that the trajectory sequence deviates from the main road segment of the predetermined road to the auxiliary road segment of the predetermined road, or from the auxiliary road segment of the predetermined road to the main road segment of the predetermined road. Therefore, this embodiment can accurately identify the possibility of the vehicle deviating and update the navigation path in a timely manner according to the deviation probability, thereby improving navigation accuracy.

[0245] Figure 17 This is a schematic diagram of an electronic device according to the fifth embodiment of the present invention. Figure 17 The illustrated electronic device is a general-purpose data processing device, comprising a general-purpose computer hardware architecture, including at least a processor 1701 and a memory 1702. The processor 1701 and memory 1702 are connected via a bus 1703. The memory 1702 is adapted to store instructions or programs executable by the processor 1701. The processor 1701 can be a standalone microprocessor or a collection of one or more microprocessors. Thus, the processor 1701 executes the commands stored in the memory 1702, thereby performing the method flow of the embodiments of the present invention as described above to process data and control other devices. The bus 1703 connects the aforementioned components together, and also connects these components to a display controller 1704, a display device, and an input / output (I / O) device 1705. The input / output (I / O) device 1705 can be a mouse, keyboard, modem, network interface, touch input device, motion-sensing input device, printer, and other devices known in the art. Typically, the input / output (I / O) device 1705 is connected to the system via the input / output (I / O) controller 1706.

[0246] The memory 1702 can store software components, such as an operating system, a communication module, an interaction module, and application programs. Each of the modules and application programs described above corresponds to a set of executable program instructions that perform one or more functions and the methods described in the embodiments of the invention.

[0247] The flowcharts and / or block diagrams of the methods, apparatus (systems), and computer program products according to embodiments of the present invention describe various aspects of the invention. It should be understood that each block of the flowcharts and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus to produce a machine such that the instructions (executed via the processor of the computer or other programmable data processing apparatus) create means for implementing the functions / actions specified in the flowchart and / or block diagram blocks or blocks.

[0248] Furthermore, as those skilled in the art will recognize, various aspects of the embodiments of the present invention can be implemented as a system, method, or computer program product. Therefore, various aspects of the embodiments of the present invention can take the form of a completely hardware implementation, a completely software implementation (including firmware, resident software, microcode, etc.), or an implementation combining software and hardware aspects, which may generally be referred to herein as a "circuit," "module," or "system." Additionally, aspects of the present invention can take the form of a computer program product implemented in one or more computer-readable media having computer-readable program code implemented thereon.

[0249] Any combination of one or more computer-readable media can be used. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example, (but not limited to) an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or apparatus, or any suitable combination thereof. More specific examples (not an exhaustive list) of computer-readable storage media will include: an electrical connection having one or more wires, a portable computer floppy disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable optical disc read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In the context of embodiments of the present invention, a computer-readable storage medium can be any tangible medium capable of containing or storing a program used by or in conjunction with an instruction execution system, device, or apparatus.

[0250] Computer-readable signal media may include propagated data signals having computer-readable program code implemented therein, such as in baseband or as part of a carrier wave. Such propagated signals may take any of a variety of forms, including, but not limited to, electromagnetic, optical, or any suitable combination thereof. A computer-readable signal medium may be any computer-readable medium that is not a computer-readable storage medium and can communicate, propagate, or transmit a program used by or in conjunction with an instruction execution system, device, or apparatus.

[0251] Computer program code used to perform operations relating to various aspects of this invention can be written in any combination of one or more programming languages, including: object-oriented programming languages ​​such as Java, Smalltalk, C++, PHP, Python, etc.; and conventional procedural programming languages ​​such as the "C" programming language or similar programming languages. The program code can be executed as a standalone software package entirely on the user's computer, partially on the user's computer, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the latter case, the remote computer can be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet provided by an Internet service provider).

[0252] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. For those skilled in the art, the present invention can be modified and varied in various ways. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of protection of the present invention.

Claims

1. An interaction method, characterized in that, The method includes: Obtain the trajectory sequence of the target task, wherein the trajectory sequence is a sequence of multiple coordinate points; In response to determining that the trajectory sequence has deviated, the navigation path association information of the target task is presented on the navigation page according to the target path planning result of the target task. The deviation of the trajectory sequence is determined according to the deviation probability of the target task. The deviation probability is used to characterize the probability that the trajectory sequence deviates from the target road segment to a non-target road segment. The target road segment is the main road segment or auxiliary road segment of the predetermined road, and the non-target road segment is the auxiliary road segment or main road segment of the predetermined road. The yaw probability is determined based on the road segment features of the target road segment and the coordinate point features of each coordinate point, which are input into a pre-trained yaw probability prediction model. The yaw probability prediction model is trained on a training sample set, which is determined based on multiple historical trajectory sequences, preset moving road segments corresponding to each historical trajectory sequence, and yaw tags. The preset moving road segments are auxiliary or main road segments of specific roads in the historical navigation paths corresponding to the historical trajectory sequences. The yaw tags are used to characterize whether the actual moving road segment deviates from the preset moving road segment. The actual moving road segment is an auxiliary or main road segment of the specific road. The yaw tags are determined based on at least one of the road network connectivity relationship and image recognition results corresponding to the historical trajectory sequences. The image recognition results are obtained by recognizing a target image sequence, which is an image sequence synchronously acquired along with the corresponding historical trajectory sequence.

2. The method according to claim 1, characterized in that, The method further includes: Send the trajectory sequence of the target task.

3. The method according to claim 2, characterized in that, The trajectory sequence for sending the target task includes: Obtain the current path planning result for the target task; Determine the current road segment corresponding to the trajectory sequence, and determine the target road segment corresponding to the trajectory sequence based on the current path planning result; In response to a mismatch between the current road segment and the target road segment, the trajectory sequence is sent.

4. The method according to claim 1, characterized in that, The response to determining that the trajectory sequence has veered off course includes: Obtain the yaw probability; In response to the yaw probability satisfying the first probability condition, it is determined that the trajectory sequence has yawed.

5. The method according to claim 4, characterized in that, The associated information for presenting the navigation path of the target task on the navigation page based on the target path planning result of the target task includes: Obtain the target path planning results; Update the navigation path on the navigation page based on the target path planning results.

6. The method according to claim 5, characterized in that, The method of presenting the navigation path association information of the target task on the navigation page based on the target path planning result of the target task also includes: A pre-defined pop-up window is displayed to indicate that the navigation route has changed. In response to receiving a path restoration command, the navigation path is restored to the previous path planning result.

7. The method according to claim 6, characterized in that, The received path restoration instruction specifically includes: Receive a first speech sequence and perform intent recognition on the first speech sequence; In response to the recognition result of the first speech sequence being used to characterize the reconstruction of the navigation path, it is determined that the path reconstruction instruction has been received; or In response to the path restore control being triggered, confirm that a path restore command has been received.

8. The method according to claim 6, characterized in that, The pop-up window for displaying the scheduled content includes: The predetermined pop-up window is displayed at a first position on the navigation page, where the first position is any position on the navigation page that does not obscure the position indicator.

9. The method according to claim 1, characterized in that, The response to determining that the trajectory sequence has veered off course includes: Obtain the yaw probability; In response to the yaw probability satisfying the second probability condition, a position confirmation control is rendered and displayed. The position confirmation control is used to confirm whether the current road segment is the target road segment. In response to receiving a path update command, it is determined that the trajectory sequence has deviated.

10. The method according to claim 9, characterized in that, The location confirmation control includes a location update control; The received path update instruction specifically includes: Receive a second speech sequence and perform intent recognition on the second speech sequence; In response to the recognition result of the second speech sequence indicating a mismatch between the current road segment and the target road segment, it is determined that the path update instruction has been received; or In response to the location update control being triggered, it is determined that the path update instruction has been received.

11. The method according to claim 9, characterized in that, The rendering display position confirmation control includes: The location confirmation control is rendered and displayed at a second location on the navigation page, where the second location is any location on the navigation page that does not obscure the location indicator.

12. The method according to claim 9, characterized in that, The associated information for presenting the navigation path of the target task on the navigation page based on the target path planning result of the target task includes: Obtain the target path planning results; Update the navigation path on the navigation page based on the target path planning results.

13. The method according to claim 9, characterized in that, The method further includes: In response to determining that the trajectory sequence has not deviated, the navigation path remains unchanged.

14. The method according to claim 13, characterized in that, Determining that the trajectory sequence has not deviated includes: In response to receiving a path hold instruction, or if the path update instruction is not received within a predetermined time period.

15. The method according to claim 14, characterized in that, The location confirmation control includes a location hold control; The received path-keeping instruction specifically includes: Receive a third speech sequence and perform intent recognition on the third speech sequence; In response to the recognition result of the third speech sequence used to characterize the match between the current road segment and the target road segment, it is determined that the path-keeping instruction has been received; or In response to the position holding control being triggered, it is determined that the path holding instruction has been received.

16. The method according to claim 1, characterized in that, The target path planning result is determined based on the road segment identifiers of the non-target road segments and the target location, where the target location is the termination location of the target task.

17. The method according to claim 1, characterized in that, The yaw tag is determined in the following way: For each of the historical trajectory sequences, a first-class road segment connected to the actual moving road segment is determined; The connection relationship between the preset mobile road segment and the first type of road segment is determined based on the road network connectivity relationship; In response to the connectivity relationship being disconnected, it is determined that the yaw tag is used to characterize that the historical trajectory sequence has veered off course; In response to the connectivity being connected, the yaw tag is determined to be used to characterize that the historical trajectory sequence has not deviated.

18. The method according to claim 17, characterized in that, The first type of road segment that is determined to be connected to the actual moving road segment includes: The first type of road segment is determined based on the road network connectivity; or Image recognition is performed on the target image sequence to determine the first type of road segment.

19. The method according to claim 18, characterized in that, The step of performing image recognition on the target image sequence to determine the first type of road segment includes: Image recognition is performed on the target image sequence to determine whether there is a road segment connected to the actual moving road segment; In response to the existence of a road segment connected to the actual moving road segment, the first type of road segment is determined according to the road network connectivity relationship.

20. The method according to claim 1, characterized in that, The target image sequence is an image sequence collected within a predetermined range, and the predetermined range is at least one of the range where the road segment intersection is located and the range where the road segment entrance and exit are located; The yaw tag is determined in the following way: Image recognition is performed on the target image sequence to determine the first main and auxiliary road label corresponding to the actual moving road segment; Determine the second main and auxiliary road label corresponding to the preset mobile road segment; In response to a mismatch between the first main and auxiliary road label and the second main and auxiliary road label, it is determined that the yaw label is used to characterize a yaw in the historical trajectory sequence; In response to the matching of the first main and auxiliary road label and the second main and auxiliary road label, it is determined that the yaw label is used to characterize that the historical trajectory sequence has not deviated.

21. The method according to claim 1, characterized in that, The yaw tag is determined in the following way: The yaw label of the corresponding first type trajectory sequence is determined according to the road network connectivity relationship corresponding to each first type trajectory sequence, and the first type trajectory sequence belongs to the plurality of historical trajectory sequences; The image recognition model is trained based on the target image sequence corresponding to each of the first type of trajectory sequences and the yaw label to obtain the trained image recognition model; Based on the target image sequence corresponding to each second type of trajectory sequence, and using the trained image recognition model, the yaw label of the corresponding second type of trajectory sequence is determined.

22. An interaction method, characterized in that, The method includes: In response to receiving a trajectory sequence of a target task, the target road segment corresponding to the trajectory sequence is determined based on the current path planning result corresponding to the target task. The trajectory sequence is a sequence of multiple coordinate points, and the target road segment is a main road segment or an auxiliary road segment of a predetermined road. Based on the road segment characteristics of the target road segment and the coordinate point characteristics of each coordinate point, and based on a pre-trained yaw probability prediction model, the yaw probability corresponding to the trajectory sequence is determined. The yaw probability is used to characterize the probability that the trajectory sequence deviates from the target road segment to a non-target road segment. The non-target road segment is the auxiliary road segment or the main road segment of the predetermined road. In response to the yaw probability satisfying the first probability condition, path planning is performed based on the road segment identifier of the non-target road segment and the target location to determine the target path planning result, wherein the target location is the termination location of the target task; Send the target path planning results; The yaw probability prediction model is trained based on a training sample set, which is determined according to multiple historical trajectory sequences, preset moving segments corresponding to each historical trajectory sequence, and yaw labels. The preset moving segments are the main or auxiliary segments of a specific road in the historical navigation path corresponding to the historical trajectory sequence. The yaw labels are used to characterize whether the actual moving segment deviates from the preset moving segment. The actual moving segment is the auxiliary or main segment of the specific road. The yaw labels are determined based on at least one of the road network connectivity relationship and image recognition results corresponding to the historical trajectory sequence. The image recognition results are obtained by recognizing a target image sequence, which is an image sequence acquired synchronously with the corresponding historical trajectory sequence.

23. The method according to claim 22, characterized in that, The yaw tag is determined in the following way: For each of the historical trajectory sequences, a first-class road segment connected to the actual moving road segment is determined; The connection relationship between the preset mobile road segment and the first type of road segment is determined based on the road network connectivity relationship; In response to the connectivity relationship being disconnected, it is determined that the yaw tag is used to characterize that the historical trajectory sequence has veered off course; In response to the connectivity being connected, the yaw tag is determined to be used to characterize that the historical trajectory sequence has not deviated.

24. The method according to claim 23, characterized in that, The first type of road segment that is determined to be connected to the actual moving road segment includes: The first type of road segment is determined based on the road network connectivity; or Image recognition is performed on the target image sequence to determine the first type of road segment.

25. The method according to claim 24, characterized in that, The step of performing image recognition on the target image sequence to determine the first type of road segment includes: Image recognition is performed on the target image sequence to determine whether there is a road segment connected to the actual moving road segment; In response to the existence of a road segment connected to the actual moving road segment, the first type of road segment is determined according to the road network connectivity relationship.

26. The method according to claim 22, characterized in that, The target image sequence is an image sequence collected within a predetermined range, and the predetermined range is at least one of the range where the road segment intersection is located and the range where the road segment entrance and exit are located; The yaw tag is determined in the following way: Image recognition is performed on the target image sequence to determine the first main and auxiliary road label corresponding to the actual moving road segment; Determine the second main and auxiliary road label corresponding to the preset mobile road segment; In response to a mismatch between the first main and auxiliary road label and the second main and auxiliary road label, it is determined that the yaw label is used to characterize a yaw in the historical trajectory sequence; In response to the matching of the first main and auxiliary road label and the second main and auxiliary road label, it is determined that the yaw label is used to characterize that the historical trajectory sequence has not deviated.

27. The method according to claim 22, characterized in that, The yaw tag is determined in the following way: The yaw label of the corresponding first type trajectory sequence is determined according to the road network connectivity relationship corresponding to each first type trajectory sequence, and the first type trajectory sequence belongs to the plurality of historical trajectory sequences; The image recognition model is trained based on the target image sequence corresponding to each of the first type of trajectory sequences and the yaw label to obtain the trained image recognition model; Based on the target image sequence corresponding to each second type of trajectory sequence, and based on the trained image recognition model, the yaw label of the corresponding second type of trajectory sequence is determined, wherein the second type of trajectory sequence belongs to the plurality of historical trajectory sequences.

28. The method according to claim 22, characterized in that, The method further includes: Send the yaw probability.

29. The method according to claim 22, characterized in that, The method further includes: In response to receiving a path restoration command, the target path planning result is restored to the previous path planning result.

30. The method according to claim 22, characterized in that, The method further includes: In response to receiving a route update request, route planning is performed based on the road segment identifier and the target location in the route update request, and the target route planning result is determined. Send the target path planning results.

31. An interactive device, characterized in that, The device includes: A sequence acquisition unit is used to acquire the trajectory sequence of the target task, wherein the trajectory sequence is a sequence of multiple coordinate points; A path update unit is configured to, in response to determining that the trajectory sequence has deviated, present the associated information of the navigation path of the target task in the navigation page according to the target path planning result of the target task. The deviation of the trajectory sequence is determined according to the deviation probability of the target task. The deviation probability is used to characterize the probability that the trajectory sequence deviates from the target road segment to a non-target road segment. The target road segment is the main road segment or auxiliary road segment of a predetermined road, and the non-target road segment is the auxiliary road segment or main road segment of the predetermined road. The yaw probability is determined based on the road segment characteristics of the target road segment and the coordinate point characteristics of each coordinate point, which are input into a pre-trained yaw probability prediction model. The yaw probability prediction model is trained based on a training sample set, which is determined according to multiple historical trajectory sequences, preset moving segments corresponding to each historical trajectory sequence, and yaw labels. The preset moving segments are auxiliary or main segments of specific roads in the historical navigation paths corresponding to the historical trajectory sequences. The yaw labels are used to characterize whether the actual moving segments deviate from the preset moving segments. The actual moving segments are auxiliary or main segments of the specific roads. The yaw labels are determined based on at least one of the road network connectivity relationship and image recognition results corresponding to the historical trajectory sequences. The image recognition results are obtained by recognizing a target image sequence, which is an image sequence acquired synchronously with the corresponding historical trajectory sequences.

32. An interactive device, characterized in that, The device includes: A road segment determination unit is used to respond to receiving a trajectory sequence of a target task and determine the target road segment corresponding to the trajectory sequence based on the current path planning result corresponding to the target task. The trajectory sequence is a sequence of multiple coordinate points, and the target road segment is a main road segment or an auxiliary road segment of a predetermined road. The probability prediction unit is used to determine the yaw probability corresponding to the trajectory sequence based on the road segment characteristics of the target road segment and the coordinate point characteristics of each coordinate point, and on a pre-trained yaw probability prediction model. The yaw probability is used to characterize the probability that the trajectory sequence deviates from the target road segment to a non-target road segment, where the non-target road segment is an auxiliary road segment or a main road segment of the predetermined road. The planning result determination unit is used to respond to the yaw probability satisfying the first probability condition, perform path planning based on the road segment identifier of the non-target road segment and the target location, and determine the target path planning result, wherein the target location is the termination location of the target task; The planning result sending unit is used to send the target path planning result; The yaw probability prediction model is trained based on a training sample set, which is determined according to multiple historical trajectory sequences, preset moving segments corresponding to each historical trajectory sequence, and yaw labels. The preset moving segments are auxiliary or main segments of specific roads in the historical navigation paths corresponding to the historical trajectory sequences. The yaw labels are used to characterize whether the actual moving segments deviate from the preset moving segments. The actual moving segments are auxiliary or main segments of the specific roads. The yaw labels are determined based on at least one of the road network connectivity relationship and image recognition results corresponding to the historical trajectory sequences. The image recognition results are obtained by recognizing a target image sequence, which is an image sequence acquired synchronously with the corresponding historical trajectory sequences.

33. A computer-readable storage medium storing computer program instructions thereon, characterized in that, The computer program instructions, when executed by a processor, implement the method as described in any one of claims 1-30.

34. An electronic device comprising a memory and a processor, characterized in that, The memory is used to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method as described in any one of claims 1-30.

35. A computer program product comprising a computer program / instructions, characterized in that, The computer program / instructions are executed by the processor to implement the method as described in any one of claims 1-30.

Images