CN113723656A

CN113723656A - Method and device for repairing driving track, electronic equipment and storage medium

Info

Publication number: CN113723656A
Application number: CN202011622139.1A
Authority: CN
Inventors: 任慧敏; 李瑞远; 鲍捷; 谭楚婧
Original assignee: Jingdong City Beijing Digital Technology Co Ltd
Current assignee: Jingdong City Beijing Digital Technology Co Ltd
Priority date: 2020-12-31
Filing date: 2020-12-31
Publication date: 2021-11-30

Abstract

The invention provides a method and a device for repairing a driving track, electronic equipment and a storage medium, wherein the positions of the repairing tracks are determined according to the known track positions, each repairing track position is positioned on a road network section, the repairing track positions are prevented from being positioned in an unreachable area in the driving process, the repairing track positions are more fit with the reality, and the repairing driving track containing the repairing track positions also has better authenticity and usability. In addition, the repair running track has more running position sampling points relative to the running track only containing the known track position, and the sampling rate of the track position in the running track is improved.

Description

Method and device for repairing driving track, electronic equipment and storage medium

Technical Field

The invention relates to the technical field of internet, in particular to a method and a device for repairing a driving track, electronic equipment and a storage medium.

Background

With the popularization of internet economy and the continuous development of positioning technology, people generate a great deal of trajectory data in daily life, such as: store card data, social media location sharing data, sharing travel and taxi GPS data, express deliverer PDA trajectory data, and the like. By deeply mining the track data, more convenience can be brought to the life of people, such as: forecasting the traffic flow of a region in a future period of time in the city through the taxi track; the express delivery time is predicted by mining PDA track data of express delivery distributors.

Deep mining of track data depends on richer driving tracks (for example, driving tracks with higher sampling rate of driving positions), but in reality, a large number of scenes can only acquire sparse track positions (namely, driving tracks with low sampling rate), such as store card punching data and social media position sharing data. In the prior art, when a low-sampling-rate driving track is restored to a high-sampling-rate driving track, the position of the track is known to be directly presumed, and the presumed position of the track is often a position which cannot be reached (for example, the position of the track in a flower bed, the position of the track in lake water). Therefore, the track position restored by the prior art is often not practical and has low research value.

Therefore, the recovery of the track position in the prior art is often deviated from the reality, the authenticity of the recovered driving track is poor, and the usability of the driving track is reduced.

Disclosure of Invention

The invention provides a method and a device for repairing a driving track, electronic equipment and a storage medium, which are used for solving the defects that the track position is recovered by the prior art and is often deviated from the actual state, the authenticity of the recovered driving track is poor, and the usability of the driving track is reduced, so that the repaired track position is more fit with the actual state, and the sampling rate of the track position in the driving track is improved.

The invention provides a method for repairing a driving track, which comprises the following steps:

acquiring a known track position on a target object driving track as data to be corrected;

determining the track position on the road network section according to the data to be repaired to be used as a repair track position;

and generating a driving track comprising the repair track position as a repair driving track.

The invention provides a method for repairing a driving track, which is characterized in that on the basis, the method for obtaining the known track position on the driving track of a target object as data to be repaired comprises the following steps:

acquiring the running state information and/or the environment information of the target object at the known track position;

and taking the known track position, the running state information and/or the environment information as the data to be corrected.

On the basis, the method for repairing the driving track, which is used for determining the track position on the road network section according to the data to be repaired and used as the repairing track position, comprises the following steps:

inputting the data to be repaired into a repair model to obtain a repair result output by the repair model; wherein the repair result at least comprises any one of the following: the road network road section where the track position is located and the position of the track position in the road section;

taking the track position determined according to the repair result as the repair track position;

the track repairing model is used for outputting a repairing result used for determining the position of a repairing track based on input data to be repaired.

The invention provides a method for repairing a driving track, which is characterized in that on the basis, the data to be repaired are input into a repairing model to obtain a repairing result output by the repairing model, and the method comprises the following steps:

inputting the data to be repaired into a coding module in the repair model to obtain characteristic data output by the coding module; the coding module comprises a bidirectional long-time and short-time memory network;

inputting the characteristic data into a decoding module in the repair model to obtain a repair result output by the decoding module; the decoding module comprises a unidirectional long-time and short-time memory network.

The invention provides a method for repairing a driving track, which is characterized in that on the basis, the data to be repaired are input into a coding module in a repairing model to obtain characteristic data output by the coding module, and the method comprises the following steps:

inputting the known track position in the data to be repaired and the driving state information at the known track position into a first submodule in the coding module to obtain first characteristic data output by the first submodule;

inputting the environmental information in the data to be repaired into a second submodule in the coding module to obtain second characteristic data output by the second submodule;

and splicing the first characteristic data and the second characteristic data to form data serving as the characteristic data.

On the basis, before the data to be repaired is input into the repair model, the method for repairing the driving track further comprises the following steps:

matching a first known running track with the road network section, acquiring a track position on the road network section in the first known running track, taking the track position as a matched track position, and generating a sample repairing result representing the matched track position;

acquiring data comprising a track position on a second known running track as sample data to be repaired; wherein the second known travel track is obtained by reducing a sampling rate of track positions in the first known travel track;

generating training samples, and performing machine learning on a plurality of training samples to obtain the repairing model; the training sample takes the data to be repaired of the sample as input, and takes the sample repairing result as expected output.

According to the present invention, in addition to the above, before matching the first known travel track with the road network segment, the method for repairing a travel track further includes:

and determining an abnormal track position in the first known running track according to the interval distance between any two track positions in the first known running track, the interval sampling time and a speed threshold value so as to remove the abnormal track position in the first known running track.

The present invention also provides a travel track restoration device, including:

the system comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring a known track position on a target object driving track as data to be corrected;

the determining unit is used for determining the track position on the road network section according to the data to be repaired to be used as the repair track position;

a generating unit configured to generate a travel locus including the repair locus position as a repair travel locus.

The invention also provides an electronic device, which comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein the processor executes the program to realize the steps of any one of the driving track repairing methods.

The present invention also provides a non-transitory computer-readable storage medium having stored thereon a computer program which, when being executed by a processor, carries out the steps of the method for repairing a driving trajectory as described in any one of the above.

According to the driving track repairing method, the driving track repairing device, the electronic equipment and the storage medium, the repairing track positions are determined according to the known track positions, each repairing track position is located on the road network section, the repairing track positions are prevented from being located in the region which cannot be reached in the driving process, the repairing track positions are enabled to be more fit with the reality, and the repairing driving track containing the repairing track positions also has better authenticity and usability. In addition, the repair running track has more running position sampling points relative to the running track only containing the known track position, and the sampling rate of the track position in the running track is improved.

Drawings

In order to more clearly illustrate the technical solutions of the present invention or the prior art, the drawings needed for the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.

FIG. 1 is a schematic flow chart of a method for repairing a driving track according to the present invention;

FIG. 2 is a schematic illustration of a repair model provided by the present invention;

FIG. 3 is a block diagram of a driving trajectory repairing apparatus according to the present invention;

FIG. 4 is a schematic physical structure diagram of an electronic device provided by the present invention;

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

Fig. 1 is a schematic flow chart of a method for repairing a driving trajectory according to this embodiment, where the method for repairing a driving trajectory may be executed by any device capable of acquiring a known trajectory position, for example, by a server or a terminal. Referring to fig. 1, the method for repairing a driving trajectory includes:

step 101: and acquiring a known track position on the running track of the target object as data to be corrected.

The target object may be any person who travels, such as a courier, a person who goes out for shopping, or a person who travels, or the like, or the target object may be an unmanned vehicle. The known track positions are typically a few sparse track positions in the target object travel track, for example, the known track positions are data opened after the target object arrives at some positions (e.g., tourist attractions). The known trajectory position may be regarded as a trajectory position obtained by sampling the trajectory position of the travel trajectory of the target object at a low sampling rate. The data to be repaired includes a known track position, and may also include data that is beneficial to repairing other track positions besides the known track position, which is not specifically limited in this embodiment.

Step 102: and determining the track position on the road network section according to the data to be repaired to be used as a repair track position.

A road network segment, i.e. a known road network segment, is typically a segment of the road network in which a known trajectory position is located.

In a conventional repair process, a repair location repaired from a known trajectory location is likely not located in a road segment, but in a location that is not likely to be reached by some other target object, for example, in a lake or in a flower bed. In order to avoid such deviation in the conventional repairing process, in this embodiment, each repaired track position is located on a road network road segment, that is, each repaired driving track can be guaranteed to be a position that a target object can reach, so that the repaired track position is more practical, and the authenticity of the repaired driving track is further improved.

Step 103: and generating a driving track comprising the repair track position as a repair driving track.

The known track position in the repair travel track has a known track position in addition to the repair track position, and therefore, the sampling rate of the track position in the repair travel track is higher than that of the travel track including only the known track position. Therefore, the track position sampling rate of the driving track is improved by repairing the track position.

The embodiment provides a method for repairing a driving track, wherein the positions of the repaired tracks are determined according to the positions of the known tracks, each position of the repaired track is located on a road network section, and the positions of the repaired tracks are prevented from being located in an unreachable area in the driving process, so that the positions of the repaired tracks are more fit with the reality, and the repaired driving track containing the positions of the repaired tracks also has better authenticity and usability. In addition, the repair running track has more running position sampling points relative to the running track only containing the known track position, and the sampling rate of the track position in the running track is improved.

Further, on the basis of the foregoing embodiment, the acquiring a known track position on a target object driving track as data to be repaired includes:

Further, the environment information may be represented by One-Hot encoding (One-Hot encoding). Where one-hot encoding uses an N-bit status register to encode N states, each state having its own independent register bit and only one of which is active at any one time. Different environment information can be represented through one-hot coding, and size comparison relation does not exist in each environment information.

In order to further improve the accuracy of the determined position of the repair track, some data influencing the travel track, for example, travel state information of a known track position (specifically, the current speed, acceleration, traveling direction, and the like at the known track position) may be added to the data to be repaired. The environmental data may include, for example, weather, time of day, week (e.g., weekdays and non-weekdays may have an impact on travel trajectories), points of interest (i.e., POIs, e.g., in areas such as business, suburban, industrial, etc.).

In the embodiment, the accuracy of determining the position of the repair track according to the data to be repaired is improved by adding the state information and/or the environment information into the data to be repaired.

Further, on the basis of the foregoing embodiments, the determining, according to the data to be repaired, a track position on a road network segment as a repair track position includes:

The repair result output by the repair model is used to indicate the specific position of the repair track position in the road network segment, for example, the repair result may include at least any one of the following: the road network segment where the track location is located (i.e., road network segment ID), and the location of the track location in the segment (e.g., the distance between the location of the track location in the road network segment and the segment start or end of a certain ID may be expressed as a percentage of the total length of the segment). In order to be able to more accurately locate the repair track position, the repair model may be caused to output the repair result in a plurality of tasks, for example, the plurality of tasks output the link ID and output the percentage, respectively, to more accurately locate the repair track position by the link ID and the percentage.

In this embodiment, the repairing model obtained through machine learning is used for repairing the track position, and the repairing model is obtained through training of a large amount of real data, so that the track position close to the reality can be accurately repaired. In addition, the repair model outputs the repair result in a multi-task mode, the repair track position can be more accurately positioned by combining the repair result of multi-task data, and the authenticity of the track position is improved.

Further, on the basis of the foregoing embodiments, the inputting the data to be repaired into a repair model to obtain a repair result output by the repair model includes:

In this embodiment, the repair model adopts a Seq2Seq network architecture, such as an encoding module and a decoding module, and since the known track position is usually data of sequence driving, the Seq2Seq network architecture has a better processing result.

The coding module comprises a bidirectional long-short time memory network (Bi-LSTM), and can perform more accurate and fine feature extraction on the serialized data of the known track position, so that the repair track position on the driving track can be more accurately presumed through the unidirectional long-short time memory network (LSTM) in the decoding module.

In this embodiment, the network architecture of Seq2Seq is used as a repair model, so that the characteristics of the serialized data, that is, the known track position, can be sufficiently mined, and the repair track position can be more accurately inferred.

Further, on the basis of the foregoing embodiments, the inputting the data to be repaired into the coding module in the repair model to obtain the feature data output by the coding module includes:

The first submodule comprises a bidirectional long-time and short-time memory network, and the second submodule comprises a depth perception machine.

In order to enable the coding module to fully extract the features of the data to be repaired, the data to be repaired in different forms can be input into different structures for feature extraction. For example, the first submodule may input serialized data of a known trajectory position and travel state information at the known trajectory position, and perform feature extraction through the bidirectional long-and-short-term memory network. Since the environmental data is relatively stable during normal driving, for example, weather conditions, whether the driving day is a working day, and the like, the environmental information is usually nonserialized information, and therefore feature extraction can be performed on the environmental information through the depth perception engine in the second sub-module.

Fig. 2 is a schematic diagram of a principle of the repair model provided in this embodiment, referring to fig. 2, feature extraction is performed on environmental information through a depth sensing machine to obtain second feature data, feature extraction is performed on a known track position and driving state information that changes along with the known track position through a bidirectional long-and-short-term memory network to obtain first feature data, and data obtained by splicing the first feature data and the second feature data is input to a decoder, so that the repair track positions are sequentially output through the decoder. The LSTM in the decoder outputs the position of the repair track in a circulating output mode, and the position of the repair track output last time is used as a parameter of the position of the repair track output next time. Understandably, the sampling rate can be set, so that the LSTM sequentially outputs the repair track positions according to the sampling rate until the end point of the travel track is reached, and thus the travel track with any sampling rate can be obtained.

In the embodiment, the data to be repaired is subjected to feature extraction according to different characteristics of the data to be repaired, so that a more accurate feature extraction process of the data to be repaired can be realized, and the authenticity and the accuracy of a track position of subsequent repair can be improved.

In particular, the repair model comprises an encoding module and a decoding module, wherein the encoding module comprises a first submodule for feature extraction of serialized data (e.g., serialized data composed of known trajectory positions) and a second submodule for feature extraction of environmental information. Firstly, inputting a track with a low sampling rate into a first sub-module in a Seq2Seq model (namely a repair model) to obtain an embedded layer (containing first characteristic data); meanwhile, the external information (namely the environment information) passes through the second submodule to obtain another embedded layer (containing second characteristic data); and then the two embedded layers are merged and input into a decoder module in a Seq2Seq model, the decoder module is embedded with a multitask model, namely, a decoder generates road section IDs with high frequency sampling rate and the percentages of corresponding road sections one by one, and finally, the two outputs are combined and converted into GPS points, so that the high sampling rate track of the matched road network is obtained finally.

With regard to the first sub-module in the encoding module, it is intended to learn the feature data of the low sample rate trajectory. The first sub-module in the coding module is used for sequence information, and coding sequence information with any length into a vector, namely an embedded layer. When extracting sequence information, the coding module usually uses RNN and LSTM structures, but the RNN and LSTM structures can predict the output of the next time only according to the time sequence information of the previous time, however, in some sequence problems, the output of the current time is not only related to the previous state, but also related to the future state, and Bi-LSTM (bidirectional LSTM) can well solve the problem. In the problem to be solved, the known track position is strong sequence information, and the global information of the low-sampling-rate track needs to be better known in order to better realize track interpolation, so that the Bi-LSTM is used for extracting the embedded layer of the low-sampling-rate track in the invention. For better learning track characteristics, each Bi-LSTM node input not only contains the longitude and latitude of the track, but also adds the behavior characteristics of the track, such as the current speed, acceleration, traveling direction and the like of each node.

Regarding the second sub-module in the encoding module, it is to extract more external information to improve the model accuracy. The external information has great influence on the path selection and the speed of the vehicle operation, and the path selection and the speed determine the accuracy of track recovery, so that the external information feature extraction module is added in the method and the device to improve the accuracy of track interpolation. Since the extrinsic features are classification values, one-hot encoding is first employed, and then the embedded layer information of each feature is extracted through MLP. The embedded layer information is combined with the embedded layer information obtained by the encoder module as input to the decoder module.

And a decoding module for generating a sequence of an arbitrary length from the input embedded layer, based on the sequence information. Only historical information is of interest when generating the high-sampling track sequence, so the LSTM structure is selected as the decoding module in the present invention. The coding module does not directly generate each GPS point corresponding to the high sampling rate track, but distributes two generating tasks for the decoding module, wherein one task is to generate the road section ID corresponding to each point in the sequence, the other task is to output the percentage of the corresponding road section, and finally, the multi-task training model is used to convert the road section ID and the road section percentage into the specific GPS point, so that each recovered GPS point can be mapped to the road network, and the accuracy of model recovery is improved.

Further, on the basis of the foregoing embodiments, before inputting the data to be repaired into the repair model, the method further includes:

For the training process of the repairing model, in order to ensure the correctness of the first known driving track, the first known driving track may be matched with the road network section before training, and then the matched track position is selected from the first known driving track to generate a sample repairing result, so as to determine the expected output in the training sample.

Specifically, the first known travel track actually refers to an "original high-sampling-rate track without map matching", the matching track position actually refers to a "track position on the high-sampling-rate track after the first known travel track is matched through a road network segment", and the second known travel track actually refers to a "low-sampling-rate track obtained by down-sampling data in the first known travel track". The process of training the restoration model is actually a process of supervised learning, and the data to be restored of the samples determined according to the low-sampling-rate driving track of the second known driving track is used as input, and the restoration result of the samples determined according to the position of the matching track is used as expected data to obtain the restoration model through modeling. The repairing model after training can repair the low sampling rate track, and the low sampling rate track is restored to the track which has high sampling rate and is subjected to map matching.

In the embodiment, the first known running track is matched with the road network road section, and then the repairing result is determined according to the position of the matched track, so that the authenticity of the training sample is ensured, and the accuracy and the authenticity of the position of the repairing track output by the repairing model are further improved.

Further, in addition to the above embodiments, before matching the first known travel track with the road network segment, the method further includes:

The speed threshold may be dynamically set, and is determined according to the average speed of the current track, for example, the speed threshold is 2 times the average speed of the current track. This is because, in an actual scene, the vehicle is driven with a very small probability, and the vehicle is accelerated to twice the average speed of the section in a short time and then rapidly decelerated to the original speed.

Further, after determining the abnormal track position in the first known running track, two track positions adjacent to the abnormal track position are determined, and the abnormal track position is replaced by the middle position (namely the position determined by the coordinate mean value) of the two adjacent track positions.

In this embodiment, before the model training, the abnormal track position of the first known running track is removed, which is beneficial to improving the accuracy of the trained restoration model.

Specifically, the travel trajectory used to determine the training samples may be processed by a process:

a) track denoising

Based on the trajectory data generated by the GPS device, there are a small number of noisy points, i.e., there may be some points in a continuous trajectory whose recorded values deviate from the true values by more than a few hundred meters. In order to remove noise points, a speed threshold V is set in the present embodiment, and in the original trajectory data, if the distance between two consecutive points divided by the interval time is greater than the speed threshold V, the average of two adjacent points of the abnormal point is used instead. Where the setting of V is dynamic, determined by the average speed of the current trajectory, V is set to twice the average speed of the trajectory.

b) Trajectory down-sampling

In order to verify the effectiveness of the invention, after the high sampling rate track is obtained, the low sampling frequency is set, the low sampling rate track is randomly extracted from the high sampling rate track data before map matching and used as the model input, the high sampling frequency track mapped to the road network is generated through model learning, and the comparison verification is carried out with the reference true value in d).

c) Efficient feature extraction

The driving route and speed of the vehicle are greatly influenced by the external factors such as weather, time, week (weekday and non-weekday), and point of interest (POI, such as commercial area, suburban area, industrial area, etc.) in addition to individual driving habits. In order to further improve the accuracy of the model, relevant features are extracted to be used as a part of the input of the model. Since this part of the information is not a continuous value, but a classification value, it is one-hot encoded.

d) Trajectory map matching

Map matching is a process of associating trajectory data to a map road network, and if map matching is not performed, a vehicle running trajectory is likely not to be displayed on the road network (such as a building, a pond and the like), which is not beneficial to subsequent data mining and analysis, so that a low-sampling-rate trajectory is matched with the road network while being restored to a high-sampling-rate trajectory. The hidden Markov map matching algorithm has good performance on the track with high sampling rate, so the method is directly used for matching the track with high sampling rate to a road network to obtain a reference true value.

The invention trains a multi-task model by using a deep learning Seq2Seq structure framework, realizes the low sampling rate track recovery based on the road network, effectively solves the problems of the prior art that the recovery information is not fine enough and is separated from the road network, and ensures that the track recovery has usability in reality.

Fig. 3 is a block diagram of a driving trajectory repairing apparatus according to the present embodiment, and referring to fig. 3, the apparatus includes an obtaining unit 301, a determining unit 302, and a generating unit 303, wherein,

an obtaining unit 301, configured to obtain a known track position on a target object driving track as data to be repaired;

a determining unit 302, configured to determine, according to the data to be repaired, a track position on a road network section as a repair track position;

a generating unit 303, configured to generate a travel track including the repair track position as a repair travel track.

The device for repairing a driving track provided in this embodiment is suitable for the method for repairing a driving track provided in each of the embodiments, and is not described herein again.

The embodiment provides a travel track repairing device, and a repairing track position is determined according to a known track position, each repairing track position is located on a road network road section, so that the repairing track position is prevented from being located in an unreachable area in a traveling process, the repairing track position is more fit with the reality, and the repairing travel track including the repairing track position also has better authenticity and usability. In addition, the repair running track has more running position sampling points relative to the running track only containing the known track position, and the sampling rate of the track position in the running track is improved.

According to the present invention, on the basis of the above, the acquiring a known track position on a target object travel track as data to be repaired includes:

The invention provides a driving track repairing device, which determines a track position on a road network section according to the data to be repaired as a repairing track position on the basis of the driving track repairing device, and comprises the following components:

According to the present invention, on the basis of the above, the apparatus for repairing a driving track, which inputs the data to be repaired into a repair model to obtain a repair result output by the repair model, includes:

The invention provides a driving track repairing device, which is characterized in that on the basis, the data to be repaired are input into a coding module in a repairing model to obtain characteristic data output by the coding module, and the device comprises:

According to the present invention, there is provided a travel track restoration device, wherein before the data to be restored is input to a restoration model, the device further includes:

According to the present invention, in addition to the above, before matching a first known travel track with the road network link, the travel track repair device further includes:

Fig. 4 illustrates a physical structure diagram of an electronic device, which may include, as shown in fig. 4: a processor (processor)410, a communication Interface 420, a memory (memory)430 and a communication bus 440, wherein the processor 410, the communication Interface 420 and the memory 430 are communicated with each other via the communication bus 440. Processor 410 may invoke logic instructions in memory 430 to perform a method of travel track restoration, the method comprising:

In addition, the logic instructions in the memory 430 may be implemented in the form of software functional units and stored in a computer readable storage medium when the software functional units are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.

In another aspect, the present invention also provides a computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions which, when executed by a computer, enable the computer to perform a method of travel trajectory restoration, the method comprising:

In yet another aspect, the present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, is implemented to perform

The method for repairing the driving track comprises the following steps:

The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.

Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.

Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims

1. A travel track restoration method, characterized by comprising:

2. The method for repairing a driving track according to claim 1, wherein the acquiring a known track position on a driving track of a target object as data to be repaired includes:

3. The method for repairing a driving track according to claim 1 or 2, wherein the determining a track position on a road network section according to the data to be repaired as a repaired track position comprises:

4. The method for repairing a driving track according to claim 3, wherein the inputting the data to be repaired into a repair model to obtain the repair result output by the repair model comprises:

5. The method for repairing a driving track according to claim 4, wherein the inputting the data to be repaired into an encoding module in the repair model to obtain the feature data output by the encoding module comprises:

6. The method according to claim 3, before inputting the data to be repaired into a repair model, further comprising:

7. The method for repairing a travel track according to claim 6, further comprising, before matching a first known travel track with the road network segment:

8. A travel track restoration device characterized by comprising:

9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method of repairing a driving trajectory according to any one of claims 1 to 7 when executing the program.

10. A non-transitory readable storage medium on which a computer program is stored, the computer program, when being executed by a processor, implementing the steps of the travel track restoration method according to any one of claims 1 to 7.