CN114647799A

CN114647799A - Path recommendation method, and path recommendation model training method and device

Info

Publication number: CN114647799A
Application number: CN202210325797.7A
Authority: CN
Inventors: 陈晓龙; 肖飞
Original assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Current assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Priority date: 2022-03-30
Filing date: 2022-03-30
Publication date: 2022-06-21

Abstract

The disclosure provides a path recommendation method, a path recommendation model training method and a device, a data processing technology and a map technology, in particular to automatic driving, intelligent transportation and deep learning, and the method and the device can be applied to scenes such as vehicle navigation and path planning. The specific implementation scheme is as follows: the method comprises the steps of obtaining a vehicle running path recommendation request, determining a recommendation domain sequence according to the vehicle running path recommendation request, wherein the recommendation domain sequence is a recommendation running transfer relation of a vehicle between domains, the domains comprise named domains determined on the basis of named road sections with the same road name in a road network and virtual domains determined on the basis of unknown road sections in the road network, generating and outputting the vehicle running path according to the recommendation domain sequence, namely determining the vehicle running path by taking the domain as granularity, and compared with training by taking the road sections as the granularity, the technical effect of improving the recommendation efficiency is realized, and the solution space of path recommendation is relatively small.

Description

Path recommendation method, and path recommendation model training method and device

Technical Field

The present disclosure relates to a data processing technology and a map technology, and in particular, to an automatic driving, an intelligent transportation, and a deep learning, which can be applied to a vehicle navigation and a route planning, and in particular, to a route recommendation method, a method and an apparatus for training a route recommendation model.

Background

With the intelligentization of vehicle travel, the vehicle can travel from the origin to the destination in a navigation mode.

In the related art, a historical vehicle travel path may be acquired to be a recommended vehicle travel path based on the acquired vehicle travel path.

Disclosure of Invention

The disclosure provides a path recommendation method for improving recommendation efficiency, and a training method and device of a path recommendation model.

According to a first aspect of the present disclosure, there is provided a path recommendation method, including:

acquiring a vehicle running path recommendation request;

determining a recommended domain sequence according to the vehicle driving path recommendation request, wherein the recommended domain sequence is a recommended driving transfer relationship of vehicles among domains, and the domains comprise named domains determined based on named road sections with the same road name in a road network and obtained virtual domains determined based on unknown road sections in the road network;

and generating and outputting a vehicle driving path according to the recommended domain sequence.

According to a second aspect of the present disclosure, there is provided a training method of a path recommendation model, including:

carrying out aggregation processing on named road sections with the same road name in a road network to obtain a named domain; carrying out aggregation processing on the unknown road sections in the road network to obtain a virtual domain;

determining a driving transfer relation of vehicles among domains according to the road network and the acquired sample vehicle driving path which drives the road network, wherein the domains comprise the named domain and the virtual domain;

and training according to the driving transfer relation and the sample vehicle driving path to obtain a path recommendation model, wherein the path recommendation model is used for recommending the vehicle driving path.

According to a third aspect of the present disclosure, there is provided a path recommendation device including:

the vehicle driving path recommendation system comprises a first acquisition unit, a second acquisition unit and a recommendation unit, wherein the first acquisition unit is used for acquiring a vehicle driving path recommendation request;

the vehicle driving path recommendation system comprises a first determination unit, a second determination unit and a third determination unit, wherein the first determination unit is used for determining a recommendation domain sequence according to the vehicle driving path recommendation request, the recommendation domain sequence is a recommendation driving transfer relationship of vehicles among domains, and the domains comprise named domains determined based on named road sections with the same road name in a road network and obtained virtual domains determined based on unknown road sections in the road network;

the generating unit is used for generating a vehicle driving path according to the recommended domain sequence;

an output unit for outputting the vehicle travel path.

According to a fourth aspect of the present disclosure, there is provided a training apparatus for a path recommendation model, including:

the aggregation unit is used for performing aggregation processing on the famous road sections with the same road name in the road network to obtain a famous domain; carrying out aggregation processing on the unknown road sections in the road network to obtain a virtual domain;

a second determining unit, configured to determine a driving transfer relationship between domains of a vehicle according to the road network and an obtained sample vehicle driving path that runs through the road network, where the domains include the named domain and the virtual domain;

and the training unit is used for obtaining a path recommendation model according to the driving transfer relation and the sample vehicle driving path training, wherein the path recommendation model is used for recommending the vehicle driving path.

According to a fifth aspect of the present disclosure, there is provided an electronic device comprising:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,

the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of the first aspect or the second aspect.

According to a sixth aspect of the present disclosure, there is provided a non-transitory computer readable storage medium having stored thereon computer instructions for causing the computer to perform the method according to the first or second aspect.

According to a seventh aspect of the present disclosure, there is provided a computer program product comprising: a computer program, stored in a readable storage medium, from which at least one processor of an electronic device can read the computer program, execution of the computer program by the at least one processor causing the electronic device to perform the method of the first aspect or the second aspect.

According to the recommendation domain sequence which is used for determining the representation of the recommended driving transfer relation based on the named domain and the virtual domain, the vehicle driving path is determined based on the recommendation domain sequence, namely the technical characteristics of the vehicle driving path are determined by taking the domain as granularity, compared with training by taking the road section as the granularity, the 'solution space' of path recommendation is relatively small, and the technical effect of improving the recommendation efficiency can be achieved.

It should be understood that the statements in this section do not necessarily identify key or critical features of the embodiments of the present disclosure, nor do they limit the scope of the present disclosure. Other features of the present disclosure will become apparent from the following description.

Drawings

The drawings are included to provide a better understanding of the present solution and are not to be construed as limiting the present disclosure. Wherein:

FIG. 1 is a schematic diagram according to a first embodiment of the present disclosure;

FIG. 2 is a schematic diagram according to a second embodiment of the present disclosure;

FIG. 3 is a schematic diagram according to a third embodiment of the present disclosure;

FIG. 4 is a schematic diagram according to a fourth embodiment of the present disclosure;

FIG. 5 is a schematic diagram according to a fifth embodiment of the present disclosure;

FIG. 6 is a schematic diagram according to a sixth embodiment of the present disclosure;

FIG. 7 is a schematic diagram according to a seventh embodiment of the present disclosure;

FIG. 8 is a schematic diagram according to an eighth embodiment of the present disclosure;

FIG. 9 is a schematic illustration according to a ninth embodiment of the present disclosure;

fig. 10 is a block diagram of an electronic device for implementing a path recommendation method and a training method of a path recommendation model according to an embodiment of the present disclosure.

Detailed Description

Exemplary embodiments of the present disclosure are described below with reference to the accompanying drawings, in which various details of the embodiments of the disclosure are included to assist understanding, and which are to be considered as merely exemplary. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the present disclosure. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.

With the intellectualization of vehicle traveling, vehicle users have higher demands on the safety, efficiency and the like of the vehicle traveling. In order to meet the vehicle traveling requirements of vehicle users and improve the safety and efficiency of vehicle traveling, at least the following embodiments may be adopted to implement recommendation of a vehicle traveling path.

In some embodiments, the vehicle travel path with the smallest total cost may be searched and recommended based on a graph theory algorithm.

Illustratively, the traffic cost of each road of the road network is mined based on the historical driving track or the road attribute, the path with the minimum traffic cost is quickly searched through a graph theory algorithm, and the path with the minimum traffic cost is determined as the recommended driving path of the vehicle.

The road attribute includes the type of the road, such as an expressway, and the like, and may also include the traffic congestion degree of the road, and the like. The passing cost can be set by the user based on the requirement, for example, the minimum passing cost can be the shortest passing time, can also be the most smooth, and the passing cost is the minimum.

However, the traffic cost of the vehicle driving path is not completely equal to the sum of the road and intersection costs of the vehicle driving path, and the road sequence relationship in the vehicle driving path is also a cost influence factor, for example, the traffic cost of merging from a side road into a main road and turning left is different from the traffic cost of turning left on the main road, so that the method has the disadvantage that the accuracy of the recommended vehicle driving path is low.

In other embodiments, the path recommendation model may be trained to generate and recommend a vehicle travel path based on the path recommendation model.

For example, sample data is acquired, wherein the sample data is a historical vehicle traveling path, the basic network model is trained according to the sample data to obtain a path recommendation model, and a vehicle traveling path is recommended according to the path recommendation model, so that a vehicle travels based on the vehicle traveling path, or a user controls the vehicle to travel based on the vehicle traveling path.

However, the road network has a wide coverage area, including millions or more road segments, and there are disadvantages of low efficiency and large resource consumption in both training the path recommendation model and predicting the vehicle driving path based on the path recommendation model.

In order to avoid the above technical problems, the inventors of the present disclosure have made creative efforts to obtain the inventive concept of the present disclosure: and determining a recommended domain sequence according to the vehicle driving path recommendation request, namely determining a recommended driving transfer relationship of the vehicle between domains to generate a vehicle driving path based on the transfer relationship.

Based on the inventive concept, the invention provides a path recommendation method, a path recommendation model training method and a path recommendation model training device, relates to a data processing technology and a map technology, and particularly relates to automatic driving, intelligent transportation and deep learning.

Fig. 1 is a schematic diagram according to a first embodiment of the present disclosure, and as shown in fig. 1, a path recommendation method according to an embodiment of the present disclosure includes:

s101: and acquiring a vehicle running path recommendation request.

For example, the execution subject of this embodiment may be a path recommendation device (hereinafter, referred to as a recommendation device for short), and the recommendation device may be a server (such as a cloud server, a local server, or a server cluster), a computer, a terminal device, a processor, a chip, or the like, which is not listed here.

The method for obtaining the vehicle travel path recommendation request in this embodiment is not limited, and may be obtained based on a setting and a vehicle-mounted terminal of the vehicle, or obtained based on a user device of a vehicle user, or obtained based on a sound pickup device provided in the vehicle, and so on, which are not listed here.

S102: and determining a recommended domain sequence according to the vehicle running path recommendation request. The recommended domain sequence is a recommended driving transfer relation of vehicles among domains, and the domains comprise named domains determined based on named road sections with the same road name in a road network and acquired virtual domains determined based on unknown road sections in the road network.

A road network is understood to be a road system consisting of various road segments interconnected and interwoven in a mesh. For example, the road network includes road segments, the road segments may or may not have road names, and different road segments may have the same road name.

A named domain can be understood as a region consisting of named road segments having the same road name. A virtual domain is understood to be a region consisting of unknown road segments.

The recommended travel shift relationship is a relative concept with the sample travel shift relationship in the following, and "recommended" and "sample" cannot be understood as a definition of the travel shift relationship.

For example, if a vehicle travels from one domain to another domain, the travel relationship of the vehicle traveling from one domain to another domain may refer to the travel transfer relationship between one domain and another domain. If the vehicle can travel from domain a to domain B, it can be said that there is a travel transition relationship between domain a and domain B.

And because the domain comprises the famous domain and the virtual domain, correspondingly, the driving transfer relation can be the driving transfer relation between the famous domain and the famous domain, the driving transfer relation between the famous domain and the virtual domain, and the driving transfer relation between the virtual domain and the virtual domain.

S103: and generating and outputting a vehicle driving path according to the recommended domain sequence.

Based on the above analysis, an embodiment of the present disclosure provides a path recommendation method, including: the method comprises the steps of obtaining a vehicle driving path recommendation request, determining a recommendation domain sequence according to the vehicle driving path recommendation request, wherein the recommendation domain sequence is a recommendation driving transfer relation of a vehicle between domains, the domains comprise named domains determined based on named road sections with the same road name in a road network and virtual domains determined based on unknown road sections in the road network, and generating and outputting a vehicle driving path according to the recommendation domain sequence.

For the reader to more deeply understand the implementation principle of the present disclosure, the embodiment shown in fig. 1 will now be explained in more detail with reference to fig. 2.

Fig. 2 is a schematic diagram according to a second embodiment of the present disclosure, and as shown in fig. 2, a path recommendation method according to an embodiment of the present disclosure includes:

s201: and acquiring a vehicle driving path recommendation request, wherein the driving path recommendation request comprises a driving starting point and a driving end point.

It should be understood that, in order to avoid cumbersome statements, the present embodiment will not be described again with respect to the same technical features as those in the above embodiments.

The travel starting point is a start point of travel of the vehicle, and the travel end point is a destination of travel of the vehicle.

S202: and determining a recommended field sequence according to the driving starting point and the driving end point. The recommended domain sequence is a recommended driving transfer relation of vehicles among domains, and the domains comprise named domains determined based on named road sections with the same road name in a road network and acquired virtual domains determined based on unknown road sections in the road network.

For example, a domain (including a nameless domain and/or a nameless domain) between the driving start point and the driving end point may be determined, and a driving transition relationship between the domains, which may be referred to as a recommended driving transition relationship, may be determined for the vehicle to travel from the driving start point to the driving end point.

In some embodiments, the named fields are obtained by performing aggregation processing on named road segments with the same road name in a road network; the virtual domain is obtained by carrying out aggregation processing on the unknown road sections in the road network.

For example, the recommendation device may determine each of the famous road segments and each of the unknown road segments from the road network. The recommendation device may acquire, for each of the named links, a road name corresponding to each of the named links, to extract the named links having the same road name from the named links, and aggregate the extracted named links to obtain a named domain, which may be understood as a region composed of the named links having the same road name.

For example, if the road names of the named links a1, a2, and A3 are all a, the named links a1, a2, and A3 may be aggregated to obtain named domains. Accordingly, the named zones can be understood as connected zones based on the named links a1, a2, and A3.

Similarly, for each unknown road segment, the recommending device may perform aggregation processing on each unknown road segment to obtain a virtual domain, and the virtual domain may be understood as an area composed of the unknown road segments.

In the embodiment, the named domain and the nameless domain are obtained by combining the aggregation processing, so that the determined named domain and the nameless domain have higher accuracy and reliability, and further, the vehicle driving path recommended by taking the domain as the granularity has the technical effect of higher accuracy and reliability.

In some embodiments, the number of the unknown road segments is plural; the virtual domain is determined by taking any unknown road section as an exploration starting point until the named road section is explored from the undirected graph according to the obtained undirected graph corresponding to the road network, and the virtual domain represents the region between any unknown road section and the explored named road section.

That is, the nameless road segments can be aggregated in a connected component manner according to the undirected graph, thereby virtualizing the domain.

For example, in an undirected graph corresponding to a road network, a peripheral area is searched from a current unknown link as a starting point until a known link is searched, a virtual domain including an area from the starting point (i.e., the current unknown link) to the searched known link is created, and the process is repeated until all virtual domains are obtained.

In the embodiment, the virtual domain is obtained by combining the undirected graph exploration, so that the virtual domain has higher accuracy and effectiveness.

S203: and generating and outputting a vehicle driving path according to the recommended domain sequence and a preset mapping relation. The mapping relation is used for representing the mapping relation between the named road sections and the road names in the road network.

In some embodiments, the mapping relationship is a correspondence relationship between named road sections and road names, which are obtained from a road network and correspond to the named road sections respectively.

Illustratively, the recommended domain sequence is a recommended driving transfer relationship among vehicle domains, the domains comprise a named domain and an unknown domain, and the named domains are obtained based on the aggregation of the named road sections with the same road name, so that after the recommended domain sequence is determined, the vehicle driving path can be determined from the granularity of the 'domains' in combination with the mapping relationship, and the efficiency of determining the vehicle driving path is improved.

In some embodiments, S203 may include: and determining road sections corresponding to the domains in the recommended domain sequence according to the mapping relation, and generating and outputting a vehicle driving path according to the determined road sections.

By combining the mapping relation, the recommended domain sequence with the domain granularity can be converted into the vehicle running path with the road section granularity, so that the vehicle running path with relatively more fine granularity is recommended, the vehicle running requirement of a user is met, and the safety and the reliability of vehicle running are improved.

In some embodiments, the vehicle travel path recommendation request further includes user characteristics of the vehicle user, and the user characteristics are used for characterizing the age, gender, travel preference and the like of the vehicle user so as to determine the vehicle travel path by combining the user characteristics, thereby meeting personalized requirements of the vehicle user and improving the diversity and flexibility of recommendation.

In some embodiments, the path recommendation model may be further constructed to recommend the vehicle travel path based on the path recommendation model, for example, a vehicle travel path recommendation request (including a travel starting point and a travel destination, and also including a user characteristic) may be input to the path recommendation model, and the recommendation domain sequence may be output, so as to generate and output the vehicle travel path according to the recommendation domain sequence and the mapping relationship.

Illustratively, splicing processing is carried out on a driving starting point, a driving end point and user characteristics to obtain splicing characteristics, the splicing characteristics are input into a path recommendation model, a recommendation domain sequence is output, and a vehicle driving path is generated and output according to the recommendation domain sequence and a mapping relation.

Now, a training method of a path recommendation model is described with reference to fig. 3, where fig. 3 is a schematic diagram according to a third embodiment of the disclosure, and as shown in fig. 3, the training method of the path recommendation model according to the embodiment of the disclosure includes:

s301: and carrying out aggregation processing on the famous road sections with the same road name in the road network to obtain the famous domain. And carrying out aggregation processing on the unknown road sections in the road network to obtain the virtual domain.

For example, the executing subject of the present embodiment is a training device of the path recommendation model (hereinafter, referred to as a training device for short), the training device may be the same device as the path recommendation device, or may be a different device, and the present embodiment is not limited thereto.

Similarly, in order to avoid the tedious statements, the technical features of the present embodiment that are the same as those of the above embodiments are not repeated.

In some embodiments, the training device may determine each named segment and each unknown segment from the road network. The training device may acquire, for each of the named links, a road name corresponding to each of the named links, to extract the named links having the same road name from the named links, and aggregate the extracted named links to obtain a named domain, which may be understood as a region composed of the named links having the same road name.

Similarly, for each unknown road section, the training device may perform aggregation processing on each unknown road section to obtain a virtual domain, and the virtual domain may be understood as an area composed of the unknown road sections.

S302: and determining the driving transfer relationship of the vehicles among the domains according to the road network and the acquired sample vehicle driving paths running on the road network. Wherein, the domain comprises a name domain and a virtual domain.

The sample vehicle travel path may be understood as a vehicle travel path generated by a vehicle traveling in a road network. The number of the sample vehicle travel paths is not limited in this embodiment, and may be determined based on a demand, a history, a test, and the like.

For example, the number of sample vehicle travel paths may be relatively greater for scenarios with higher relative training requirements, and may be relatively less for scenarios with lower relative training requirements.

This embodiment can be understood as: according to the road system and the sample vehicle driving path, it can be determined that the vehicle drives from one domain to another domain, and the driving relationship of the vehicle driving from one domain to another domain can refer to that a driving transfer relationship exists between one domain and another domain. If the vehicle can travel from domain a to domain B, it can be said that there is a travel transition relationship between domain a and domain B.

S303: and training according to the driving transfer relation and the sample vehicle driving path to obtain a path recommendation model. The path recommendation model is used for recommending a vehicle driving path.

Based on the above analysis, an embodiment of the present disclosure provides a method for training a path recommendation model, including: and carrying out aggregation processing on the famous road sections with the same road name in the road network to obtain the famous domain. The method comprises the steps of carrying out aggregation processing on unknown road sections in a road network to obtain virtual domains, determining a driving transfer relation of vehicles among the domains according to the road network and obtained sample vehicle driving paths, wherein the domains comprise the famous domains and the virtual domains, and training according to the driving transfer relation and the sample vehicle driving paths to obtain a path recommendation model. In the embodiment, by generating a domain with name and a virtual domain, the driving transfer relationship of the vehicle between the domains is determined, that is, the technical characteristics of the path recommendation model are obtained by training with the domain as a granularity, and compared with training with a road section as the granularity, the training has a relatively small solution space, so that the technical effect of the training efficiency can be improved.

For the reader to more deeply understand the implementation principle of the present disclosure, the above-mentioned embodiment is now explained in more detail with reference to fig. 4.

Fig. 4 is a schematic diagram according to a fourth embodiment of the present disclosure, and as shown in fig. 4, the method for training a path recommendation model according to the embodiment of the present disclosure includes:

s401: and acquiring road names corresponding to the road sections from the road network, and performing aggregation processing on the named road sections with the same road name to obtain named domains.

S402: and acquiring an undirected graph corresponding to the road network, and acquiring unknown road sections without road names in the road network.

The undirected graph includes points and edges, the points can be understood as intersections in the road network, and the edges between two points can be understood as road segments between the two points.

S403: and taking any unknown road section as an exploration starting point until a named road section is explored from the undirected graph, and determining an area between any unknown road section and the explored named road section as a virtual domain.

In the embodiment, the virtual domain is obtained by combining undirected graph exploration, so that the virtual domain has higher accuracy and effectiveness.

S404: and determining the driving transfer relationship of the vehicles among the domains according to the road network and the acquired sample vehicle driving paths running on the road network, and generating a new undirected graph according to the driving transfer relationship.

Wherein the domain comprises the named domain and the virtual domain.

For example, this step may be understood as updating the undirected graph referred to in S402, where in S402 the granularity of the undirected graph is a road segment and the granularity of the new undirected graph is a domain.

In some embodiments, determining a driving transfer relationship between vehicles in a domain according to a road network and an acquired sample vehicle driving path for driving the road network may include: and extracting the connection relation among all domains in the road network, and determining the driving transfer relation according to the connection relation and the sample vehicle driving path.

Illustratively, the driving transfer relationship is determined according to the connection relationship between the named domain and the named domain, the connection relationship between the named domain and the nameless domain, and the connection relationship between the nameless domain and the nameless domain, so that the driving transfer relationship is the transfer information of the vehicle driving with the domain as the granularity.

Because the connection relation can be between the named domains, between the nameless domains or between the named domains and the nameless domains, the driving transfer relation determined by combining the connection relation has the technical effects of higher accuracy and reliability.

In some embodiments, the path recommendation model may be obtained according to a new undirected graph, a driving transfer relationship, and a sample vehicle driving path training, that is, the path recommendation model is obtained by training with "domain" as a granularity, so as to improve the training efficiency.

S405: and constructing a mapping relation between the named road sections and the road names.

For example, in combination with the above analysis, a road name may correspond to a plurality of named road segments, and therefore, in the mapping relationship, a road name may correspond to one named road segment or a plurality of named road segments, but generally, a named road segment corresponds to one road name.

S406: and training to obtain a path recommendation model according to the new undirected graph, the driving transfer relation, the mapping relation and the sample vehicle driving path. The path recommendation model is used for recommending a vehicle driving path.

In combination with the analysis, the path recommendation model is obtained through the new undirected graph training, the training efficiency can be improved, and the mapping relation can represent the corresponding relation between the named road sections and the road names, so that the path recommendation model is obtained through the mapping relation training, the training with relatively fine granularity can be realized, and the accuracy and the effectiveness of the path recommendation model obtained through the training are improved.

In some embodiments, S406 may include the steps of:

the first step is as follows: and determining a road section track sequence of each road section in the road network driven by the vehicle according to the sample vehicle driving path.

The road section track sequence is used for representing each road section driven by the vehicle and the sequence relation among the road sections.

The second step is as follows: and converting the road section track sequence into a domain track sequence according to the mapping relation and the new undirected graph, and training to obtain a path recommendation model according to the road section track sequence, the driving transfer relation and the domain track sequence.

The road section track sequence and the domain track sequence are relative concepts, the road section track sequence is a track sequence with road sections as granularity, the domain track sequence is a track sequence with domains as granularity, if a domain is obtained by aggregating road sections, the named domain is obtained by aggregating the named road sections with common road names, and the virtual domain is obtained by aggregating the unknown road sections, so that the training convergence based on the domain track sequence is faster and the efficiency is higher.

In some embodiments, converting the road segment track sequence into the domain track sequence according to the mapping relation and the new undirected graph may include the following steps:

the first step is as follows: and determining the corresponding domain of each road section according to the mapping relation aiming at each road section in the road section track sequence.

For example, the link track sequence may include a named link and a nameless link, the mapping relationship may represent a correspondence between the named link and the link name, and the named domain is obtained based on aggregation processing of the named links having the same link name, so that for the named link in the link track sequence, the corresponding named domain may be determined according to the mapping relationship, and for the nameless link in the link track sequence, the nameless domain corresponding to the nameless link may be determined.

The second step is as follows: and acquiring the determined connection relation between the domains corresponding to the road sections according to the new undirected graph, and generating a domain track sequence according to the connection relation.

In combination with the analysis, the new undirected graph is the undirected graph with the domain as the granularity, and the domain track sequence is generated by combining the new undirected graph, so that the technical effects of the accuracy and the reliability of the domain track sequence can be improved.

In some embodiments, the road segment track sequence has a track start point and a track end point; according to the road section track sequence, the driving transfer relation and the domain track sequence, training to obtain a path recommendation model, which may include: and training to obtain a path recommendation model according to the track starting point, the track end point, the driving transfer relation and the domain track sequence.

Since the vehicle driving paths corresponding to different track starting points and track ending points may be different, in order to improve the accuracy and reliability of vehicle driving path recommendation, a path recommendation model may be obtained by combining the track starting points and the track ending points.

In some embodiments, user characteristics of the vehicle users corresponding to the road section track sequence, that is, user characteristics of the vehicle users forming the driving path of the sample vehicle, may also be obtained, and the user characteristics are used to characterize the age, gender, driving preference, and the like of the users, and a path recommendation model is obtained by combining with user characteristic training, so as to meet the personalized requirements of the path recommendation model for path recommendation, and improve the flexibility and diversity of path recommendation.

Illustratively, feature splicing processing can be performed on user features, a track starting point, a track end point and a domain track sequence to obtain splicing features, and a path recommendation model is obtained according to a driving transfer relation and the splicing feature training.

The splicing characteristics can be used as input characteristics, the input characteristics can be predicted based on a driving transfer relation to obtain a predicted vehicle driving path, and iterative training is carried out based on a loss value by calculating the loss value between the predicted vehicle driving path and a real vehicle driving path calibrated in advance to obtain a path recommendation model.

Fig. 5 is a schematic diagram according to a fifth embodiment of the present disclosure, and as shown in fig. 5, a route recommendation device 500 of the embodiment of the present disclosure includes:

a first obtaining unit 501, configured to obtain a vehicle travel path recommendation request.

A first determining unit 502, configured to determine a recommended domain sequence according to the vehicle driving path recommendation request, where the recommended domain sequence is a recommended driving transfer relationship between domains of the vehicle, and the domains include a named domain determined based on named road segments in the road network having the same road name and a derived virtual domain determined based on unknown road segments in the road network.

A generating unit 503, configured to generate a vehicle travel path according to the recommended domain sequence.

An output unit 504 for outputting a vehicle travel path.

Fig. 6 is a schematic diagram according to a sixth embodiment of the present disclosure, and as shown in fig. 6, a route recommendation device 600 of the embodiment of the present disclosure includes:

a first obtaining unit 601, configured to obtain a vehicle travel path recommendation request.

In some embodiments, the vehicle travel path recommendation request includes a travel start point, a travel end point, and a user characteristic of a vehicle user of the vehicle travel path recommendation request.

A first determining unit 602, configured to determine a recommended domain sequence according to the vehicle driving path recommendation request, where the recommended domain sequence is a recommended driving transfer relationship between domains of the vehicle, and the domains include a named domain determined based on a named road segment in the road network having the same road name and a derived virtual domain determined based on an unknown road segment in the road network.

As can be seen in fig. 6, in some embodiments, the first determining unit 602 includes:

an input subunit 6021 for inputting a vehicle travel path recommendation request to a pre-trained path recommendation model.

An output subunit 6022 for outputting the recommended domain sequence.

The path recommendation model is obtained by training according to a sample driving transfer relation and a sample vehicle driving path, and the sample driving transfer relation is the driving transfer relation between the vehicle domains determined according to a road network and the obtained sample vehicle driving path.

A generating unit 603 configured to generate a vehicle travel path according to the recommended domain sequence.

In some embodiments, the generating unit 603 is configured to generate the vehicle driving path according to the recommended domain sequence and a preset mapping relationship, where the mapping relationship is used to represent a mapping relationship between a named road segment and a road name in a road network.

As can be seen in fig. 6, in some embodiments, the generating unit 603 includes:

a first determining subunit 6031, configured to determine, according to the mapping relationship, a road segment corresponding to each domain in the recommended domain sequence.

A first generation subunit 6032 configured to generate a vehicle travel path from each of the determined road segments.

An output unit 604 for outputting a vehicle travel path.

Fig. 7 is a schematic diagram of a seventh embodiment of the present disclosure, and as shown, a training apparatus 700 of a path recommendation model according to an embodiment of the present disclosure includes:

the aggregation unit 701 is configured to perform aggregation processing on named road segments with the same road name in a road network to obtain a named domain; and carrying out aggregation processing on the unknown road sections in the road network to obtain the virtual domain.

A second determining unit 702, configured to determine a driving transfer relationship between domains of a vehicle according to a road network and an obtained sample vehicle driving path that runs through the road network, where the domains include a name domain and a virtual domain.

The training unit 703 is configured to obtain a path recommendation model according to the driving transfer relationship and the sample vehicle driving path training, where the path recommendation model is used to recommend the vehicle driving path.

Fig. 8 is a schematic diagram of an eighth embodiment of the present disclosure, and as shown in fig. 8, an apparatus 800 for training a path recommendation model according to an embodiment of the present disclosure includes:

the aggregation unit 801 is configured to perform aggregation processing on named road segments with the same road name in a road network to obtain a named domain; and carrying out aggregation processing on the unknown road sections in the road network to obtain the virtual domain.

In some embodiments, the number of the unknown road segments is plural; a polymerization unit 801 comprising:

an obtaining subunit 8011 is configured to obtain an undirected graph corresponding to a road network.

The exploration sub-unit 8012 is configured to explore any unknown road segment as an exploration starting point until the unknown road segment is explored from the undirected graph.

A third determining subunit 8013 is configured to determine an area between any of the unknown road segments and the explored named road segment as a virtual domain.

A second determining unit 802, configured to determine a driving transfer relationship between domains of a vehicle according to a road network and an obtained sample vehicle driving path that runs through the road network, where the domains include a name domain and a virtual domain.

As can be seen in fig. 8, in some embodiments, the second determining unit 802 includes:

the extracting subunit 8021 is configured to extract a connection relationship between domains in the road network.

And a second determining subunit 8022, configured to determine the driving transfer relationship according to the connection relationship and the sample vehicle driving path.

A second obtaining unit 803, configured to obtain road names corresponding to the named road segments from the road network.

The building unit 804 is configured to build a mapping relationship between the named road segments and the road names.

A third obtaining unit 805, configured to obtain a user characteristic of a vehicle user corresponding to the road segment track sequence.

And a training unit 806, configured to obtain a path recommendation model according to the driving transfer relationship and the sample vehicle driving path training, where the path recommendation model is used to recommend a vehicle driving path.

As can be seen in conjunction with fig. 8, in some embodiments, the training unit 806 includes:

the second generating subunit 8061 is configured to generate an undirected graph corresponding to the road network according to the driving transfer relationship.

The training subunit 8062 is configured to obtain a path recommendation model according to the undirected graph, the driving transfer relationship, and the training of the sample vehicle driving path.

The training subunit 8062 is configured to train to obtain a path recommendation model according to the undirected graph, the driving transfer relationship, the mapping relationship, and the sample vehicle driving path.

In some embodiments, training subunit 8062 includes:

and the determining module is used for determining the road section track sequence of each road section in the road network driven by the vehicle according to the sample vehicle driving path.

And the conversion module is used for converting the road section track sequence into the domain track sequence according to the undirected graph and the mapping relation.

In some embodiments, a conversion module, comprising:

and the determining submodule is used for determining a domain corresponding to each road section according to the mapping relation aiming at each road section in the road section track sequence.

And the obtaining subunit is used for obtaining the determined communication relation between the domains corresponding to the road sections according to the undirected graph.

And the generation submodule is used for generating a domain track sequence according to the communication relation.

And the training module is used for training to obtain a path recommendation model according to the road section track sequence, the driving transfer relation and the domain track sequence.

In some embodiments, the road segment track sequence has a track start point and a track end point; the training module is used for training to obtain a path recommendation model according to the track starting point, the track end point, the driving transfer relation and the domain track sequence.

The training module is used for training to obtain a path recommendation model according to the user characteristics, the track starting point, the track end point, the driving transfer relation and the domain track sequence.

In some embodiments, a training module comprises:

and the splicing submodule is used for performing characteristic splicing processing on the user characteristics, the track starting point, the track end point and the domain track sequence to obtain splicing characteristics.

And the training submodule is used for training according to the driving transfer relation and the splicing characteristic to obtain a path recommendation model.

Fig. 9 is a schematic diagram according to a ninth embodiment of the present disclosure, and as shown in fig. 9, an electronic device 900 in the present disclosure may include: a processor 901 and a memory 902.

A memory 902 for storing programs; the Memory 902 may include a volatile Memory (RAM), such as a Static Random Access Memory (SRAM), a Double Data Rate Synchronous Dynamic Random Access Memory (DDR SDRAM), and the like; the memory may also comprise a non-volatile memory, such as a flash memory. The memory 902 is used to store computer programs (e.g., applications, functional modules, etc. that implement the methods described above), computer instructions, etc., which may be stored in partitions in the one or more memories 902. And the above-described computer programs, computer instructions, data, and the like can be called by the processor 901.

The computer programs, computer instructions, etc. described above may be stored in one or more memories 902 in partitions. And the above-mentioned computer program, computer instruction, etc. can be called by the processor 901.

A processor 901 for executing the computer program stored in the memory 902 to implement the steps of the method according to the above embodiments.

Reference may be made in particular to the description relating to the preceding method embodiment.

The processor 901 and the memory 902 may be separate structures or may be an integrated structure integrated together. When the processor 901 and the memory 902 are separate structures, the memory 902 and the processor 901 may be coupled by a bus 903.

The electronic device of this embodiment may execute the technical solution in the method, and the specific implementation process and technical principle are the same, which are not described herein again.

In the technical scheme of the disclosure, the collection, storage, use, processing, transmission, provision, disclosure and other processing of the personal information (such as user characteristics) of the related user all conform to the regulations of related laws and regulations, and do not violate the customs of the public order.

The present disclosure also provides an electronic device, a readable storage medium, and a computer program product according to embodiments of the present disclosure.

According to an embodiment of the present disclosure, the present disclosure also provides a computer program product comprising: a computer program, stored in a readable storage medium, from which at least one processor of the electronic device can read the computer program, the at least one processor executing the computer program causing the electronic device to perform the solution provided by any of the embodiments described above.

FIG. 10 illustrates a schematic block diagram of an example electronic device 1000 that can be used to implement embodiments of the present disclosure. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital assistants, cellular telephones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be examples only, and are not intended to limit implementations of the disclosure described and/or claimed herein.

As shown in fig. 10, the apparatus 1000 includes a computing unit 1001 that can perform various appropriate actions and processes according to a computer program stored in a Read Only Memory (ROM)1002 or a computer program loaded from a storage unit 1008 into a Random Access Memory (RAM) 1003. In the RAM 1003, various programs and data necessary for the operation of the device 1000 can be stored. The calculation unit 1001, the ROM 1002, and the RAM 1003 are connected to each other by a bus 1004. An input/output (I/O) interface 1005 is also connected to bus 1004.

A number of components in device 1000 are connected to I/O interface 1005, including: an input unit 1006 such as a keyboard, a mouse, and the like; an output unit 1007 such as various types of displays, speakers, and the like; a storage unit 1008 such as a magnetic disk, an optical disk, or the like; and a communication unit 1009 such as a network card, a modem, a wireless communication transceiver, or the like. The communication unit 1009 allows the device 1000 to exchange information/data with other devices through a computer network such as the internet and/or various telecommunication networks.

Computing unit 1001 may be a variety of general and/or special purpose processing components with processing and computing capabilities. Some examples of the computing unit 1001 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various dedicated Artificial Intelligence (AI) computing chips, various computing units running machine learning model algorithms, a Digital Signal Processor (DSP), and any suitable processor, controller, microcontroller, and so forth. The calculation unit 1001 executes the respective methods and processes described above, such as the path recommendation method, the training method of the path recommendation model. For example, in some embodiments, the path recommendation method, the training method of the path recommendation model, may be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 1008. In some embodiments, part or all of the computer program may be loaded and/or installed onto device 1000 via ROM 1002 and/or communications unit 1009. When the computer program is loaded into the RAM 1003 and executed by the computing unit 1001, one or more steps of the path recommendation method, the training method of the path recommendation model described above may be performed. Alternatively, in other embodiments, the computing unit 1001 may be configured by any other suitable means (e.g., by means of firmware) to perform the path recommendation method, the training method of the path recommendation model.

Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuitry, Field Programmable Gate Arrays (FPGAs), Application Specific Integrated Circuits (ASICs), Application Specific Standard Products (ASSPs), system on a chip (SOCs), Complex Programmable Logic Devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs that are executable and/or interpretable on a programmable system including at least one programmable processor, which may be special or general purpose, receiving data and instructions from, and transmitting data and instructions to, a storage system, at least one input device, and at least one output device.

Program code for implementing the methods of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the program codes, when executed by the processor or controller, cause the functions/operations specified in the flowchart and/or block diagram to be performed. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.

In the context of this disclosure, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic, speech, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a back-end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), Wide Area Networks (WANs), and the Internet.

The computer system may include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The Server can be a cloud Server, also called a cloud computing Server or a cloud host, and is a host product in a cloud computing service system, so as to solve the defects of high management difficulty and weak service expansibility in the traditional physical host and VPS service ("Virtual Private Server", or simply "VPS"). The server may also be a server of a distributed system, or a server incorporating a blockchain.

It should be understood that various forms of the flows shown above may be used, with steps reordered, added, or deleted. For example, the steps described in the present disclosure may be executed in parallel, sequentially, or in different orders, and are not limited herein as long as the desired results of the technical solutions disclosed in the present disclosure can be achieved.

The above detailed description should not be construed as limiting the scope of the disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may be made, depending on design requirements and other factors. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present disclosure should be included in the scope of protection of the present disclosure.

Claims

1. A path recommendation method, comprising:

acquiring a vehicle running path recommendation request;

2. The method of claim 1, wherein the generating and outputting the vehicle travel path according to the recommended domain sequence comprises:

and generating and outputting the vehicle driving path according to the recommended domain sequence and a preset mapping relation, wherein the mapping relation is used for representing the mapping relation between the named road sections in the road network and the road names.

3. The method of claim 2, wherein the generating and outputting the vehicle travel path according to the recommended domain sequence and a preset mapping relation comprises:

and determining road sections corresponding to all domains in the recommended domain sequence according to the mapping relation, and generating and outputting the vehicle driving path according to the determined road sections.

4. The method according to claim 2 or 3, wherein the mapping relationship is a correspondence relationship between named road sections and road names, which are obtained from the road network and correspond to the named road sections.

5. The method according to any one of claims 1-4, wherein the named domains are obtained by performing aggregation processing on named road segments with the same road name in the road network; the virtual domain is obtained by performing aggregation processing on the unknown road sections in the road network.

6. The method of claim 5, wherein the number of the anonymous segments is plural; the virtual domain is determined by taking any unknown road section as an exploration starting point until a named road section is explored from the undirected graph according to the acquired undirected graph corresponding to the road network, and the virtual domain represents an area between the any unknown road section and the explored named road section.

7. The method of any of claims 1-6, wherein the determining a recommended domain sequence from the vehicle travel path recommendation request comprises:

and inputting the vehicle driving path recommendation request into a pre-trained path recommendation model, and outputting the recommendation domain sequence, wherein the path recommendation model is obtained by training according to a sample driving transfer relationship and an obtained sample vehicle driving path which is driven to the road network, and the sample driving transfer relationship is the driving transfer relationship between the domains of the vehicle determined according to the road network and the sample vehicle driving path.

8. The method of claim 7, wherein the vehicle travel path recommendation request includes a travel start point, a travel end point, and a user characteristic of a vehicle user of the vehicle travel path recommendation request.

9. A training method of a path recommendation model comprises the following steps:

determining a driving transfer relation of vehicles among domains according to the road network and the obtained sample driving paths of the vehicles driving on the road network, wherein the domains comprise the named domains and the virtual domains;

10. The method of claim 9, wherein the training from the travel transfer relationship and the sample vehicle travel path to derive a path recommendation model comprises:

generating an undirected graph corresponding to the road network according to the driving transfer relation;

and training according to the undirected graph, the driving transfer relation and the sample vehicle driving path to obtain the path recommendation model.

11. The method of claim 10, further comprising:

acquiring road names corresponding to the named road sections from the road network, and constructing a mapping relation between the named road sections and the road names;

and the step of obtaining the path recommendation model according to the undirected graph, the driving transfer relation and the sample vehicle driving path training comprises the following steps: and training to obtain the path recommendation model according to the undirected graph, the driving transfer relation, the mapping relation and the sample vehicle driving path.

12. The method of claim 11, wherein the training the path recommendation model according to the undirected graph, the travel transfer relationship, the mapping relationship, and the sample vehicle travel path comprises:

determining a road section track sequence of each road section of the road network driven by the vehicle according to the sample vehicle driving path;

and converting the road section track sequence into a domain track sequence according to the undirected graph and the mapping relation, and training to obtain the path recommendation model according to the road section track sequence, the driving transfer relation and the domain track sequence.

13. The method of claim 12, wherein the converting the sequence of road segment trajectories to a sequence of domain trajectories according to the undirected graph and the mapping relationship comprises:

determining a domain corresponding to each road section according to the mapping relation aiming at each road section in the road section track sequence;

and acquiring the determined communication relation between the domains corresponding to the road sections according to the undirected graph, and generating the domain track sequence according to the communication relation.

14. The method according to claim 12 or 13, wherein the sequence of road segment trajectories has a trajectory start point and a trajectory end point; the training to obtain the path recommendation model according to the road section track sequence, the driving transfer relation and the domain track sequence comprises:

and training to obtain the path recommendation model according to the track starting point, the track end point, the driving transfer relation and the domain track sequence.

15. The method of claim 14, further comprising:

acquiring user characteristics of vehicle users corresponding to the road section track sequence;

and training to obtain the path recommendation model according to the track starting point, the track end point, the driving transfer relation and the domain track sequence, wherein the training comprises: and training to obtain the path recommendation model according to the user characteristics, the track starting point, the track end point, the driving transfer relation and the domain track sequence.

16. The method of claim 15, wherein the training the path recommendation model according to the user characteristics, the track start point, the track end point, the driving transfer relationship, and the domain track sequence comprises:

performing feature splicing processing on the user features, the track starting point, the track end point and the domain track sequence to obtain splicing features;

and training according to the driving transfer relation and the splicing characteristic to obtain the path recommendation model.

17. The method according to any one of claims 9 to 16, wherein determining a driving transfer relationship between vehicles in a domain according to the road network and the acquired sample driving paths of the vehicles driving the road network comprises:

and extracting the connection relation among all domains in the road network, and determining the driving transfer relation according to the connection relation and the driving path of the sample vehicle.

18. The method of any of claims 9-17, wherein the number of the unknown road segments is plural; the aggregating the unknown road sections in the road network to obtain the virtual domain comprises:

obtaining an undirected graph corresponding to the road network;

and taking any unknown road section as an exploration starting point until a named road section is explored from the undirected graph, and determining an area between the unknown road section and the explored named road section as a virtual domain.

19. A path recommendation device comprising:

an output unit for outputting the vehicle travel path.

20. The device of claim 19, wherein the generating unit is configured to generate the vehicle driving path according to the recommended domain sequence and a preset mapping relationship, wherein the mapping relationship is used for representing a mapping relationship between a named road segment in a road network and a road name.

21. The apparatus of claim 20, wherein the generating unit comprises:

the first determining subunit is configured to determine, according to the mapping relationship, a road segment corresponding to each domain in the recommended domain sequence;

and the first generation subunit is used for generating the vehicle running path according to the determined road sections.

22. The apparatus according to claim 20 or 21, wherein the mapping relationship is a correspondence relationship between named road sections and road names, which are obtained from the road network and are corresponding to the named road sections.

23. The apparatus according to any one of claims 19-22, wherein the named domain is obtained by performing an aggregation process on named road segments having the same road name in the road network; the virtual domain is obtained by performing aggregation processing on the unknown road sections in the road network.

24. The apparatus of claim 23, wherein the number of the anonymous segments is plural; the virtual domain is determined by taking any unknown road section as an exploration starting point until a named road section is explored from the undirected graph according to the acquired undirected graph corresponding to the road network, and the virtual domain represents an area between the any unknown road section and the explored named road section.

25. The apparatus according to any one of claims 19-24, wherein the first determining unit comprises:

the input subunit is used for inputting the vehicle running path recommendation request to a pre-trained path recommendation model;

an output subunit, configured to output the recommended domain sequence;

the path recommendation model is obtained by training according to a sample driving transfer relationship and an acquired sample vehicle driving path which drives the road network, and the sample driving transfer relationship is the driving transfer relationship between the domains of the vehicles determined according to the road network and the sample vehicle driving path.

26. The apparatus of claim 25, wherein the vehicle travel path recommendation request includes a travel start point, a travel end point, and a user characteristic of a vehicle user of the vehicle travel path recommendation request.

27. A training apparatus for a path recommendation model, comprising:

28. The apparatus of claim 27, wherein the training unit comprises:

a second generation subunit, configured to generate an undirected graph corresponding to the road network according to the driving transfer relationship;

and the training subunit is used for obtaining the path recommendation model according to the undirected graph, the driving transfer relation and the sample vehicle driving path training.

29. The apparatus of claim 28, the apparatus further comprising:

the second acquisition unit is used for acquiring road names corresponding to the named road sections from the road network;

the construction unit is used for constructing a mapping relation between the named road sections and the road names;

and the training subunit is configured to train to obtain the path recommendation model according to the undirected graph, the driving transfer relationship, the mapping relationship, and the sample vehicle driving path.

30. The apparatus of claim 29, wherein the training subunit comprises:

the determining module is used for determining a road section track sequence of each road section in the road network driven by the vehicle according to the sample vehicle driving path;

the conversion module is used for converting the road section track sequence into a domain track sequence according to the undirected graph and the mapping relation;

and the training module is used for training to obtain the path recommendation model according to the road section track sequence, the driving transfer relation and the domain track sequence.

31. The apparatus of claim 30, wherein the conversion module comprises:

the determining submodule is used for determining a domain corresponding to each road section according to the mapping relation aiming at each road section in the road section track sequence;

the obtaining submodule is used for obtaining the determined communication relation between the domains corresponding to the road sections according to the undirected graph;

and the generation submodule is used for generating the domain track sequence according to the communication relation.

32. The apparatus of claim 30 or 31, wherein the sequence of segment trajectories has a trajectory start point and a trajectory end point; the training module is used for training to obtain the path recommendation model according to the track starting point, the track end point, the driving transfer relation and the domain track sequence.

33. The apparatus of claim 32, the apparatus further comprising:

the third acquisition unit is used for acquiring the user characteristics of the vehicle users corresponding to the road section track sequence;

and the training module is used for training to obtain the path recommendation model according to the user characteristics, the track starting point, the track end point, the driving transfer relation and the domain track sequence.

34. The apparatus of claim 33, wherein the training module comprises:

the splicing submodule is used for performing characteristic splicing processing on the user characteristics, the track starting point, the track end point and the domain track sequence to obtain splicing characteristics;

and the training sub-module is used for training according to the driving transfer relation and the splicing characteristic to obtain the path recommendation model.

35. The apparatus according to any of claims 27-34, wherein the second determining unit comprises:

the extraction subunit is used for extracting the connection relation among the domains in the road network;

and the second determining subunit is used for determining the driving transfer relationship according to the connection relationship and the sample vehicle driving path.

36. The apparatus of any one of claims 27-35, wherein the number of the unknown road segments is plural; the polymerization unit includes:

an obtaining subunit, configured to obtain an undirected graph corresponding to the road network;

the exploration subunit is used for taking any unknown road section as an exploration starting point until a named road section is explored from the undirected graph;

and the third determining subunit is used for determining an area between the unknown road section and the searched named road section as a virtual domain.

37. An electronic device, comprising:

at least one processor; and

the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-8; or to enable the at least one processor to perform the method of any of claims 9-18.

38. A non-transitory computer readable storage medium having stored thereon computer instructions for causing the computer to perform the method of any one of claims 1-8; alternatively, the computer instructions are for causing the computer to perform the method of any of claims 9-18.

39. A computer program product comprising a computer program which, when executed by a processor, carries out the steps of the method of any one of claims 1 to 8; alternatively, the computer program when executed by a processor implements the steps of the method of any one of claims 9 to 18.