CN113538920B

CN113538920B - Method and device for determining target driving route, electronic equipment and storage medium

Info

Publication number: CN113538920B
Application number: CN202111075695.6A
Authority: CN
Inventors: 刘志煌
Original assignee: Tencent Technology Shenzhen Co Ltd
Current assignee: Tencent Technology Shenzhen Co Ltd
Priority date: 2021-09-14
Filing date: 2021-09-14
Publication date: 2021-12-10
Anticipated expiration: 2041-09-14
Also published as: CN113538920A

Abstract

The application relates to the technical field of artificial intelligence, in particular to a method and a device for determining a target driving route, electronic equipment and a storage medium, which are applied to a vehicle-mounted scene, wherein the method comprises the following steps: responding to a destination selection operation triggered by a target object, and obtaining at least two candidate driving routes between the target object and the destination; obtaining at least one candidate place with designated resources corresponding to at least two candidate driving routes respectively, wherein each candidate place is located in a peripheral preset range of a destination; respectively obtaining flow characteristic association degrees and resource characteristic association degrees between at least two candidate driving routes and each reference recommended route by adopting a trained route recommendation model; and selecting the candidate driving route of which the corresponding flow characteristic association degree and the corresponding resource characteristic association degree both meet the preset association degree condition from the at least two candidate driving routes as a target driving route. In this way, the accuracy of determining the target travel route can be improved.

Description

Method and device for determining target driving route, electronic equipment and storage medium

Technical Field

The present application relates to the field of artificial intelligence technologies, and in particular, to a method and an apparatus for determining a target driving route, an electronic device, and a storage medium.

Background

With the rapid increase of the vehicle holding amount, parking difficulty becomes one of the problems to be solved.

In the related art, in order to solve the above problem, a prediction model based on a deep neural network is generally used to predict the number of vacant parking spaces when a vehicle reaches a parking lot during driving, so that a route from a current position to the parking lot with a large number of vacant parking spaces can be generated.

However, if the actual time of arrival at the parking lot is later than the predicted time while the vehicle is traveling along the generated route, there is a possibility that the parking lot may have no vacant parking spaces.

Therefore, such a route recommendation method in the related art is not highly accurate.

Disclosure of Invention

The embodiment of the application provides a method and a device for determining a target driving route, electronic equipment and a storage medium, so as to improve the accuracy of determining the target driving route.

The embodiment of the application provides the following specific technical scheme:

a method of determining a target travel route, comprising:

responding to a destination selection operation triggered by a target object, and obtaining at least two candidate driving routes between the target object and the destination;

obtaining at least one candidate place with designated resources corresponding to the at least two candidate driving routes respectively, wherein each candidate place is located in a peripheral preset range of the destination;

respectively obtaining flow characteristic association degrees and resource characteristic association degrees between at least two candidate driving routes and each reference recommended route by adopting a trained route recommendation model; each reference recommended route is a route which meets recommended conditions and is determined according to historical flow and historical resource quantity;

and selecting the candidate driving route of which the corresponding flow characteristic association degree and the corresponding resource characteristic association degree both meet the preset association degree condition from the at least two candidate driving routes as a target driving route.

A target travel route determination apparatus, comprising:

the first obtaining module is used for responding to destination selection operation triggered by a target object and obtaining at least two candidate running routes between the target object and the destination;

the second obtaining module is used for obtaining at least one candidate place with specified resources corresponding to the at least two candidate driving routes respectively, wherein each candidate place is located in a peripheral preset range of the destination;

the first processing module is used for respectively obtaining the traffic characteristic association degree and the resource characteristic association degree between at least two candidate driving routes and each reference recommended route by adopting the trained route recommendation model; each reference recommended route is a route which meets recommended conditions and is determined according to historical flow and historical resource quantity;

and the screening module is used for selecting the candidate driving route of which the corresponding traffic characteristic association degree and the corresponding resource characteristic association degree both meet the preset association degree condition from the at least two candidate driving routes as a target driving route.

Optionally, the first processing module is further configured to:

respectively obtaining flow characteristics and resource characteristics corresponding to the at least two candidate driving routes;

adopting the trained route recommendation model, and aiming at the at least two candidate driving routes, respectively executing the following operations:

respectively obtaining the flow characteristic correlation degree between the flow characteristic of one candidate driving route and the flow characteristic of each reference recommended route; and the number of the first and second groups,

and respectively obtaining the resource characteristic correlation degree between the resource characteristic of the candidate driving route and the resource characteristic of each reference recommended route.

Optionally, when the flow characteristics corresponding to the at least two candidate driving routes are respectively obtained, the first processing module is further configured to:

for the at least two candidate driving routes, respectively performing the following operations:

respectively obtaining each running road section contained in one candidate running route, historical flow corresponding to each historical sampling time period, and respectively obtaining current flow corresponding to each running road section in the current sampling time period;

respectively determining flow change information of each driving road section according to each corresponding historical flow of each driving road section;

and determining the flow characteristic of the candidate driving route according to the obtained current flow and flow change information corresponding to each driving road section and the object density characteristic of the candidate driving route.

Optionally, when the resource features corresponding to the at least two candidate driving routes are respectively obtained, the first processing module is further configured to:

respectively obtaining resource change information corresponding to at least one candidate of one candidate running route in each historical sampling period;

and determining the resource characteristics corresponding to the candidate driving route based on the obtained resource change information according to a preset characteristic extraction mode.

Optionally, when the resource feature corresponding to the candidate driving route is determined based on the obtained resource change information according to a preset feature extraction manner, the first processing module is further configured to:

for each historical sampling time period, generating a corresponding information sequence based on the resource change information of at least one candidate corresponding to the candidate running route in the corresponding historical sampling time period;

obtaining a candidate information set corresponding to the at least one candidate place based on the obtained information sequences, wherein each candidate information is resource change information of which the occurrence frequency meets a preset frequency condition in the resource change information corresponding to the corresponding candidate place;

sequentially reading each candidate information contained in the candidate information set, and executing the following operations when reading each candidate information: taking each information sequence containing one currently read candidate information as a candidate information sequence, and respectively generating corresponding information sequence characteristics in each candidate information sequence based on the candidate information and other resource change information; the other resource change information is resource change information which is arranged behind the candidate information in each candidate information sequence according to a preset sequence;

and selecting each target information sequence feature which meets the condition of preset quantity from each information sequence feature as a resource feature corresponding to the candidate driving route based on the total number of the resource change information associated with each information sequence feature.

Optionally, when each information sequence including one currently read candidate information is used as a candidate information sequence, and corresponding information sequence features are generated in each candidate information sequence based on the one candidate information and other resource change information, the first processing module is further configured to:

taking the currently read candidate information as an initial information sequence characteristic;

based on each candidate information sequence obtained currently, the following steps are executed in a circulating manner until the occurrence frequency of each other resource change information is determined to not meet the preset frequency condition:

adding other resource change information of which the occurrence frequency meets the frequency condition in the other resource change information contained in each candidate information sequence to corresponding initial information sequence characteristics respectively to obtain each updated information sequence characteristic;

respectively screening out each candidate information sequence containing each updated information sequence characteristic from each candidate information sequence to serve as a corresponding updated candidate information sequence;

and aiming at each updated information sequence characteristic, respectively executing the following operations: respectively arranging resource change information behind the characteristics of one updated information sequence in each updated candidate information sequence according to a preset sequence to serve as other updated resource change information;

and taking each updated information sequence characteristic as each new initial information sequence characteristic respectively.

Optionally, when obtaining a flow characteristic association degree between a flow characteristic of a candidate driving route and a flow characteristic of a reference recommended route, the first processing module is specifically configured to:

sequentially reading each flow sub-feature corresponding to the flow feature of one candidate driving route, and when one flow sub-feature is read, determining the reference flow sub-feature correlation degree corresponding to the reference flow sub-feature as the flow sub-feature correlation degree corresponding to the one flow sub-feature when the reference flow sub-feature same as the one flow sub-feature exists in each reference flow sub-feature corresponding to the flow feature of one reference recommended route, wherein each reference flow sub-feature correlation degree characterizes the correlation degree between the reference flow sub-feature and each reference recommended route;

and accumulating the obtained flow sub-feature association degrees, and taking an accumulated result as the flow feature association degree corresponding to the candidate driving route and the reference recommended route.

Optionally, when obtaining the resource feature association degree between the flow feature of one candidate driving route and the resource feature of one reference recommended route, the first processing module is further configured to:

sequentially reading each resource sub-feature corresponding to the resource feature of one candidate driving route, and when one resource sub-feature is read, determining the reference resource sub-feature correlation degree corresponding to the reference resource sub-feature as the resource sub-feature correlation degree corresponding to one resource sub-feature when the reference resource sub-feature same as the one resource sub-feature exists in each reference resource sub-feature corresponding to the resource feature of one reference recommended route, wherein the reference resource sub-feature correlation degree corresponding to each reference resource sub-feature characterizes the correlation degree between the reference resource sub-feature and each reference recommended route;

and accumulating the obtained resource sub-feature association degrees, and taking an accumulated result as the resource feature association degree corresponding to the candidate driving route and the reference recommended route.

Optionally, when training the route recommendation model, the method further includes:

a third obtaining module, configured to obtain a sample set, where the sample set includes each reference recommended route and each non-recommended route, each reference recommended route and each non-recommended route respectively corresponds to a candidate feature, the candidate feature includes multiple candidate sub-features, and each candidate sub-feature is a resource sub-feature or a traffic sub-feature;

the classification module is used for classifying the candidate sub-features of the reference recommended routes and the non-recommended routes respectively to obtain various candidate sub-features;

the second processing module is used for respectively obtaining sub-feature association degrees corresponding to the various candidate sub-features, and taking the candidate sub-features with the sub-feature association degrees larger than a preset sub-feature association degree threshold value as reference sub-features, wherein the sub-feature association degrees are probabilities that the corresponding sub-features are included in the candidate sub-features of the reference recommended routes;

and the training module is used for taking the flow sub-features in the reference sub-features as the reference flow sub-features learned by the route recommendation model, and taking the resource sub-features in the reference sub-features as the reference resource sub-features learned by the route recommendation model.

Optionally, when obtaining the respective sub-feature association degrees corresponding to the various candidate sub-features, the second processing module is further configured to:

for the various candidate sub-features, performing the following operations, respectively:

determining a first sample probability according to the number of the reference recommended routes and the total number of samples of the sample set, wherein the first sample probability characterizes the occurrence probability of the reference recommended routes in the sample set;

obtaining a second sample probability according to the number of the reference recommended routes and the non-recommended routes containing the candidate sub-feature and the total number of the samples, wherein the second sample probability is characterized in the sample set and the occurrence probability of the candidate sub-feature is obtained;

determining a third sample probability according to the number of the reference recommended routes containing the candidate sub-feature and the number of the reference recommended routes, wherein the third sample probability characterizes the occurrence probability of the candidate sub-feature in the reference recommended routes;

and determining the association degree of the sub-features corresponding to the candidate sub-features according to the first sample probability, the second sample probability and the third sample probability.

Optionally, the screening module is further configured to:

respectively aiming at the at least two candidate driving routes, respectively executing the following operations: accumulating the traffic characteristic association degrees corresponding to one candidate driving route to obtain a traffic characteristic association degree sum, and accumulating the resource characteristic association degrees corresponding to the candidate driving route to obtain a resource characteristic association degree sum;

selecting a corresponding flow characteristic association degree sum and a candidate driving route of which the resource characteristic association degree sum is not less than a preset association degree threshold from the at least two candidate driving routes;

aiming at least two candidate driving routes which are selected, the following operations are respectively executed: adding the sum of the flow characteristic relevance degrees of the selected candidate driving route and the sum of the resource characteristic relevance degrees to obtain a corresponding sum of the characteristic relevance degrees;

and taking the candidate driving route corresponding to the characteristic association degree sum with the maximum value as a target driving route.

An electronic device provided by an embodiment of the present application includes a processor and a memory, where the memory stores program codes, and when the program codes are executed by the processor, the processor is caused to execute any one of the steps of the method for determining a target driving route.

A computer-readable storage medium provided in an embodiment of the present application includes program code, when the program code runs on an electronic device, the program code is configured to cause the electronic device to perform any one of the steps of the method for determining a target driving route.

The beneficial effect of this application is as follows:

the embodiment of the application provides a method and a device for determining a target driving route, electronic equipment and a storage medium. In the embodiment of the application, in response to a destination selection operation triggered by a target object, at least two candidate driving routes between the target object and the destination are obtained, at least one candidate place with a specified resource and corresponding to each of the at least two candidate driving routes is obtained, then, a trained route recommendation model is adopted, flow characteristic association degrees and resource characteristic association degrees between the at least two candidate driving routes and each reference recommended route are respectively obtained, and finally, a candidate driving route, of which the corresponding flow characteristic association degrees and the corresponding resource characteristic association degrees meet a preset association degree condition, is selected from the at least two candidate driving routes to serve as the target driving route.

Therefore, the target driving route is determined from at least two candidate driving routes according to the flow characteristic of each candidate driving route and the resource characteristic of the candidate place which can be reached by each candidate driving route, so that the error between the actual time of the target object reaching the candidate place and the predicted time can be reduced, the sufficient specified resources in the candidate place are ensured when the target object reaches the candidate place, and the accuracy of recommending the route to the candidate place to the target object is improved; and the route recommendation is carried out by determining the traffic characteristic correlation degree and the resource characteristic correlation degree between the at least two candidate driving routes and each item mark driving route, so that the calculation amount of the route recommendation can be reduced, the real-time performance of the route recommendation is ensured, and the accuracy of the route recommendation is further improved.

Additional features and advantages of the application will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by the practice of the application. The objectives and other advantages of the application may be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.

Drawings

FIG. 1 is a schematic diagram of an application scenario in an embodiment of the present application;

FIG. 2A is a schematic diagram of a training process of a route recommendation model in an embodiment of the present application;

FIG. 2B is a schematic flow chart illustrating the process of obtaining sub-feature correlation in the embodiment of the present application;

FIG. 3A is a schematic diagram illustrating a process of determining a target driving route according to an embodiment of the present application;

FIG. 3B is a schematic diagram of a candidate driving route in the embodiment of the present application;

fig. 4A is a schematic flow chart illustrating obtaining a traffic characteristic relevance degree and a resource characteristic relevance degree of at least two candidate driving routes in the embodiment of the present application;

FIG. 4B is a schematic flow chart illustrating the flow characteristic obtained in the embodiment of the present application;

FIG. 4C is a schematic diagram illustrating a relationship between a candidate driving route and a driving road segment in the embodiment of the present application;

FIG. 4D is a schematic diagram of a driving section included in a candidate driving route in the embodiment of the present application;

fig. 4E is a schematic diagram of a historical traffic corresponding relationship in the embodiment of the present application;

fig. 4F is a schematic diagram of a traffic sequence information mapping relationship in the embodiment of the present application;

FIG. 4G is a schematic flow chart illustrating resource feature acquisition according to an embodiment of the present application;

FIG. 4H is a schematic diagram illustrating an association relationship between a candidate driving route and a parking lot in an embodiment of the application;

fig. 4I is a schematic diagram illustrating a corresponding relationship between vacant parking spaces in a parking lot in the embodiment of the present application;

FIG. 4J is a schematic diagram illustrating a resource data mapping relationship of a candidate driving route 1 according to an embodiment of the present application;

FIG. 4K is a schematic diagram illustrating a change in the number of vacant parking spaces in the embodiment of the present application;

FIG. 4L is a schematic diagram illustrating resource data increase and decrease in an embodiment of the present application;

fig. 4M is a schematic diagram illustrating an increase and decrease trend of an empty parking space in the embodiment of the present application;

FIG. 4N is a schematic flow chart illustrating the determination of resource characteristics according to an embodiment of the present application;

FIG. 4O is a diagram illustrating an information sequence in an embodiment of the present application;

FIG. 4P is a schematic diagram of an information sequence of a candidate driving route 1 according to an embodiment of the present application;

fig. 4Q is a schematic diagram of increasing and decreasing parking spaces of the candidate driving route 1 in the embodiment of the present application;

fig. 4R is a schematic diagram of a corresponding relationship between resource change information and occurrence frequency in the embodiment of the present application;

FIG. 4S is a schematic flow chart illustrating the generation of an information sequence in an embodiment of the present application;

FIG. 4T is a schematic flowchart illustrating an exemplary process for determining an initial information sequence according to an embodiment of the present disclosure;

FIG. 4U is a schematic diagram illustrating a process of mining resource features in an embodiment of the present application;

FIG. 4V is a schematic flow chart illustrating a process for determining relevance of flow characteristics in the practice of the present application;

FIG. 4W is a schematic view of a process for determining resource feature association in the practice of the present application;

FIG. 5A is a schematic flow chart illustrating the determination of a target driving route according to an embodiment of the present application;

FIG. 5B is a flow chart illustrating route recommendation in an embodiment of the present application;

fig. 6A is a schematic structural diagram of a target driving route determining apparatus according to an embodiment of the present application;

FIG. 6B is a schematic structural diagram of a training apparatus for a route recommendation model according to an embodiment of the present application;

fig. 7 is a schematic diagram of a hardware component structure of an electronic device in an embodiment of the present application;

fig. 8 is a schematic structural diagram of a computing device in an embodiment of the present application.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some embodiments, but not all embodiments, of the technical solutions of the present application. All other embodiments obtained by a person skilled in the art without any inventive step based on the embodiments described in the present application are within the scope of the protection of the present application.

The terms "first," "second," and the like in the description and in the claims of the present application and in the above-described drawings are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are capable of operation in sequences other than those illustrated or described herein.

Some terms in the embodiments of the present application are explained below to facilitate understanding by those skilled in the art.

Target object: refers to a terminal for triggering a destination selection operation. For example, when the intelligent parking navigation system is applied to an intelligent parking navigation scene, the target object is a terminal used by a user in the process of driving a vehicle.

Candidate driving routes: refers to a candidate route for the target object to travel from the current location to the destination.

Alternatively: refers to a place having a designated resource within a preset range around a destination. For example, when applied to an intelligent parking guidance scenario, the candidate is a parking lot disposed within a preset range around the destination, and the resource is designated as each parking space disposed within the parking lot.

Referring to the recommended route: the route meeting the preset recommendation condition is obtained from the historical driving route according to the historical flow and the historical resource quantity. The reference recommended route may be manually pre-labeled.

Flow characteristic correlation degree: the feature correlation degree corresponding to the flow feature between one candidate driving route and one reference recommended route is referred to.

Resource feature association degree: the feature correlation degree corresponding to the resource feature between one candidate driving route and one reference recommended route is referred to.

Flow characteristics: the flow characteristic of each candidate driving route is formed by a plurality of flow sub-characteristics and is determined according to the flow change information, the current flow and the object density characteristics of the candidate driving routes.

Resource characteristics: the resource feature of each candidate driving route is determined according to the resource change information corresponding to at least one candidate of the candidate driving routes, and the resource feature of each candidate driving route is composed of a plurality of resource sub-features.

Reference flow sub-feature: the flow rate estimation method is obtained from a plurality of candidate flow rate sub-features contained in candidate flow rate features corresponding to all reference recommended routes, and the reference flow rate sub-features represent the candidate flow rate sub-features with the occurrence probability exceeding a threshold value in all the reference recommended routes.

Reference resource sub-feature: the method is obtained from a plurality of candidate resource sub-features contained in candidate resource features corresponding to all reference recommended routes, and the reference resource sub-features represent the candidate resource sub-features with the occurrence probability exceeding a threshold value in all the reference recommended routes.

The following briefly introduces the design concept of the embodiments of the present application:

with the rapid development of cities, the holding amount of vehicles is also increased. When the vehicle owner drives to the destination, the vehicle owner may re-drive to a new parking lot due to inconvenience of parking at the destination, and therefore, the vehicle owner usually searches the number of vacant parking spaces in the parking lot before departure, and generates a driving route from the current position to the parking lot with more vacant parking spaces.

In the related art, a prediction model based on a deep neural network is usually adopted, and in the driving process, the number of vacant parking spaces when a vehicle reaches a parking lot is predicted, so that a driving route from a current position to the parking lot with more vacant parking spaces is generated.

However, when the vehicle travels along the predetermined route, if the actual time of arrival at the parking lot is later than the predicted time, the parking lot may have no free space. The car owner needs to pocket and turn in peripheral parking area to the parking area that seeks to have vacant parking stall carries out parking of vehicle. Therefore, the accuracy of such a route recommendation method in the related art is not high.

In view of this, the present application provides a method, an apparatus, an electronic device, and a storage medium for determining a target driving route. According to the method and the device, the target driving route is determined from at least two candidate driving routes based on the flow characteristics of each candidate driving route and the resource characteristics of the candidate places which can be reached, so that route recommendation is performed by combining the flow characteristics and the resource characteristics of the candidate driving routes, the time difference between the actual time when the target object reaches the candidate places and the predicted time can be reduced, sufficient specified resources can be guaranteed when the candidate places are reached, and the accuracy of the route recommendation is improved. In addition, in the embodiment of the application, the traffic characteristic association degree and the resource characteristic association degree are used for recommending the route, so that the calculation amount during the route recommendation can be reduced, and the real-time performance of the route recommendation can be ensured.

The preferred embodiments of the present application will be described in conjunction with the drawings of the specification, it should be understood that the preferred embodiments described herein are for purposes of illustration and explanation only and are not intended to limit the present application, and features of the embodiments and examples of the present application may be combined with each other without conflict.

Fig. 1 is a schematic view of an application scenario in the embodiment of the present application. The application scenario diagram includes terminal device 110 and server 120. The terminal device 110 and the server 120 may communicate with each other via a communication network.

In the embodiment of the present application, when the terminal device 110 is a vehicle-mounted terminal device, the terminal device 110 is a device embedded in a vehicle. When terminal device 110 is a client, then terminal device 110 is a device used by the user. The terminal device 110 is pre-installed with a target application having a function of recommending a target travel route, and the function of the target application is not limited to recommending a target travel route. The target application may be a pre-installed client application, a web page version application, an applet, or the like. Terminal device 110 may include one or more processors 1101, memory 1102, I/O interface 1103 and display 1104, among other things, that interact with server 120. The terminal device 110 includes, but is not limited to, a mobile phone, a computer, an intelligent voice interaction device, an intelligent appliance, a vehicle-mounted terminal, and the like.

The server is a background server corresponding to the target application and provides service for the target application. Server 120 may include one or more processors 1201, memory 1202, and I/O interface 1203 or the like that interacts with terminal device 110. The server 120 may be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server providing basic cloud computing services such as a cloud service, a cloud database, cloud computing, a cloud function, cloud storage, a Network service, cloud communication, a middleware service, a domain name service, a security service, a Content Delivery Network (CDN), a big data and artificial intelligence platform, and the like. The terminal device 110 and the server 120 may be directly or indirectly connected through wired or wireless communication, and the embodiment of the present application is not limited herein.

It should be noted that the method for determining a target driving route disclosed in the present application may include a plurality of servers, wherein the plurality of servers may be configured as a blockchain, and each server is a node on the blockchain.

The determination of the target driving route in the embodiment of the present application may be performed on the terminal device 110, or may be performed on the server 120. When the terminal device 110 determines the target driving route, the terminal device 110 responds to a destination selection operation triggered by the target object to obtain at least two candidate driving routes between the target object and the destination, and selects a candidate driving route from the at least two candidate driving routes, wherein the corresponding traffic characteristic correlation degree and the corresponding resource characteristic correlation degree both meet a preset correlation degree condition, as the target driving route by using a trained route recommendation model. When the server 120 determines the target driving route, when the terminal device 110 obtains a destination selection operation triggered by the target object, the obtained destination selection operation is sent to the server 120, and then the server 120 responds to the destination selection operation to obtain at least two candidate driving routes between the target object and the destination, and selects a candidate driving route, of which the corresponding traffic characteristic correlation degree and the resource characteristic correlation degree both meet a preset correlation degree condition, from the at least two candidate driving routes by using a trained route recommendation model, as the target driving route, and feeds the target driving route back to the terminal device 110.

In the embodiment of the application, the target driving route determination method can be applied to scenes of intelligent parking, for example, when a vehicle owner needs to park near a destination, the destination selection operation can be triggered and generated in the driving process, the destination selection operation is sent to the server through the terminal device, the server responds to the destination selection operation triggered by the vehicle owner, the target driving route is determined from at least two candidate driving routes, and then the vehicle owner can drive to a parking lot near the destination according to the determined target driving route, so that the situation that the actual arrival time of the vehicle owner to the parking lot is later than the predicted time due to traffic jam can be avoided, the situation that the parking lot has no vacant parking spaces is caused, and the route recommendation accuracy is improved.

In some possible embodiments of the present application, the determination of the target driving route may also be applied to a scenario of planning a traffic route, for example, a traffic road planner may implement a rational planning of a traffic road according to a traffic flow, a destination, and a number of designated resources when reaching the destination.

A process of training a route recommendation model in the embodiment of the present application is described below with reference to the drawings, where the training process in the embodiment of the present application is applicable to the terminal device 110 or the server 120 shown in fig. 1, and a specific training process is as follows:

in the embodiment of the application, an initial route recommendation model can be selectively built based on a Bayesian formula, and inputting the candidate sub-features corresponding to each reference recommended route and each non-recommended route in the sample set into an initial route recommendation model, according to the Bayesian formula in the initial route recommendation model, determining the reference sub-features which are required to be learned by the initial route recommendation model from the candidate sub-features, and training the initial route recommendation model by using the determined reference sub-characteristics to obtain a trained route recommendation model, the resource characteristic association degree and the flow characteristic association degree between the at least two candidate driving routes and each reference recommended route are determined according to the trained route recommendation model, and then the corresponding target driving route is determined from the at least two candidate driving routes.

Referring to fig. 2A, a schematic diagram of a training process of a route recommendation model in the embodiment of the present application is shown, and the following describes the training process of the route recommendation model in the embodiment of the present application with reference to fig. 2A:

step 20: obtaining a sample set; the sample set comprises all reference recommended routes and all non-recommended routes, all the reference recommended routes and all the non-recommended routes respectively correspond to candidate features, the candidate features comprise a plurality of candidate sub-features, and each candidate sub-feature is a resource sub-feature or a flow sub-feature.

In the embodiment of the application, a sample set comprising each reference recommended route and each non-recommended route is obtained, each reference recommended route corresponds to a candidate feature, each non-recommended route corresponds to a candidate feature, and each candidate feature is composed of a plurality of candidate sub-features.

It should be noted that the candidate feature is a resource feature or a flow feature, when the candidate feature is the resource feature, each candidate sub-feature of the candidate feature is a resource sub-feature, and when the candidate feature is the flow sub-feature, each candidate sub-feature of the candidate feature is a flow sub-feature.

The candidate sub-feature may be a resource sub-feature, for example, a may be a, increase-B, decrease-C, increase-D, and the like, and the candidate sub-feature may also be a flow sub-feature, for example, may be [45,50,70], [0], [1], and the like, which is not limited in the embodiment of the present application.

Step 21: and classifying the candidate sub-features of each reference recommended route and each non-recommended route respectively to obtain various candidate sub-features.

In the embodiment of the application, since the candidate sub-features corresponding to each reference recommended route and each non-recommended route may include the same candidate sub-features, before calculating the sub-feature association degree corresponding to the candidate sub-features, the candidate sub-features need to be classified first, and the category corresponding to each candidate sub-feature is determined, so as to obtain various candidate sub-features.

For example, assuming that a plurality of candidate sub-features of "a increase-B decrease-C increase" appear in each reference recommended route, the plurality of "a increase-B decrease-C increase" appearing are classified into one category.

Step 22: respectively obtaining sub-feature association degrees corresponding to various candidate sub-features, and taking the candidate sub-features with the sub-feature association degrees larger than a preset sub-feature association degree threshold value as reference sub-features; wherein the sub-feature relevance is a probability that the corresponding sub-feature is included in each candidate sub-feature of each reference recommended route.

In this embodiment of the present application, first, when step 22 is executed, it is required to obtain a sub-feature association degree corresponding to each candidate sub-feature for each candidate sub-feature, specifically, taking any one candidate sub-feature (hereinafter, referred to as a candidate sub-feature n) as an example, a process of obtaining the corresponding sub-feature association degree is described as follows, referring to fig. 2B, a schematic flow chart of obtaining the sub-feature association degree in this embodiment of the present application is shown, and the process of obtaining the sub-feature association degrees corresponding to each candidate sub-feature in this embodiment of the present application is described in detail below with reference to fig. 2B:

step 221: determining a first sample probability according to the number of each reference recommended route and the total number of samples of the sample set; wherein the first sample probability characterizes a probability of occurrence of each reference recommended route in the sample set.

In the embodiment of the application, the ratio between the number of each reference recommended route and the total number of the samples in the sample set is determined, and the determined ratio is used as the first sample probability corresponding to the candidate sub-feature.

Optionally, in this embodiment of the present application, an initial route recommendation model is constructed using a bayesian formula, which may be expressed as:

wherein P (Q) is the occurrence probability of the candidate feature in the sample set, such as the occurrence probability of the resource feature and the occurrence probability of the traffic feature, P (a) is the occurrence probability of all the reference recommended routes in the sample set, P (a | Q) is the occurrence probability of the candidate feature in all the reference recommended routes, and P (Q | a) is the occurrence probability of the candidate feature in the sample set.

Optionally, the first sample probability in the embodiment of the present application may be expressed as:

wherein, P (a)₁) For the first sample probability, x is the number of recommended routes for each reference and y is the number of non-recommended routes.

Step 222: obtaining a second sample probability according to the number of the reference recommended routes and the non-recommended routes containing the candidate sub-features n and the total number of samples; and the second sample probability characterizes the occurrence probability of the candidate sub-feature n in the sample set.

In the embodiment of the present application, a reference recommended route including the candidate sub-feature n and a corresponding number are determined, a non-recommended route including the candidate sub-feature n and a corresponding number are determined, then, a sum of the number of the reference recommended routes including the candidate sub-feature n and the number of the non-recommended routes including the candidate sub-feature n is calculated, a ratio between the obtained sum and a total number of samples is calculated, and the calculated ratio is used as a second sample probability.

Optionally, in the embodiment of the present application,

wherein, P (q)₁) For the probability of occurrence of a flow characteristic, P (q)₂) Is the probability of occurrence of the resource feature, therefore, for the candidate sub-feature n, the second sample probability may be expressed as:

where P (n) is a second sample probability, x is the number of recommended routes for each reference, y is the number of non-recommended routes, x_qFor references containing candidate sub-features nNumber of recommended routes, y_qIs the number of non-recommended routes that contain the candidate sub-feature n.

Step 223: determining a third sample probability according to the number of the reference recommended routes containing the candidate sub-feature n and the number of each reference recommended route; and the third sample probability characterizes the occurrence probability of the candidate sub-feature n in each reference recommended route.

In the embodiment of the present application, a ratio between the number of reference recommended routes including the candidate sub-feature n and the number of each reference recommended route is calculated, and the calculated ratio is used as the third sample probability.

Optionally, in this embodiment of the application, the probability of occurrence of the candidate feature in each reference recommended route is:

wherein, P (q)_i) As the probability of occurrence of the candidate feature, P (q)₁A) is the probability of occurrence of the candidate feature in each reference recommended route, and p (a) is the probability of occurrence of each reference recommended route in the sample set.

Therefore, the third sample probability in the embodiment of the present application may be expressed as:

wherein, P (n | a)₁) Is the third sample probability, x is the number of recommended routes for each reference, x_qThe number of reference recommended routes containing the candidate sub-feature n.

Step 224: and determining the sub-feature association degree corresponding to the candidate sub-feature n according to the first sample probability, the second sample probability and the third sample probability.

In the embodiment of the application, a product between the first sample probability and the third sample probability is calculated to obtain a product result, a ratio between the product result and the second sample probability is calculated, and the calculated ratio is used as a sub-feature association degree corresponding to the candidate sub-feature i.

Optionally, the feature association degree in the embodiment of the present application may be expressed as:

thus, the sub-feature relevance may be expressed as:

wherein, P (a)₁N) is the degree of association of the sub-features, P (n | a)₁) Is the third sample probability, P (a)₁) P (n) is the second sample probability.

Thus, an initial route recommendation model is constructed based on a Bayesian formula, and according to the occurrence probability of each reference recommendation route in a sample set, the occurrence probability of candidate sub-features in the sample set, and the occurrence probability of the candidate sub-features in each reference recommendation route, the sub-feature association degree between each candidate sub-feature and each reference recommendation route can be determined, so that the reference sub-features and the reference sub-features required to be learned by the initial route recommendation model can be determined according to the sub-feature association degree, so that the initial route recommendation model learns the determined reference sub-features, thereby realizing the training process of the route recommendation model, therefore, in the embodiment of the application, the route recommendation model is constructed based on the Bayesian formula, and the target driving route is determined by using the reference sub-features learned by the trained route recommendation model, the route recommendation efficiency can be improved, and the real-time performance of the route recommendation is guaranteed.

The following describes in detail a process of calculating the sub-feature relevance degree corresponding to the candidate sub-feature n in the embodiment of the present application by using a specific example.

Assume that the number of reference recommended routes is 600, wherein the number of reference recommended routes including the candidate sub-feature n is 400, the number of reference recommended routes not including the candidate sub-feature n is 200, and the number of non-recommended routes is 400, wherein the number of non-recommended routes including the candidate sub-feature n is 50, and the number of non-recommended routes not including the candidate sub-feature n is 350.

The first sample probability is:

the second sample probability is:

the third sample probability is:

the sub-feature association degree is as follows:

then, after obtaining the sub-feature association degree corresponding to each candidate sub-feature, sequentially reading the sub-feature association degree corresponding to each candidate sub-feature, judging whether the sub-feature association degree is greater than a preset sub-feature association degree threshold value or not every time one candidate sub-feature association degree is read so as to obtain a judgment result of the candidate sub-feature, obtaining the judgment result of each candidate sub-feature after the reading of each candidate sub-feature association degree is completed, selecting the candidate sub-feature of which the judgment result is that the sub-feature association degree is greater than the preset sub-feature association degree threshold value from each candidate sub-feature, and taking the selected candidate sub-feature as a reference sub-feature.

For example, if the preset sub-feature relevance degree threshold is 0.89, the candidate sub-feature with the sub-feature relevance degree greater than 0.89 is used as the reference sub-feature.

Step 23: and taking the flow sub-features in the reference sub-features as reference flow sub-features learned by the route recommendation model, and taking the resource sub-features in the reference sub-features as reference resource sub-features learned by the route recommendation model.

In the embodiment of the application, the category corresponding to each reference sub-feature is determined, the flow sub-features in the various reference sub-features are used as the reference flow sub-features learned by the route recommendation model, and the resource sub-features in the various reference sub-features are used as the reference resource sub-features learned by the route recommendation model.

In the embodiment of the application, the initial route recommendation model is trained through the method, the trained route recommendation model is obtained, the route recommendation model is built based on the Bayesian formula, and the efficiency of the route recommendation model training can be improved. During subsequent specific implementation, the flow characteristic association degree and the resource characteristic association degree between at least two candidate driving routes and each reference recommended route can be respectively determined according to the trained route recommendation model, and then a corresponding target driving route is determined from the at least two candidate driving routes.

Referring to fig. 3A, a schematic diagram of a process for determining a target driving route according to an embodiment of the present application is shown, and the following description is made in detail with reference to fig. 3A:

step 30: and responding to a destination selection operation triggered by the target object, and obtaining at least two candidate driving routes between the target object and the destination.

In the embodiment of the application, a target object triggers and generates a destination selection operation in a terminal device, so that the terminal device sends the destination selection operation triggered by the target object to a server, and the server responds to the destination selection operation when acquiring the destination selection operation, determines the current position of the target object, and acquires at least two candidate driving routes from the current position of the target object to the destination.

Wherein the at least two candidate driving routes are driving routes from the current position of the target object to the destination.

It should be noted that there are multiple candidate driving routes between the current position of the target object and the destination, and in order to reduce the calculation amount and improve the route recommendation efficiency, at least two candidate driving routes may be selected from the multiple candidate driving routes, and the target driving route may be determined from the selected at least two candidate driving routes.

For example, at least two candidate travel routes are selected from the plurality of candidate travel routes according to travel distances of the candidate travel routes.

Step 31: obtaining at least one candidate place with specified resources corresponding to at least two candidate driving routes respectively; wherein each candidate is located within a peripheral preset range of the destination.

In the embodiment of the application, because the driving sections passed by each candidate driving route are different, the candidate places which can be reached by different candidate driving routes are also different, and at least one candidate place with the specified resource corresponding to at least two candidate driving routes is obtained.

When the intelligent parking system is applied to a scene of intelligent parking, the designated resource is a parking space, the candidate place is a parking lot, and the target object is an owner of the vehicle.

For example, referring to fig. 3B, a schematic diagram of a candidate driving route in the embodiment of the present application is shown, and a map includes a parking lot a, a parking lot B, a parking lot C, a parking lot D, a parking lot E, a candidate ground F, and a candidate ground G. After the target object triggers and generates a destination selection operation, responding to the destination selection operation triggered by the target object, determining the current position of the target object, and obtaining two candidate driving routes from the current position of the target object to the destination, namely a candidate driving route 1 and a candidate driving route 2.

As shown in fig. 3B, the parking lots that the candidate travel route 1 can reach are: E. f and a, parking lots that the candidate travel route 2 can reach are: E. a and D. Since the parking lots within the preset range around the destination are A, B, C and D, respectively, the parking lot a corresponding to the candidate travel route 1 is taken as the candidate place corresponding to the candidate travel route 1, and the parking lots a and D corresponding to the candidate travel route 2 are taken as the candidate places corresponding to the candidate travel route 2.

Step 32: respectively obtaining flow characteristic association degrees and resource characteristic association degrees between at least two candidate driving routes and each reference recommended route by adopting a trained route recommendation model; and each reference recommended route is a route which meets the recommended condition and is determined according to the historical flow and the historical resource quantity.

In the embodiment of the present application, when step 32 is executed, it is required to obtain a traffic characteristic association degree and a resource characteristic association degree between at least two candidate driving routes and each reference recommended route, specifically, taking any one candidate driving route (hereinafter, referred to as a candidate driving route i) as an example, a process of obtaining the traffic characteristic association degree and the resource characteristic association degree is described as follows:

and inputting the candidate driving route i into the trained route recommendation model to obtain the traffic characteristic association degree between the candidate driving route i and each reference recommended route and the resource characteristic association degree between the candidate driving route i and each reference recommended route.

It should be noted that the determined resource feature association degree between the candidate driving route i and each reference recommended route, that is, the candidate driving route i and one reference recommended route correspond to one traffic feature association degree and one resource feature association degree.

In addition, it should be noted that, because each reference recommended route is a pre-labeled route satisfying the recommendation condition, if the traffic characteristic association degree and the resource characteristic association degree determined by the candidate driving route i and each reference recommended route are higher, the probability that the candidate driving route is the target recommended route is higher, in this embodiment of the application, the traffic characteristic association degree between the candidate driving route i and each reference recommended route is respectively obtained, and the resource characteristic association degree between the candidate driving route i and each reference recommended route is respectively obtained.

That is, a traffic characteristic association degree can be determined between the candidate driving route i and a reference recommended route, and a resource characteristic association degree can be determined between the candidate driving route i and a reference recommended route.

The specific calculation method of the traffic characteristic relevance degree and the resource characteristic relevance degree will be explained in detail in the following embodiments.

Step 33: and selecting the candidate driving route of which the corresponding flow characteristic association degree and the corresponding resource characteristic association degree both meet the preset association degree condition from the at least two candidate driving routes as a target driving route.

In the embodiment of the application, according to the flow characteristic association degrees and the resource characteristic association degrees corresponding to the at least two candidate driving routes, the candidate driving routes, of which the corresponding flow characteristic association degrees and the corresponding resource characteristic association degrees both meet the preset association degree condition, are selected from the at least two candidate driving routes, and the selected candidate driving routes are used as the target driving routes which are finally recommended to the target object.

The manner in which the specific target travel route is determined will be described in detail in the following embodiments.

The association degree condition may be a preset association degree threshold, for example, the association degree threshold may be set to 0.85, which is not limited in this embodiment of the application.

In the embodiment of the application, because the flow of the candidate driving route influences the time of the target object reaching the destination, the flow characteristic and the resource characteristic of the candidate driving route are combined, the traffic jam is predicted by using the flow characteristic, the route optimization is performed, the accuracy of the route recommendation is improved, and the most convenient candidate place is selected for the target object by using the resource characteristic.

Referring to fig. 4A, a schematic flow chart illustrating obtaining of the traffic characteristic association degree and the resource characteristic association degree of at least two candidate driving routes in the embodiment of the present application is shown, and a description is given below of a process of obtaining the traffic characteristic association degree and the resource characteristic association degree between at least two candidate driving routes and each reference recommended route respectively by using a trained route recommendation model in combination with fig. 4A.

Step 40: and respectively obtaining the flow characteristics and the resource characteristics corresponding to the at least two candidate driving routes.

In the embodiment of the application, the driving flow data of at least two candidate driving routes are respectively subjected to feature extraction to obtain respective corresponding flow features, and the resource data of at least two candidate driving routes are respectively subjected to feature extraction to obtain respective corresponding resource features.

Referring to fig. 4B, a detailed description is given below to a process of obtaining flow characteristics corresponding to at least two candidate driving routes respectively, and referring to fig. 4B, a schematic flow chart of obtaining flow characteristics in the embodiment of the present application is shown, specifically, a process of obtaining flow characteristics of a candidate driving route is described as follows, taking the candidate driving route i as an example:

step 401: and respectively obtaining each driving road section contained in the candidate driving route i, the historical flow corresponding to each historical sampling time period, and the current flow corresponding to each driving road section in the current sampling time period.

In the embodiment of the present application, because there are many factors that affect the flow rate corresponding to each traveling road segment, when performing feature extraction on the flow rate of the candidate traveling route i, it is necessary to determine each traveling road segment included in the candidate traveling route i first. Fig. 4C is a schematic diagram illustrating a correspondence relationship between a candidate driving route and a driving road segment according to an embodiment of the present application.

As in the candidate travel route-travel-link correspondence table in fig. 4C, the correspondence between the plurality of candidate travel routes and the respective travel links included therein are stored in the candidate travel route-travel-link correspondence table, and the travel links included in the candidate travel route 1 are A, F, D, E respectively, and the travel links included in the candidate travel route 2 are B, C, D, E respectively.

It should be noted that, in the embodiment of the present application, each candidate driving route may have a partial overlap, and therefore, the driving routes included in each candidate driving route may have a partial identity. Referring to fig. 4D, which is a schematic diagram of the driving links included in the candidate driving route according to the embodiment of the present application, the driving link included in the candidate driving route 1 is A, F, D, E, the driving link included in the candidate driving route 2 is B, C, D, E, and both the candidate driving route 1 and the candidate driving route 2 include the driving links D and E.

Then, historical flow rates of each candidate driving route corresponding to each historical sampling time point are obtained, each driving section of each candidate driving route corresponds to one flow rate data table, and the following description will be made on the flow rate data table in the embodiment of the present application by taking the driving section a of the candidate driving route 1 as an example, and refer to fig. 4E, which is a schematic diagram of a historical flow rate correspondence relationship in the embodiment of the present application.

As shown in the flow rate data table in fig. 4E, the historical flow rates of the travel links a of the candidate travel route 1 at the historical sampling time points are stored, and the number of the flow rate data tables of each candidate travel route is related to the number of the travel links included in the candidate travel route, that is, one flow rate data table is corresponding to each travel link.

When the historical flow rate is stored, a flow rate data table for each travel route on each natural day is also obtained. For example, a travel link a of day 9.2 corresponds to one flow rate data table, and a travel link a of day 9.1 corresponds to one flow rate data table.

Optionally, in this embodiment of the application, historical flow rates of the travel road segments included in each candidate travel route at each historical sampling time point may also be stored in the same flow rate data table, which is not limited herein.

For example, when the method is applied to an intelligent parking scene, a traffic camera and other collection devices are used for identifying and acquiring 24-hour historical flow of each driving road section of urban road traffic, flow sequence information can be generated according to the historical flow corresponding to each driving road section of each candidate driving route at each historical sampling time point after the historical flow corresponding to each driving road section included in each candidate driving route at each historical sampling time point is acquired, wherein each historical sampling time point corresponds to one flow sequence information, and further when the historical flow of each candidate driving route at each historical sampling point is determined, the historical flow corresponding to each historical sampling time point of each driving road section can be directly acquired from each flow sequence information, so that the operation of acquiring the historical flow is simplified.

Optionally, when the historical traffic corresponding to each historical sampling time point of each running road section of each candidate running route is stored, the historical traffic corresponding to each historical sampling time point of each candidate running route may be stored through the traffic sequence information mapping table, so that, in specific implementation, the historical traffic corresponding to each historical sampling time point of each candidate running route may be obtained from the traffic sequence information mapping table according to the route identifier and the historical sampling time point corresponding to the candidate running route. A detailed description is given below to the traffic sequence information mapping table in this embodiment by using a specific example, and reference is made to fig. 4F, which is a schematic diagram of a traffic sequence information mapping relationship in this embodiment.

For example, the flow rate sequence information mapping table in fig. 4F stores route identifiers corresponding to the candidate driving routes, historical flow rates corresponding to the historical sampling time points and the driving road sections included in the candidate driving routes, and flow rate sequence information corresponding to the candidate driving routes at the historical sampling time points.

The flow rate sequence information is composed of historical flow rates corresponding to all driving road sections of the candidate driving route. For example, if the historical flow rate of the travel section a corresponding to the candidate travel route 1 at 23 is 40, the historical flow rate of the travel section B is 426, the historical flow rate of the travel section C is 263, the historical flow rate of the travel section D is 126, the historical flow rate of the travel section E is 216, and the historical flow rate of the travel section F is 201, the flow rate sequence information corresponding to the candidate travel route 1 at 23 is 40-426 plus 263 plus 216 plus 201.

In this way, in specific implementation, the time point when the destination selection operation is acquired and the route identifier corresponding to the generated candidate driving route are determined, and the flow sequence information corresponding to the generated candidate driving route is acquired from the flow sequence information mapping table.

For example, assuming that a candidate travel route i is taken as an example, a destination selection operation is obtained at 8 points on 2 th and 9 th months, so as to generate a candidate travel route i from a current position to a destination, and the time of arrival at the destination is predicted to be 9 points in the morning, flow sequence information of the candidate travel route i at 8 points on 1 st and 9 th months, flow sequence information of 9 points on 1 st and 9 th days on 9 th months, flow sequence information of 8 points on 31 th and 31 th days on 8 th and 9 th days on 8 th months are obtained from a flow sequence information mapping table as historical flows corresponding to historical sampling time points, so that flow change information in each historical sampling period can be determined according to the historical flows corresponding to the historical sampling time points.

Step 402: and respectively determining the flow change information of each driving road section according to each historical flow corresponding to each driving road section.

In the embodiment of the application, the historical flow corresponding to each running road section at the historical sampling time point can be obtained from the flow sequence information mapping table, so that the flow change information of each running road section in each historical sampling time period can be determined according to the obtained historical flow corresponding to each historical sampling time point.

Wherein, the flow change information is the change of the flow of the road section which runs in the historical sampling period.

For example, if the travel link a of the candidate travel route i corresponds to the historical flow rate of 90 at 8 points on day 1/9/month, and the travel link a of the candidate travel route i corresponds to the historical flow rate of 88 at 8 points on day 1/9/month, the flow rate change information from 8 points to 9 points on day 1/9 of month is determined to be-2.

In addition, if the historical flow corresponding to the initial historical sampling time point in the historical sampling period is smaller than the historical flow corresponding to the end historical sampling time point, it is determined that the value of the acquired flow change information is a positive number, the value of the flow change information is represented by the positive number, and the flow is increased in the historical sampling period. For example, if the historical flow rate of the travel link a at 9 points is 70 and the historical flow rate at 10 points is 90, the flow rate change information of the travel link a at 9 to 10 points is 20, and the value is a positive number.

And if the historical flow corresponding to the initial historical sampling time point in the historical sampling time period is greater than the historical flow corresponding to the ending historical sampling time point, determining that the numerical value of the obtained flow change information is a negative number, representing that the numerical value of the flow change information is a negative number, and reducing the flow in the historical sampling time period. For example, if the historical flow rate of the travel link B at 10 points is 90 and the historical flow rate at 11 points is 80, the flow rate change information of the travel link B at 10-11 points is-10, and the value is a negative number.

It should be noted that the historical sampling periods in the embodiment of the present application may be different periods of time in the same natural day, for example, each historical sampling period is 9-10 points on 5.30 days and 10-11 points on 5.30 days, which is not limited in this regard.

Of course, the historical sampling period may also be the same time period in different natural days.

Step 403: and determining the flow characteristic of the candidate driving route i according to the obtained current flow and flow change information corresponding to each driving road section and the object density characteristic of the candidate driving route i.

In the embodiment of the application, the current flow corresponding to each driving road section contained in the candidate driving route i and each influence factor influencing the object density of the candidate driving route i are obtained, each influence factor is coded according to a preset coding mode, the object density characteristic of the candidate driving route i is obtained, and then the flow characteristic of the candidate driving route i is determined according to the current flow, the flow change information and the object density characteristic which respectively correspond to each driving road section.

The influence factor is a factor that influences the flow rate corresponding to the candidate travel route. For example, the influence factor is whether the peak hour is on duty, whether the peak hour is on holiday, whether a large business district exists nearby, whether a visiting scenic spot exists nearby, whether a subway bus exists nearby, whether a school hospital exists nearby, and the like, and this is not limited in the embodiment of the present application.

For example, when the influence factor is whether a large business turn exists, if a large business turn exists near the destination of the candidate driving route i, the value of the object density characteristic corresponding to the influence factor is set to 1, otherwise, the value of the object density characteristic corresponding to the influence factor is set to 0.

Optionally, the encoding manner in the embodiment of the present application may be, for example, one-hot encoding, but is not limited thereto.

For example, assume that the flow characteristic corresponding to the candidate driving route i can be expressed as: [15,243,241,313,84,474, -5,42,51,51, -121,24,0,1,0,1 … ].

Then [15,243,241,313,84,474] is the current flow rate for each travel segment of the candidate travel route i, and [ -5,42,51,51, -121,24] is the flow rate change information corresponding to 8-9 points on 5.30 days for each travel segment of the candidate travel route i.

It should be noted that, because the flow rate in each historical sampling period in the cycle is dynamically changed, and the object density characteristics corresponding to each historical sampling period are different, the flow rate characteristics corresponding to the same candidate driving route in each historical sampling period are also different. For example, the flow characteristics over 8:00-9:00 on day 2 of 9 month are different from the flow characteristics over 8:00-9:00 on day 3 of 9 month, and the flow characteristics over 8:00-9:00 on day 2 of 9 month are different from the flow characteristics over 9:00-10:00 on day 2 of 9 month.

The flow rate feature includes a plurality of flow rate sub-features, for example, four flow rate sub-features are included in the flow rate feature [5,42,51,51, -121,24,0,1], where [5,42,51] is a flow rate sub-feature corresponding to the current flow rate of the candidate travel route i, [51, -121,24] is a flow rate sub-feature corresponding to the flow rate change information of the candidate travel route i within 5 months and 30 days 9:00-10:00, [0] is a flow rate sub-feature corresponding to a large supermarket, and [1] is a flow rate sub-feature corresponding to a peak period of up-and-down shifts.

In the implementation of the application, the flow characteristic is determined according to the flow change information, the current flow and the object density characteristic corresponding to the candidate driving route, so that the object density characteristic capable of influencing the object density is added in the flow characteristic, when the flow characteristic is used for carrying out route recommendation, the flow of the candidate driving route obtained through prediction can better accord with the actual flow, the time for predicting the candidate driving route to the destination is more accurate, and the accuracy of the route recommendation is improved.

The process of obtaining the resource features corresponding to at least two candidate driving routes respectively will be described in detail below with reference to fig. 4G, which is a schematic flow chart of obtaining the resource features in the embodiment of the present application and is shown in fig. 4G while obtaining the resource features corresponding to the candidate driving routes.

In the embodiment of the present application, when obtaining the candidate driving routes, feature extraction operations need to be performed on at least two candidate driving routes, specifically, taking the candidate driving route i as an example, a process of obtaining resource features of the candidate driving routes is described as follows:

step 404: and respectively obtaining resource change information corresponding to at least one candidate of the candidate driving route i in each historical sampling period.

In the embodiment of the present application, first, candidate places to which the candidate travel routes can reach are determined.

For example, when the candidate place is a parking lot and the designated resource is a parking lot, the parking lots through which the candidate driving routes can pass are obtained, and an association relation table between the candidate driving routes and the parking lots is established, which is shown in fig. 4H and is a schematic diagram illustrating an association relation between the candidate driving routes and the parking lots in the embodiment of the present application.

As shown in the candidate travel route/parking lot association table in fig. 4H, the parking lots through which the candidate travel route 1 can pass are stored as the parking lot a, the parking lot B, the parking lot C, the parking lot D, the parking lot E, and the parking lot F, and therefore, in concrete implementation, each parking lot that can be reached by at least two candidate travel routes can be obtained from the candidate travel route/parking lot association table.

Then, resource data of the parking lot at each historical sampling time point are obtained through the identification device, and resource change information corresponding to at least one candidate place of the candidate driving route i in each historical sampling time period is obtained according to the resource data at each historical sampling time point.

For example, when the candidate place is a parking lot and the designated resource is a parking lot, the traffic camera acquires resource data corresponding to different historical sampling time points of the parking lots where the candidate driving route 1 can reach near the destination. Fig. 4I is a schematic view illustrating a corresponding relationship between vacant parking spaces in a parking lot in the embodiment of the present application.

The parking lot vacant parking space corresponding relation table in fig. 4I stores resource data of the parking lot a of the candidate driving route 1 at each historical sampling time point, and each candidate driving route corresponds to at least one parking lot, so that at least one parking lot of each candidate driving route corresponds to the parking lot vacant parking space data table shown in fig. 4I, and resource data of at least one parking lot of the candidate driving route corresponding to each historical sampling time point can be obtained according to the parking lot vacant parking space data table.

Then, after obtaining at least one resource data corresponding to each historical sampling time point of at least one candidate of the candidate driving route i, generating a resource data comparison table of the candidate driving route i and at least one candidate, wherein the resource data comparison table records a route identifier corresponding to the candidate driving route, the historical sampling time point, resource data corresponding to at least one candidate of the candidate driving route, and resource sequence information generated according to the resource data corresponding to at least one candidate, so that in the subsequent concrete implementation, the resource data corresponding to at least one candidate of the candidate driving route can be directly searched from the resource data comparison table.

For example, when the candidate place is a parking lot, the specified resource is a parking space, and the resource data is the number of vacant parking spaces, the resource data table of the candidate travel route 1 is generated. Fig. 4J is a schematic diagram illustrating a resource data correspondence relationship of the candidate driving route 1 according to the embodiment of the present application.

As shown in fig. 4J, the resource data correspondence table stores route identifiers corresponding to the candidate driving routes, resource data corresponding to at least one candidate of the historical sampling time points and the candidate driving routes, and resource sequence information corresponding to the candidate driving routes at the historical sampling time points.

Wherein the resource sequence information is composed of resource data corresponding to at least one candidate of the candidate driving routes. For example, when the candidate driving route 1 is 23, the resource data of the parking lot a is 40, the resource data of the parking lot B is 426, the resource data of the parking lot C is 263, the resource data of the parking lot D is 126, the resource data of the parking lot E is 216, and the resource data of the parking lot F is 201, the corresponding resource sequence information of the candidate driving route 1 is 40-426 and 263 and 216 and 201 respectively.

In this way, in specific implementation, since each candidate driving route generates the resource data table as shown in fig. 4J in different historical sampling periods, the resource sequence information corresponding to the candidate driving route can be obtained from the resource data table according to the time point of obtaining the destination selection operation and the route identifier corresponding to the generated candidate driving route.

Finally, the resource change information corresponding to at least one candidate of the candidate driving route i in each historical sampling time period can be determined according to the resource data corresponding to the candidate driving route i in each historical sampling time point.

In specific implementation, a sampling time interval is preset, each historical sampling time period is obtained according to the sampling time interval, then, the difference value between the resource data of the ending historical sampling time point and the resource data of the starting historical sampling time point in each historical sampling time period is calculated, and the resource change information in the corresponding historical sampling time period is obtained.

For example, taking the candidate place as a parking lot and the resource data as the number of vacant parking spaces as an example, the resource change information is obtained based on the number of vacant parking spaces in the parking lot near the destination in each historical sampling period, and as shown in fig. 4K, the change diagram of the number of vacant parking spaces in the embodiment of the present application is shown.

As can be seen from fig. 4K, the free space amount change table stores resource change information corresponding to each candidate of the historical sampling time period and the candidate travel route.

In addition, it should be noted that, if the resource data corresponding to the start historical sampling time point in the historical sampling time period is smaller than the resource data corresponding to the end historical sampling time point, it is determined that the value of the acquired resource change information is a positive number, and the value of the resource change information is a positive number representation, and the resource data is increased in the historical sampling time period. For example, if the resource data of the parking lot a at point 1 is 10 and the resource data at point 2 is 21, the resource change information of the parking lot a at point 1-2 is 11, and the value is a positive number.

And if the resource data corresponding to the initial historical sampling time point in the historical sampling period is smaller than the resource data corresponding to the ending historical sampling time point, determining that the numerical value of the acquired resource change information is a negative number, representing that the numerical value of the resource change information is a negative number, and reducing the resource data in the historical sampling period. For example, if the resource data of the parking lot a at point 0 is 10 and the resource data at point 1 is 5, the resource change information of the parking lot a at point 0-1 is-5, and the value is a negative number.

Then, after obtaining the resource change information of at least one candidate place in each history sampling period, the value is represented by a symbol, the resource change information whose value is a positive number is represented as 1, and the resource change information whose value is a negative number is represented as-1, for example, as shown in fig. 4L, which is a schematic diagram of increasing and decreasing resource data in the embodiment of the present application.

As shown in the resource data increase and decrease table of FIG. 4L, 1 indicates that the number of vacant parking spaces increases, and-1 indicates that parking is a reduction in the number of vacant parking spaces.

Then, the "increase" flag and the "decrease" flag are used to indicate resource change information of each candidate place, for example, as shown in fig. 4M, which is a schematic diagram of increasing and decreasing trend of vacant parking spaces in the embodiment of the present application.

Step 405: and determining the resource characteristics corresponding to the candidate driving route i according to a preset characteristic extraction mode and based on the obtained resource change information.

In the embodiment of the application, the resource change information of each candidate place is determined firstly, and the resource characteristic corresponding to the candidate driving route i is determined according to the resource change information.

Optionally, a possible implementation manner is provided in the embodiment of the present application for obtaining the resource feature of the candidate driving route i, and a process of determining the resource feature in the embodiment of the present application is described in detail below with reference to the accompanying drawings, and as shown in fig. 4N, a schematic view of the process of determining the resource feature in the embodiment of the present application specifically includes:

step 4051: and for each historical sampling period, generating a corresponding information sequence based on the resource change information of at least one candidate corresponding to the candidate driving route i in the corresponding historical sampling period.

In the embodiment of the present application, when step 4051 is executed, it is necessary to obtain an information sequence corresponding to each historical sampling period for each historical sampling period, specifically, taking any one historical sampling period (hereinafter referred to as historical sampling period x) as an example, a process of obtaining the information sequence is described as follows:

and generating an information sequence in the historical sampling period x based on the resource change information in the historical sampling period x of at least one candidate corresponding to the candidate driving route i, wherein the information sequence comprises the resource change information corresponding to the at least one candidate.

Each historical sampling period corresponds to an information sequence, and a specific example is used below to describe the information sequence in the embodiment of the present application, and refer to fig. 4O, which is a schematic diagram of the information sequence in the embodiment of the present application.

The information sequence table in fig. 4O includes route identifiers corresponding to the candidate driving routes, historical sampling periods, and information sequences in each historical sampling period.

Optionally, in this embodiment of the application, information sequences of each candidate driving route in each preset sampling time period may be further stored in advance, so that when the information sequences in each historical sampling time period are generated, the corresponding information sequences may be directly obtained from the stored information sequences according to route identifiers corresponding to the candidate driving routes and the historical sampling time periods.

For example, an information sequence table may be generated in advance for each candidate driving route, and refer to fig. 4P, which is a schematic diagram of an information sequence of the candidate driving route 1 in the embodiment of the present application.

In the information sequence table of the candidate travel route as in fig. 4P, the information sequence of the candidate travel route 1 in each history sampling period of 9.2 days is stored, and each candidate travel section corresponds to the information sequence table of the candidate travel route as in fig. 4P on different natural days.

Step 4052: obtaining at least one candidate information set corresponding to a candidate on the basis of each obtained information sequence; and each candidate information is the resource change information of which the occurrence frequency meets the preset frequency condition in each resource change information corresponding to the corresponding candidate.

In the embodiment of the application, based on the obtained information sequences, resource change information with an occurrence frequency meeting a preset frequency condition is determined from at least one resource change information corresponding to a candidate place, and a candidate information set is generated according to the determined resource change information, that is, the candidate information set contains the resource change information with an occurrence frequency meeting the preset frequency condition.

For example, assuming that the candidate travel route is the candidate travel route 1, the time of receiving the destination selection operation is 12.3 days 8, and the time of predicting the destination is 12.3 days 9, when the historical sampling period is 12.1 days 8-9 and 12.2 days 8-9 according to the time of receiving the destination selection operation and the time of predicting the destination, reference is made to fig. 4Q, which is a schematic view of increasing and decreasing the parking space of the candidate travel route 1 in the embodiment of the present application.

Then, the frequency of occurrence of each resource change information in each information sequence is calculated.

For example, taking the above-described candidate travel route 1 as an example, the frequency of occurrence of each resource change information is calculated. Fig. 4R is a schematic diagram illustrating a corresponding relationship between resource change information and occurrence frequency in the embodiment of the present application.

The frequency of occurrence of the resource change information a plus is 1, the frequency of occurrence of the resource change information a minus is 1, the frequency of occurrence of the resource change information B plus is 2, the frequency of occurrence of the resource change information C plus is 2, the frequency of occurrence of the resource change information D minus is 2, the frequency of occurrence of the resource change information E plus is 1, the frequency of occurrence of the resource change information E minus is 1, and the frequency of occurrence of the resource change information F minus is 2.

The occurrence frequency in the embodiment of the present application is the minimum support degree, and optionally, the occurrence frequency may be represented as:

wherein n is the number of days in the time range of collecting the resource data, a is the minimum support rate, and the minimum support rate parameter is adjusted according to the number of the sample sets.

Then, whether the occurrence frequency of each resource change information is not less than a preset frequency threshold is judged.

For example, taking the above-described candidate travel route 1 as an example, when the frequency threshold is 2, the frequency of occurrence of the resource change information B increase, C increase, D decrease, and F decrease is not less than the preset frequency threshold 2.

And finally, generating a candidate information set according to the resource change information which is not less than the preset frequency threshold value and corresponds to.

In addition, in each historical sampling period, each resource change information corresponding to the same time period corresponding to different natural days can be extracted to obtain one target information sequence feature, so that each different time period corresponding to the same natural day can determine the corresponding target information sequence feature as a resource sub-feature, and further generate a resource feature composed of each resource sub-feature, for example, the target information sequence feature corresponding to 8-9 points is one resource sub-feature, and the target information sequence feature corresponding to 9-10 points is one resource sub-feature.

Step 4053: sequentially reading each candidate information contained in the candidate information set, and executing the following operations every time one candidate information is read: taking each information sequence containing one currently read candidate information as a candidate information sequence, and respectively generating corresponding information sequence characteristics in each candidate information sequence based on one candidate information and other resource change information; and the other resource change information is the resource change information arranged behind one candidate information in each candidate information sequence according to a preset sequence.

In this embodiment of the application, when step 4053 is executed, it is necessary to sequentially read each candidate information included in the candidate information set, and generate an information sequence feature corresponding to each candidate information, specifically, taking any one candidate information z as an example, a process of introducing the information sequence feature of the generated candidate information z is as follows:

and taking each information sequence containing the currently read candidate information z as a candidate information sequence, and respectively generating corresponding information sequence characteristics in each candidate information sequence based on one candidate information and other resource change information.

According to the method and the device, the corresponding information sequence characteristics are generated according to the candidate information and other resource change information, the information sequence characteristics which appear most frequently in the candidate driving routes can be determined, the determined information sequence characteristics can be used for representing the resource characteristics of the candidate driving routes, and then when route recommendation is carried out according to the resource characteristics, the influence of noise characteristics can be reduced, and therefore the accuracy of the route recommendation is improved.

Optionally, a possible implementation manner is provided in the embodiment of the present application for a process of generating an information sequence feature corresponding to candidate information, and the process of generating an information sequence feature in the embodiment of the present application is described in detail by taking the candidate information z as an example, and refer to fig. 4S, which is a schematic flow diagram for generating an information sequence in the embodiment of the present application.

Step 4053-1: and taking the currently read candidate information z as an initial information sequence characteristic.

In the embodiment of the application, the currently read candidate information z is used as the extracted initial information sequence feature.

For example, when the candidate information z is a, a is added as the initial information sequence feature.

Step 4053-2: and circularly executing the step of determining the characteristics of the new initial information sequence on the basis of each candidate information sequence obtained currently until the occurrence frequency of each piece of other resource change information is determined to not meet the preset frequency condition, and obtaining each new initial information sequence characteristic.

The following describes in detail the process of determining new initial information sequence features in the embodiment of the present application by taking the first loop as an example, and refer to fig. 4T, which is a schematic flow chart of determining an initial information sequence in the embodiment of the present application.

Step 4053-2-1: and respectively adding other resource change information of which the occurrence frequency meets the frequency condition in the other resource change information contained in each candidate information sequence to the corresponding initial information sequence characteristics to obtain each updated information sequence characteristic.

In the embodiment of the application, the change information of each other resource contained in each candidate information sequence and the occurrence frequency of each change information of each other resource in each candidate information sequence are determined, then, according to each determined occurrence frequency, the change information of other resources with the occurrence frequency larger than a preset frequency threshold is determined, and the determined change information of other resources is added to the initial information sequence feature to generate an updated information sequence feature containing the candidate information and the change information of other resources.

It should be noted that, since there may be a plurality of other determined resource change information that is greater than the preset frequency threshold, each of the determined other resource change information needs to be added to the initial information sequence feature.

For example, assuming that the candidate information is increase B, each candidate information sequence including increase candidate information B is increase B-increase C-increase D-decrease E-increase F-decrease, and increase B-increase C-increase D-decrease E-decrease F-decrease, respectively, determining the occurrence frequency of various other resource change information in each candidate information sequence, the occurrence frequency of increase C is 2, the occurrence frequency of decrease D is 2, the occurrence frequency of increase E is 1, the occurrence frequency of decrease F is 2, determining that the occurrence frequency of increase C, decrease D and decrease F is greater than a preset frequency threshold, combining increase B with increase C, generating updated information sequence feature increase B, increasing C, combining increase B with decrease D, generating updated information sequence feature increase D, decreasing B, and decreasing F, generating updated information sequence feature increase B, therefore, the generated updated information sequence features are respectively: b increases C, B increases D and B increases F.

Step 4053-2-2: and respectively screening out each candidate information sequence containing each updated information sequence characteristic from each candidate information sequence to be used as the corresponding updated candidate information sequence.

In the embodiment of the present application, since some candidate information sequences may not include updated information sequence features, for each updated information sequence feature, an updated candidate information sequence corresponding to each updated sequence feature is obtained, and specifically, a process of obtaining a corresponding updated candidate information sequence in the embodiment of the present application is described by taking any one updated information sequence feature (hereinafter, referred to as an updated information sequence feature l) as an example.

And screening each candidate information sequence corresponding to the updated information sequence characteristic l from each candidate information sequence, and taking each screened candidate information sequence as an updated candidate information sequence corresponding to the updated information sequence characteristic a.

For example, assuming that the updated information sequence characteristic l is B plus C plus, determining that the candidate information sequences containing B plus C plus are B plus-C plus-D minus E plus-F minus, and B plus-C plus-D minus-E minus-F minus, and then B plus-C plus-D minus, E plus-F minus, and B plus-C plus-D minus, E minus-F minus, are taken as the updated candidate information sequences.

Based on the above manner, each updated candidate information sequence corresponding to each updated information sequence feature can be obtained.

Step 4053-2-3: and aiming at each updated information sequence characteristic, respectively executing the following operations: and respectively taking the resource change information which is arranged in each updated candidate information sequence after the updated information sequence is characterized according to a preset sequence as the other updated resource change information.

In the embodiment of the present application, the process of obtaining updated other resource change information corresponding to each updated sequence feature in the updated candidate information sequence corresponding to each updated sequence feature is specifically described, taking any one updated information sequence feature (hereinafter, referred to as an updated information sequence feature b) as an example:

and respectively taking the resource change information arranged behind the updated information sequence characteristic b in each updated candidate information sequence according to a preset sequence as the updated other resource change information.

For example, taking the updated information sequence characteristic B as B increase-C increase as an example, the updated candidate information sequences are B increase-C increase-D decrease-E increase-F decrease, and B increase-C increase-D decrease-E decrease-F decrease, and each other resource change information arranged after B increase-C increase is taken as updated resource change information, that is, D decrease-E increase-F decrease, and D decrease-E decrease-F decrease are taken as updated resource change information.

Step 4053-2-4: and taking each updated information sequence characteristic as each new initial information sequence characteristic.

In the embodiment of the present application, each updated information sequence feature is used as a new initial information sequence feature, and the loop process of steps 4053-2-1 to 4053-2-4 is executed again.

In the embodiment of the application, the resource characteristics are determined by adopting the idea of sequence mode mining, so that the influence of noise characteristics can be reduced, and the accuracy of route recommendation is improved.

Step 4054: and selecting each target information sequence characteristic which meets the preset number condition from each information sequence characteristic as a resource characteristic corresponding to the candidate driving route i based on the total number of the resource change information associated with each information sequence characteristic.

In the embodiment of the application, the total number of the resource change information in each information sequence feature is determined, each target information sequence feature meeting the preset number condition is selected, each target information sequence feature is a resource sub-feature, and then the resource feature corresponding to the candidate driving route i is generated according to each resource sub-feature.

For example, assuming that the total number of resource change information in the information sequence feature 1 is 4 and the total number of resource change information in the information sequence feature 2 is 3, the information sequence feature 1 is set as the target information sequence feature.

Then, feature coding is performed on each resource sub-feature to construct a resource feature, for example, one-hot coding is performed on the resource change information, and then the element codes of each resource change information are spliced according to a time sequence to serve as the resource feature. For example, B is encoded as [1,0,0, … ], C is encoded as [0, 1,0, … ], D is encoded as [0,0, 1, … ], and the like, and the resource sequence information codes are concatenated in sequence order to construct the resource feature codes [1,0,0, … 0,1,0, … ].

It should be noted that, when the historical sampling time period is different time periods of different natural days, the resource feature in the embodiment of the present application includes a plurality of resource sub-features, for example, when the resource feature is [1,0,0,0,1,0,1,0,0,0,0,1], then [1,0,0,0,1,0] represents the resource sub-feature at 8-9 points in 30 days of 5 months, and increases-C, and [1,0,0,0,0, 0,1] represents the resource sub-feature at 9-10 points in 30 days of 5 months, and increases-D.

In addition, it should be noted that, in the embodiment of the present application, the sequence patterns in each information sequence may be mined based on a frequent sequence pattern mining algorithm, and the change patterns of each resource change information frequently occurring in each historical sampling period within a period of time are mined, so that the frequently occurring change patterns are used as the resource features.

For example, when determining resource features using a frequent sequence pattern mining algorithm, first, candidate information having a unit length of 1 is found as the initial information sequence features and the corresponding projection data set.

Then, counting the occurrence frequency of the initial information sequence features, and acquiring a frequent one-item set sequence mode according to the initial information sequence features with the support degree higher than the occurrence frequency threshold.

Then, recursively mining all initial information sequence features with the length of i and meeting the requirement of minimum support degree: excavating a projection data set of the initial information sequence characteristics, wherein the projection data set comprises various other resource change information, and if the projection data set is an empty set, returning to recursion; counting the occurrence frequency of each item in the corresponding projection data set, combining the change information of each other resource meeting the occurrence frequency with the initial information sequence characteristic to obtain an updated information sequence characteristic, and recursively returning if the change information does not meet the requirement of the occurrence frequency; making i = i +1, and the initial information sequence features are updated information sequence features after single items are combined, and performing recursion respectively; thereby generating resource characteristics including characteristics of each target information sequence.

The following describes in detail the step of obtaining the resource characteristics of the candidate driving route in the embodiment of the present application by using a specific example, and refer to fig. 4U, which is a schematic diagram of a process of mining the resource characteristics in the embodiment of the present application.

Firstly, constructing an information sequence of the parking space resource change information of the same route in the same time period in different periods, wherein the parking space change sequence of the route contains changed road section identifiers, and the A parking space increase is expressed as 'A increase'.

As shown in fig. 4U, the information sequence of the candidate travel route 1 during 9-10 of 12 th day 1 is a increase-B increase-C increase-D decrease-E increase-F decrease, and the information sequence of the candidate travel route 1 during 9-10 of 12 th day 2 is a increase-B increase-C increase-D decrease-E decrease-F decrease.

Then, mining sequence patterns contained in each information sequence of the candidate driving route in the same time period on different natural days, assuming that the set minimum support threshold is 0.5, the frequency threshold is 2, firstly counting the occurrence frequency of each resource change information to obtain a candidate information set, determining that B increase, C increase, D decrease and F decrease meet the frequency threshold, and determining that the frequency threshold corresponds to each other resource change information.

Digging for the first time: determining other resource change information meeting the frequency threshold from the corresponding other resource change information, adding the other resource change information to the initial information sequence feature generated by the candidate information, and determining each other resource change information, as shown in fig. 4U, the obtained updated information sequence features are B increase-C increase, B increase-D decrease, B increase-F decrease, C increase-D decrease, C increase-F decrease, D decrease-F decrease.

And (5) digging for the second time: determining other resource change information meeting the frequency threshold from the corresponding other resource change information, adding the other resource change information to the updated information sequence characteristics, and determining each other resource change information, as shown in fig. 4U, wherein the obtained updated information sequence characteristics are B increase-C increase-D decrease, B increase-C increase-F decrease, B increase-D decrease-F decrease, and C increase-D decrease-F decrease.

Digging for the third time: determining other resource change information meeting the frequency threshold from the corresponding other resource change information, adding the other resource change information to the updated information sequence characteristics, and determining each other resource change information, as shown in fig. 4U, wherein the obtained updated information sequence characteristics are B increase-C increase-D decrease-F decrease.

And finally, determining the information sequence characteristics as B increase-C increase-D decrease-F decrease, and coding the information sequence characteristics to be used as resource characteristics.

Step 41: adopting the trained route recommendation model, aiming at least two candidate driving routes, respectively executing the following operations: respectively obtaining the flow characteristic correlation degree between the flow characteristic of one candidate driving route and the flow characteristic of each reference recommended route; and respectively obtaining resource feature correlation degrees between the resource features of the candidate driving route and the resource features of the reference recommended routes.

In the embodiment of the present application, when step 41 is executed, it is required to obtain a resource feature association degree and a flow feature association degree between each candidate driving route and each reference recommended route for at least two candidate driving routes, specifically, taking the candidate driving route i as an example, a detailed process of obtaining the flow feature association degree and the resource feature association degree between the candidate driving route i and each reference recommended route is described as follows:

first, a process of obtaining a flow characteristic association degree between a candidate driving route i and each reference recommended route in the embodiment of the present application is described in detail, and fig. 4V is a schematic flow chart for determining the flow characteristic association degree in the implementation of the present application.

Reading each flow rate sub-feature corresponding to the flow rate feature of the candidate driving route i in sequence to obtain a sub-feature association degree corresponding to each flow rate sub-feature, specifically, taking any one flow rate sub-feature (hereinafter, referred to as a flow rate sub-feature c) as an example, a process of obtaining the sub-feature association degree of the flow rate sub-feature c is described as follows:

step 41-1-1: determining the reference flow sub-feature association degree corresponding to the reference flow sub-feature as the flow sub-feature association degree corresponding to the flow sub-feature c when the reference flow sub-feature which is the same as the flow sub-feature c exists in all the reference flow sub-features corresponding to the flow feature of the reference recommended route; wherein each reference flow sub-feature relevance degree characterizes a relevance degree between the reference flow sub-feature and each reference recommended route.

Firstly, judging whether a reference flow sub-feature identical to the flow sub-feature c exists in each reference flow sub-feature corresponding to the flow feature of a reference recommended route, which can be divided into the following two cases:

in the first case: there is a reference flow sub-feature that is the same as flow sub-feature c.

If it is determined that the reference flow sub-feature which is the same as the flow sub-feature c exists in the reference flow sub-features corresponding to the flow features of the reference recommended route, obtaining a reference flow sub-feature association degree corresponding to the reference flow sub-feature, and taking the obtained reference flow sub-feature association degree as a sub-feature association degree corresponding to the flow sub-feature c.

For example, if the influence factor is whether the peak hour of going to and from work, and the value of the flow sub-feature corresponding to the influence factor is 1, the candidate driving route is characterized as the peak hour of going to and from work, then, it is determined whether the reference traffic sub-feature corresponding to one reference recommended route includes an influence factor of the on-duty peak and the off-duty peak, and the value is 1, if it is determined that the flow sub-characteristic with the influence factor of the off-duty peak period exists in each reference flow sub-characteristic, and the corresponding value is 1, it is determined that the same reference flow sub-signature as the flow sub-signature exists in each reference flow sub-signature, and obtaining a reference flow sub-feature correlation degree 0.89 corresponding to the reference flow sub-feature, and taking the determined reference flow sub-feature correlation degree 0.89 as a sub-feature correlation degree corresponding to the flow sub-feature, namely, the sub-feature correlation degree corresponding to the flow sub-feature c of the candidate driving route i is 0.89.

In the second case: there is no reference flow sub-feature that is the same as flow sub-feature i.

And if the reference flow sub-feature which is the same as the flow sub-feature c does not exist in the reference flow sub-features corresponding to the flow features of the reference recommended route, determining that the sub-feature correlation degree corresponding to the flow sub-feature c is 0.

For example, if it is assumed that whether the influence factor is a large business turn, and the value of the flow sub-feature corresponding to the influence factor is 1, which indicates that a large business turn exists near the destination reached by the candidate driving route, it is determined whether each reference flow sub-feature corresponding to one reference recommended route includes a flow sub-feature corresponding to the influence factor and the value of which is 1, and if it is determined that there is no influence factor in each reference flow sub-feature and the corresponding reference flow sub-feature with the value of 1, it is determined that there is no reference flow sub-feature identical to the flow sub-feature in each reference flow sub-feature, and the association degree of the flow sub-feature corresponding to the flow sub-feature c of the candidate driving route i is 0.

Step 41-1-2: and accumulating the obtained flow sub-feature association degrees to obtain an accumulation result, and taking the accumulation result as the flow feature association degree corresponding to the candidate driving route i and one reference recommended route.

In the embodiment of the application, the obtained flow sub-feature association degrees are added to obtain an accumulation result, and the finally obtained accumulation result is used as the flow feature association degree of the candidate driving route i corresponding to one reference recommended route.

For example, assume that the flow rate sub-characteristics are 0.9, 0.8 and 0.9, respectively, and therefore the flow rate characteristic association degree of the candidate driving route i with respect to one reference recommended route is 2.6.

It should be noted that, as the flow characteristic association degree is higher, the probability that the characteristic candidate driving route is the target driving route is higher.

Next, a process of obtaining the resource feature association degree between the candidate driving route i and each reference recommended route in the embodiment of the present application is described in detail, and reference is made to fig. 4W, which is a schematic flow chart illustrating the determination of the resource feature association degree in the implementation of the present application.

Step 41-2-1: determining the reference resource sub-feature correlation degree corresponding to the reference resource sub-feature as the resource sub-feature correlation degree corresponding to the resource sub-feature d when the reference resource sub-feature identical to the resource sub-feature d exists in each reference resource sub-feature corresponding to the resource feature of one reference recommended route; wherein each reference flow sub-feature relevance degree characterizes a relevance degree between the reference resource sub-feature and each reference recommended route.

In the embodiment of the present application, specifically, a process of determining a resource sub-feature relevance is introduced by taking any one resource sub-feature as an example (hereinafter referred to as a resource sub-feature d) as follows:

firstly, judging whether a reference resource sub-feature identical to the resource sub-feature d exists in each resource flow sub-feature corresponding to the resource feature of a reference recommended route, which can be divided into the following two cases:

in the first case: there is a reference resource sub-feature that is the same as resource sub-feature d.

If the reference resource sub-feature which is the same as the resource sub-feature d exists in all the reference resource sub-features corresponding to the resource feature of one reference recommended route, acquiring the reference resource sub-feature association degree corresponding to the reference resource sub-feature, and taking the acquired reference resource sub-feature association degree as the sub-feature association degree corresponding to the resource sub-feature d.

For example, if the resource sub-feature is a value a, B, C, or C, it is determined whether the reference resource sub-feature corresponding to the value a, B, C, or C is included in the reference resource sub-feature corresponding to the reference recommended route, if the reference resource sub-feature is determined to exist, the reference resource sub-feature association degree 0.77 corresponding to the reference resource sub-feature is obtained, and the determined reference resource sub-feature 0.77 is used as the sub-feature association degree corresponding to the resource sub-feature.

It should be noted that, not all resource sub-features in the resource feature of one reference recommended route are reference resource sub-features.

In the second case: there is no reference resource sub-feature that is the same as resource sub-feature d.

And if the reference resource sub-feature which is the same as the resource sub-feature d does not exist in the reference resource sub-features corresponding to the resource features of the reference recommended route, determining that the sub-feature association degree corresponding to the resource sub-feature d is 0.

For example, if the resource sub-feature is a value a, B, D, or D, it is determined whether each reference resource sub-feature corresponding to the resource feature of one reference recommended route includes a reference resource sub-feature corresponding to the value a, B, D, or D, and if it is determined that the reference resource sub-feature does not exist, the resource sub-feature association degree corresponding to the resource sub-feature D is determined to be 0.

Step 41-2-2: and accumulating the obtained resource sub-feature association degrees to obtain an accumulated result, and taking the accumulated result as the resource feature association degree corresponding to the candidate driving route i and one reference recommended route.

In the embodiment of the application, the obtained resource sub-feature association degrees are added to obtain an accumulation result, and the finally obtained accumulation result is used as the resource feature association degree corresponding to the candidate driving route i and the reference recommended route.

For example, assume that the resource sub-features are 0.9, 0.8 and 0.7, respectively, and therefore the resource feature association degree of the candidate driving route i with one reference recommended route is 2.5.

It should be noted that, as the resource feature association degree is higher, the probability that the characteristic candidate driving route is the target driving route is higher.

In the embodiment of the application, the resource characteristics and the flow characteristics are combined, the characteristic independence assumption can be met, so that the resource characteristic association degree and the flow characteristic association degree corresponding to at least two candidate driving routes are obtained through a route recommendation model based on Bayes, a target driving route is determined from the at least two candidate driving routes, and the accuracy of route recommendation can be improved by combining the resource characteristics and the flow characteristics for fusion analysis.

Referring to fig. 5A, a schematic flow chart of determining a target driving route in the embodiment of the present application is shown, and a process of selecting a corresponding target driving route from at least two candidate driving routes in the embodiment of the present application is described below with reference to fig. 5A:

step 50: respectively aiming at least two candidate driving routes, respectively executing the following operations: and accumulating the traffic characteristic association degrees corresponding to one candidate driving route to obtain a traffic characteristic association degree sum, and accumulating the resource characteristic association degrees corresponding to one candidate driving route to obtain a resource characteristic association degree sum.

In the embodiment of the present application, specifically, taking the candidate driving route i as an example, a process of determining the total of the flow characteristic association degrees is described as follows: and accumulating the traffic characteristic association degrees corresponding to the candidate driving route i to obtain a traffic characteristic association degree sum, and accumulating the resource characteristic association degrees corresponding to the candidate driving route i to obtain a resource characteristic association degree sum.

For example, if the flow characteristic association degrees corresponding to the candidate driving route i are 2, 3, and 2.1, respectively, the total of the obtained flow characteristic association degrees is 7.1.

Step 51: and selecting a corresponding flow characteristic association degree sum from at least two candidate driving routes, wherein the resource characteristic association degree sum is not less than a preset association degree threshold value.

In the embodiment of the application, the following operations are respectively executed for at least two candidate driving routes: and determining whether the sum of the flow characteristic association degrees corresponding to one candidate driving route is not less than a preset association degree threshold value or not, and determining whether the sum of the resource characteristic association degrees corresponding to one candidate driving route is not less than a preset association degree threshold value or not. Then, determining candidate driving routes of which the sum of the relevance degrees of the flow characteristics is not less than a preset relevance threshold value and the sum of the relevance degrees of the resource characteristics is not less than the preset relevance threshold value from at least two candidate driving routes.

For example, assuming that the preset association threshold is 7, the sum of the flow characteristic association degrees corresponding to the candidate driving route 1 is 5, the sum of the corresponding resource characteristic association degrees is 4, the sum of the flow characteristic association degrees corresponding to the candidate driving route 2 is 5, the sum of the corresponding resource characteristic association degrees is 3, the sum of the flow characteristic association degrees corresponding to the candidate driving route 3 is 3, and the sum of the corresponding resource characteristic association degrees is 2, it is determined that the sum of the flow characteristic association degrees and the sum of the resource characteristic association degrees corresponding to the candidate driving route 2 and the candidate driving route 3 are not less than the preset association threshold.

It should be noted that, in the embodiment of the present application, the relevance threshold of the traffic characteristic relevance and the resource characteristic relevance may be the same, and of course, different relevance thresholds may be set for the traffic characteristic relevance and the resource characteristic relevance. For example, the traffic characteristic relevance threshold is 9, and the resource characteristic relevance threshold is 6.

Step 52: aiming at least two candidate driving routes which are selected, the following operations are respectively executed: and adding the sum of the flow characteristic association degrees of the selected candidate driving route and the sum of the resource characteristic association degrees to obtain the corresponding sum of the characteristic association degrees.

In the embodiment of the application, the following operations are respectively executed for at least two selected candidate driving routes: and adding the sum of the flow characteristic association degrees of the selected candidate running route and the sum of the resource characteristic association degrees corresponding to the candidate running route to obtain the sum of the characteristic association degrees corresponding to the candidate running route.

Optionally, in this embodiment of the present application, an obtaining manner of a feature association degree sum may be expressed as:

wherein, a is a recommended route of each reference, q₁As flow characteristics of the candidate driving route, q₂As a resource characteristic of the candidate driving route, I (a, q)_i) For sum of feature relevance, I (a, q)₁) For the sum of the flow characteristic association degrees, I (a, q), between each reference recommended route and the candidate driving route₂) And the sum of the resource characteristic association degrees between each reference recommended route and the candidate driving route is obtained.

Step 53: and taking the candidate driving route corresponding to the characteristic association degree sum with the maximum value as a target driving route.

In specific implementation, after the feature association degree sum of at least two selected candidate driving routes is determined, the candidate driving route corresponding to the feature association degree sum with the largest value is used as the target driving route.

In the embodiment of the application, the candidate driving routes corresponding to the flow characteristic relevance sum and the resource characteristic relevance sum meeting the relevance threshold are selected, and the target driving route is determined from the selected candidate driving routes, so that the selected candidate driving routes can meet the relevance threshold, the characteristic relevance sum is maximum, and the accuracy of route recommendation can be improved.

Based on the foregoing embodiment, a specific example is adopted below to describe in detail the method for determining the target driving route in the embodiment of the present application, and refer to fig. 5B, which is a schematic flow chart of the route recommendation in the embodiment of the present application, taking a candidate place as an example of a parking lot.

It is assumed that the candidate travel routes are the candidate travel route 1 and the candidate travel route 2, respectively, that the candidate travel route 1 corresponds to the travel section B, C, D, E, that the reachable parking lots are the parking lot E, the parking lot a, and the parking lot D, that the candidate travel route 2 corresponds to the travel section A, F, E, and that the reachable parking lots are the parking lot F, the parking lot B, and the parking lot a.

Then, feature extraction is carried out on the number of vacant parking spaces of each parking lot corresponding to the candidate driving route 1 to obtain resource features E plus-A plus, and feature extraction is carried out on flow data corresponding to each driving road section and influence factors influencing object density to obtain flow features 40,30, -50,10,0,1, 0. And (3) performing feature extraction on the number of vacant parking spaces of each parking lot corresponding to the candidate driving route 1 to obtain a resource feature F minus-B minus-A minus, and performing feature extraction on the flow data corresponding to each driving road section and the influence factors influencing the object density to obtain flow features 90,20,50,1,1 and 0.

Then, the resource feature E corresponding to the candidate driving route 1 is added to the route recommendation model, the resource feature E is added to the route recommendation model, the resource feature association degree is 0.98, the flow feature association degree is 0.65, the resource feature F corresponding to the candidate driving route 2 is subtracted from the resource feature E, the flow feature association degree is 0.98, the flow feature association degree is 0.65, the resource feature F corresponding to the candidate driving route is subtracted from the flow feature F, the resource feature F corresponding to the candidate driving route is subtracted from the flow feature E, the flow feature association degree is 0.56, and the flow feature association degree is 0.53.

And finally, recommending the candidate driving route 1 as a target driving route to the vehicle owner.

Based on the same inventive concept, the embodiment of the application also provides a device for determining the target driving route. Referring to fig. 6A, which is a schematic structural diagram of a device for determining a target driving route according to an embodiment of the present application, the device may include:

a first obtaining module 600, configured to obtain at least two candidate driving routes between a target object and a destination in response to a destination selection operation triggered by the target object;

a second obtaining module 601, configured to obtain at least one candidate place with a specified resource, where each of the at least two candidate driving routes corresponds to one of the at least two candidate driving routes, and each candidate place is located within a preset range around the destination;

a first processing module 602, configured to obtain traffic characteristic association degrees and resource characteristic association degrees between at least two candidate driving routes and each reference recommended route respectively by using a trained route recommendation model; each reference recommended route is a route which meets recommended conditions and is determined according to historical flow and historical resource quantity;

the screening module 603 is configured to select, from the at least two candidate driving routes, a candidate driving route for which the corresponding traffic characteristic association degree and the corresponding resource characteristic association degree both satisfy a preset association degree condition as a target driving route.

Optionally, the first processing module 602 is further configured to:

Optionally, when the flow characteristics corresponding to at least two candidate driving routes are obtained respectively, the first processing module 602 is further configured to:

Optionally, when the resource features corresponding to the at least two candidate driving routes are respectively obtained, the first processing module 602 is further configured to:

for at least two candidate driving routes, respectively executing the following operations:

Optionally, when determining the resource feature corresponding to one candidate driving route based on the obtained resource change information according to a preset feature extraction manner, the first processing module 602 is further configured to:

aiming at each historical sampling time interval, respectively generating a corresponding information sequence based on the resource change information of at least one candidate corresponding to one candidate running route in the corresponding historical sampling time interval;

obtaining at least one candidate information set corresponding to a candidate place based on the obtained information sequences, wherein each candidate information is resource change information of which the occurrence frequency meets a preset frequency condition in the resource change information corresponding to the corresponding candidate place;

sequentially reading each candidate information contained in the candidate information set, and executing the following operations every time one candidate information is read: taking each information sequence containing one currently read candidate information as a candidate information sequence, and respectively generating corresponding information sequence characteristics in each candidate information sequence based on one candidate information and other resource change information; the other resource change information is resource change information which is arranged behind one candidate information in each candidate information sequence according to a preset sequence;

and selecting each target information sequence characteristic which meets the preset quantity condition from each information sequence characteristic as a resource characteristic corresponding to one candidate driving route based on the total number of the resource change information associated with each information sequence characteristic.

Optionally, when each information sequence including one currently read candidate information is used as a candidate information sequence, and corresponding information sequence features are generated in each candidate information sequence based on one candidate information and other resource change information, the first processing module 602 is further configured to:

taking a currently read candidate message as an initial message sequence feature;

respectively adding other resource change information of which the occurrence frequency meets the frequency condition in the other resource change information contained in each candidate information sequence to corresponding initial information sequence characteristics to obtain each updated information sequence characteristic;

screening out each candidate information sequence containing each updated information sequence characteristic from each candidate information sequence respectively to serve as each corresponding updated candidate information sequence;

and taking each updated information sequence characteristic as each new initial information sequence characteristic.

Optionally, when obtaining a flow characteristic association degree between a flow characteristic of a candidate driving route and a flow characteristic of a reference recommended route, the first processing module 602 is specifically configured to:

Optionally, when obtaining the resource feature association degree between the flow feature of one candidate driving route and the resource feature of one reference recommended route, the first processing module 602 is further configured to:

Optionally, the screening module 603 is further configured to:

Optionally, in an embodiment of the present application, a training device for a route recommendation model is further provided, and as shown in fig. 6B, a schematic structural diagram of the training device for the route recommendation model in the embodiment of the present application is shown:

a third obtaining module 604, configured to obtain a sample set, where the sample set includes each reference recommended route and each non-recommended route, each reference recommended route and each non-recommended route respectively correspond to a candidate feature, the candidate feature includes multiple candidate sub-features, and each candidate sub-feature is a resource sub-feature or a traffic sub-feature;

a classification module 605, configured to classify candidate sub-features of each reference recommended route and each non-recommended route, respectively, to obtain various candidate sub-features;

a second processing module 606, configured to obtain respective sub-feature relevance degrees corresponding to the various candidate sub-features, and use the candidate sub-features with the sub-feature relevance degree greater than a preset sub-feature relevance degree threshold as reference sub-features, where the sub-feature relevance degree is a probability that the corresponding sub-features are included in each candidate sub-feature of each reference recommended route;

the training module 607 is configured to use the flow sub-features in each reference sub-feature as the reference flow sub-features learned by the route recommendation model, and use the resource sub-features in each reference sub-feature as the reference resource sub-features learned by the route recommendation model.

Optionally, when obtaining the respective sub-feature association degrees corresponding to the various candidate sub-features, the second processing module 606 is further configured to:

Having described the method and apparatus for determining a target travel route according to an exemplary embodiment of the present application, an electronic device according to another exemplary embodiment of the present application will be described next.

As will be appreciated by one skilled in the art, aspects of the present application may be embodied as a system, method or program product. Accordingly, various aspects of the present application may be embodied in the form of: an entirely hardware embodiment, an entirely software embodiment (including firmware, microcode, etc.) or an embodiment combining hardware and software aspects that may all generally be referred to herein as a "circuit," module "or" system.

Based on the same inventive concept as the method embodiment, an electronic device is further provided in the embodiment of the present application, referring to fig. 7, which is a schematic diagram of a hardware component structure of an electronic device to which the embodiment of the present application is applied, and the electronic device 700 may at least include a processor 701 and a memory 702. The memory 702 stores therein program codes, which, when executed by the processor 701, cause the processor 701 to perform the steps of any one of the above-described methods for determining a target travel route.

In some possible implementations, a computing device according to the present application may include at least one processor, and at least one memory. Wherein the memory stores program code which, when executed by the processor, causes the processor to perform the steps of determining a target driving route according to various exemplary embodiments of the present application described above in the present specification. For example, a processor may perform the steps as shown in FIG. 3A.

A computing device 800 according to this embodiment of the present application is described below with reference to fig. 8. As shown in fig. 8, computing device 800 is embodied in the form of a general purpose computing device. Components of computing device 800 may include, but are not limited to: the at least one processing unit 801, the at least one memory unit 802, and a bus 803 that couples various system components including the memory unit 802 and the processing unit 801.

Bus 803 represents one or more of any of several types of bus structures, including a memory bus or memory controller, a peripheral bus, a processor, or a local bus using any of a variety of bus architectures.

The storage unit 802 may include readable media in the form of volatile memory, such as Random Access Memory (RAM) 8021 and/or cache storage unit 8022, and may further include Read Only Memory (ROM) 8023.

Storage unit 802 can also include a program/utility 8025 having a set (at least one) of program modules 8024, such program modules 8024 including, but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which, or some combination thereof, may comprise an implementation of a network environment.

The computing device 800 may also communicate with one or more external devices 804 (e.g., keyboard, pointing device, etc.), with one or more devices that enable a user to interact with the computing device 800, and/or with any devices (e.g., router, modem, etc.) that enable the computing device 800 to communicate with one or more other computing devices. Such communication may be through input/output (I/O) interfaces 805. Moreover, the computing device 800 may also communicate with one or more networks (e.g., a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network, such as the internet) via the network adapter 806. As shown, the network adapter 806 communicates with other modules for the computing device 800 over the bus 803. It should be understood that although not shown in the figures, other hardware and/or software modules may be used in conjunction with the computing device 800, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, among others.

Based on the same inventive concept as the above method embodiments, various aspects of the method for determining a target travel route provided by the present application may also be implemented in the form of a program product, which includes program code for causing an electronic device to perform the steps in the method for determining a target travel route according to various exemplary embodiments of the present application described above in this specification when the program product is run on the electronic device, for example, the electronic device may perform the steps as shown in fig. 3A.

The program product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

While the preferred embodiments of the present application have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all alterations and modifications as fall within the scope of the application.

It will be apparent to those skilled in the art that various changes and modifications may be made in the present application without departing from the spirit and scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is intended to include such modifications and variations as well.

Claims

1. A method of determining a target travel route, comprising:

for the at least two candidate driving routes, respectively performing the following operations: determining the flow characteristic of one candidate driving route and the flow characteristic association degree between the flow characteristic of the corresponding reference recommended route respectively based on the reference flow sub-characteristic which is the same as each flow sub-characteristic of one candidate driving route in each reference recommended route and the corresponding reference flow sub-characteristic association degree and determining the resource characteristic of one candidate driving route and the resource characteristic association degree between the resource characteristics of the corresponding reference recommended route respectively based on the reference resource sub-characteristic which is the same as each resource sub-characteristic of one candidate driving route in each reference recommended route and the corresponding reference resource sub-characteristic association degree by adopting a trained route recommendation model;

each reference recommended route is a route which meets recommended conditions and is determined according to historical flow and historical resource quantity; each reference flow sub-feature association degree characterizes an association degree between a reference flow sub-feature and each reference recommended route; the reference resource sub-feature association degree corresponding to each reference resource sub-feature characterizes the association degree between the reference resource sub-feature and each reference recommended route; the resource characteristics are determined based on resource change information corresponding to different candidate places;

2. The method according to claim 1, wherein obtaining the flow characteristics corresponding to the at least two candidate driving routes respectively comprises:

coding each influence factor of a candidate driving route according to a preset coding mode to obtain an object density characteristic of the candidate driving route, and determining the flow characteristic of the candidate driving route according to the obtained current flow and flow change information corresponding to each driving road section and the object density characteristic, wherein the influence factor represents the factor influencing the object density of the candidate driving route.

3. The method according to claim 1, wherein the obtaining the resource characteristics corresponding to the at least two candidate driving routes respectively comprises:

4. The method according to claim 3, wherein determining the resource feature corresponding to the candidate driving route based on the obtained resource change information according to a preset feature extraction manner specifically comprises:

5. The method according to claim 4, wherein using each information sequence containing one currently read candidate information as a candidate information sequence, and generating corresponding information sequence characteristics based on the one candidate information and other resource change information in each candidate information sequence respectively comprises:

6. The method according to any one of claims 1 to 5, wherein the determining of the flow characteristic association between the flow characteristic of one candidate driving route and the flow characteristic of the corresponding reference recommended route based on the same reference flow sub-characteristic as each flow sub-characteristic of the candidate driving route and the corresponding reference flow sub-characteristic association degree in each reference recommended route specifically comprises:

sequentially reading each flow sub-feature corresponding to the flow feature of one candidate driving route, and when a reference flow sub-feature identical to the one flow sub-feature exists in each reference flow sub-feature corresponding to the flow feature of one reference recommended route, taking the reference flow sub-feature correlation degree corresponding to the reference flow sub-feature as the flow sub-feature correlation degree corresponding to the one flow sub-feature;

7. The method according to any one of claims 1 to 5, wherein the determining the resource feature association between the resource feature of the candidate driving route and the resource feature of the corresponding reference recommended route based on the reference resource sub-feature which is the same as the resource sub-feature of the candidate driving route and the corresponding reference resource sub-feature association respectively comprises:

sequentially reading each resource sub-feature corresponding to the resource feature of one candidate driving route, and when a reference resource sub-feature identical to the resource sub-feature exists in each reference resource sub-feature corresponding to the resource feature of one reference recommended route, taking the reference resource sub-feature correlation degree corresponding to the reference resource sub-feature as the resource sub-feature correlation degree corresponding to the resource sub-feature;

8. The method of claim 1, wherein the route recommendation model is trained by:

obtaining a sample set, wherein the sample set comprises each reference recommended route and each non-recommended route, each reference recommended route and each non-recommended route respectively correspond to a candidate feature, the candidate feature comprises a plurality of candidate sub-features, and each candidate sub-feature is a resource sub-feature or a flow sub-feature;

classifying the candidate sub-features of the reference recommended routes and the non-recommended routes respectively to obtain various candidate sub-features;

respectively obtaining sub-feature association degrees corresponding to the various candidate sub-features, and taking the candidate sub-features with the sub-feature association degrees larger than a preset sub-feature association degree threshold value as reference sub-features, wherein the sub-feature association degrees are the probabilities that the corresponding sub-features are included in the various candidate sub-features of the various reference recommended routes;

and taking the flow sub-features in the reference sub-features as the reference flow sub-features learned by the route recommendation model, and taking the resource sub-features in the reference sub-features as the reference resource sub-features learned by the route recommendation model.

9. The method according to claim 8, wherein the obtaining the respective sub-feature association degrees corresponding to the various candidate sub-features respectively comprises:

10. The method according to claim 1, wherein selecting, from the at least two candidate driving routes, a candidate driving route for which the corresponding traffic characteristic relevance degree and the corresponding resource characteristic relevance degree both satisfy a preset relevance degree condition as a target driving route specifically comprises:

11. A target travel route determination device, characterized by comprising:

the first processing module is used for respectively obtaining flow characteristics and resource characteristics corresponding to the at least two candidate driving routes; for the at least two candidate driving routes, respectively performing the following operations: determining the flow characteristic of one candidate driving route and the flow characteristic association degree between the flow characteristic of the corresponding reference recommended route respectively based on the reference flow sub-characteristic which is the same as each flow sub-characteristic of one candidate driving route in each reference recommended route and the corresponding reference flow sub-characteristic association degree and determining the resource characteristic of one candidate driving route and the resource characteristic association degree between the resource characteristics of the corresponding reference recommended route respectively based on the reference resource sub-characteristic which is the same as each resource sub-characteristic of one candidate driving route in each reference recommended route and the corresponding reference resource sub-characteristic association degree by adopting a trained route recommendation model;

12. An electronic device, characterized in that it comprises a processor and a memory, wherein the memory stores program code which, when executed by the processor, causes the processor to carry out the steps of the method of any of claims 1-10.

13. A computer-readable storage medium, characterized in that it comprises program code for causing an electronic device to perform the steps of the method of any of claims 1-10, when said program code is run on the electronic device.