CN115083147A

CN115083147A - Destination prediction method and device and vehicle

Info

Publication number: CN115083147A
Application number: CN202210507730.5A
Authority: CN
Inventors: 郭和攀; 郭嘉强; 翟振威; 石静迎
Original assignee: Guangzhou Automobile Group Co Ltd
Current assignee: Guangzhou Automobile Group Co Ltd
Priority date: 2022-05-10
Filing date: 2022-05-10
Publication date: 2022-09-20
Anticipated expiration: 2042-05-10
Also published as: CN115083147B

Abstract

The embodiment of the application discloses a destination prediction method and device and a vehicle. The method comprises the following steps: acquiring a candidate destination database of the vehicle; acquiring current vehicle information of the vehicle, wherein the vehicle information comprises position information and time information; acquiring initial destination information; matching the initial destination information with the candidate destination database to obtain candidate destination information, wherein the candidate destination information comprises position information and time information; and if a plurality of candidate destinations corresponding to the candidate destination information exist, obtaining a destination prediction result based on the vehicle information and the candidate destination information. By the method, the initial destination information can be matched with the candidate destination database to obtain the candidate destination information, and the destination prediction result can be obtained based on the multi-dimensional information under the condition that a plurality of candidate destinations corresponding to the candidate destination information exist, so that the accuracy of destination prediction is improved.

Description

Destination prediction method and device and vehicle

Technical Field

The present application relates to the field of automotive technologies, and in particular, to a destination prediction method and apparatus, and a vehicle.

Background

As automobile technology develops and driving road conditions become more complex, consumer demands and expectations for vehicles become higher and higher. During the running process of the vehicle, the destination can be predicted according to the relevant information of the vehicle of the user, and important reference information can be provided for the running of the user. However, the related method also has a problem that the destination prediction is not accurate.

Disclosure of Invention

In view of the above problems, the present application provides a destination prediction method, a destination prediction apparatus, and a vehicle to improve the above problems.

In a first aspect, the present application provides a destination prediction method, including: acquiring a candidate destination database of the vehicle; acquiring current vehicle information of the vehicle, wherein the vehicle information comprises position information and time information; acquiring initial destination information; matching the initial destination information with the candidate destination database to obtain candidate destination information, wherein the candidate destination information comprises position information and time information; and if a plurality of candidate destinations corresponding to the candidate destination information exist, obtaining a destination prediction result based on the vehicle information and the candidate destination information.

In a second aspect, the present application provides a destination prediction apparatus, the apparatus comprising: a candidate destination database acquisition unit for acquiring a candidate destination database of the vehicle; the vehicle information acquisition unit is used for acquiring current vehicle information of the vehicle, and the vehicle information comprises position information and time information; a destination prediction unit for acquiring initial destination information; matching the initial destination information with the candidate destination database to obtain candidate destination information, wherein the candidate destination information comprises position information and time information; and if a plurality of candidate destinations corresponding to the candidate destination information exist, obtaining a destination prediction result based on the vehicle information and the candidate destination information.

In a third aspect, the present application provides a vehicle comprising a processor, a network module, and a memory; one or more programs are stored in the memory and configured to be executed by the processor, the one or more programs configured to perform the methods described above.

In a fourth aspect, the present application provides a computer-readable storage medium having a program code stored therein, wherein the program code performs the above method when running.

According to the destination prediction method, the destination prediction device, the vehicle and the storage medium, after a candidate destination database of the vehicle is obtained, vehicle information of the vehicle including current position information and current time information is obtained, initial destination information is obtained, the initial destination information is matched with the candidate destination database, candidate destination information including the position information and the time information is obtained, and if a plurality of candidate destinations corresponding to the candidate destination information exist, a destination prediction result is obtained based on the vehicle information and the candidate destination information. By the method, the initial destination information can be matched with the candidate destination database to obtain the candidate destination information, and when a plurality of candidate destinations corresponding to the candidate destination information exist, the destination prediction result can be obtained based on the multidimensional information (the current position information and the time information of the vehicle and the position information and the time information in the candidate destination information), so that the accuracy of destination prediction is improved.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

Fig. 1 is a schematic diagram illustrating an application scenario of a destination prediction method proposed in the present application;

fig. 2 is a flowchart illustrating a destination prediction method according to an embodiment of the present application;

FIG. 3 is a flow chart of an embodiment of the present application at S110 of FIG. 2;

FIG. 4 is a schematic diagram illustrating the division of a start-stop process proposed by the present application;

FIG. 5 is a flow chart illustrating an embodiment of S150 of FIG. 2;

fig. 6 shows a flow chart of a destination prediction method according to another embodiment of the present application;

FIG. 7 is a schematic diagram illustrating a dwell point determination proposed by the present application;

FIG. 8 is a diagram illustrating a map region block partitioning proposed in the present application;

FIG. 9 is a flowchart of an embodiment of S2060 of FIG. 6 of the present application;

fig. 10 is a flowchart illustrating a destination prediction method according to another embodiment of the present application;

fig. 11 is a block diagram showing a structure of a destination prediction apparatus according to an embodiment of the present application;

FIG. 12 is a block diagram illustrating the structure of a vehicle according to the present disclosure;

fig. 13 is a storage unit for storing or carrying program code implementing a destination prediction method according to an embodiment of the present application.

Detailed Description

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

In the embodiment of the application, the inventor provides a destination prediction method, a destination prediction device and a vehicle, after a candidate destination database of the vehicle is obtained, vehicle information of the vehicle, including current position information and current time information, is obtained, initial destination information is obtained, the initial destination information is matched with the candidate destination database, candidate destination information including position information and time information is obtained, and if a plurality of candidate destinations corresponding to the candidate destination information exist, a destination prediction result is obtained based on the vehicle information and the candidate destination information. By the method, the initial destination information can be matched with the candidate destination database to obtain the candidate destination information, and when a plurality of candidate destinations corresponding to the candidate destination information exist, the destination prediction result can be obtained based on the multidimensional information (the current position information and the time information of the vehicle and the position information and the time information in the candidate destination information), so that the accuracy of destination prediction is improved.

Fig. 1 is a schematic diagram of an application environment according to an embodiment of the present application. Fig. 1 provides a destination prediction system 10, which includes a base station 100, a vehicle 200 connected to the base station 100, and a cloud platform 300 connected to the base station 100.

The vehicle 200 may include a vehicle having data transmission and data processing functions. A Telematics BOX (T-BOX) may be disposed in the vehicle 200, and the T-BOX may transmit data of the vehicle 200 to the cloud platform 300 through the base station 100 and receive data transmitted by the cloud platform 300 to the vehicle 200 through the base station 100. The T-box and the vehicle 200 may be connected by a CAN (Controller Area Network) bus.

In some embodiments, the cloud platform 300 may obtain a candidate destination database for the vehicle 200 and current vehicle information for the vehicle 200, and the vehicle information and the candidate destination database predict the destination of the vehicle 200.

In addition, in the embodiment of the present application, the provided destination prediction method may be executed by the vehicle 200 independently, and may also be executed by the vehicle 200 together with the cloud platform 300.

Embodiments of the present application will be described below with reference to the accompanying drawings.

Referring to fig. 2, a destination prediction method provided in the present application includes:

s110: a database of candidate destinations for the vehicle is obtained.

As one way, as shown in fig. 3, obtaining a candidate destination database of a vehicle includes:

s111: historical driving data of the vehicle is acquired.

The driving data may be data representing an identity of the vehicle and a driving situation. Optionally, the driving data may include at least one of a Vehicle identifier, a start time, an end time, and a driving track of the Vehicle, where the Vehicle identifier may refer to a Vehicle Identification Number (VIN) of the Vehicle, each VIN may uniquely represent a Vehicle, and the driving track may represent a change in position of the Vehicle during driving. The historical travel data refers to travel data collected within a fixed time window prior to the current time.

Here, the current time may be understood as the time when step S111 is performed. The time length of the time window can be set in various ways.

Alternatively, the time length may be set based on the update frequency of the candidate destination database. For example, the candidate destination database may be updated once every month, and the time period may be set to one month.

Alternatively, the time length may be set based on the trip frequency of the vehicle, and the lower the trip frequency, the longer the time length. For example, assuming that the trip frequency of the vehicle a is higher than that of the vehicle B, if the time lengths of the vehicle a and the vehicle B are the same, the travel data corresponding to the vehicle B may be much less than that corresponding to the vehicle a, and therefore, in order to acquire more travel data about the vehicle B to improve the accuracy of the destination prediction of the vehicle B, the time length corresponding to the vehicle B may be set to be longer than that of the vehicle a.

In the embodiment of the application, the time length is set based on the travel frequency of the vehicle, so that enough historical travel data can be obtained for the vehicle with lower travel frequency, and the accuracy of destination prediction is improved. In addition, in the embodiment of the application, the time length corresponding to the historical data can be set based on various modes, so that the flexibility of the destination prediction method provided by the application is improved.

As a mode, after the vehicle is awakened through the modes of powering on a whole vehicle key, remote control and the like, the running data can be uploaded to the cloud platform in real time through the vehicle-mounted T-box, and after the cloud platform receives the running data of the vehicle, the running data can be stored in different regions according to the vehicle identification, so that when the cloud platform or the vehicle needs to use the historical running data of the vehicle, the historical running data can be acquired from the corresponding storage region based on the vehicle identification.

S112: and acquiring a plurality of historical trips corresponding to the vehicle and historical trip data corresponding to each of the plurality of historical trips based on the historical driving data, wherein each historical trip represents a historical driving route of the vehicle.

The historical travel data may include, among other things, a start time and an end time for each travel of the vehicle. The historical trip data may be data that characterizes the identity of the vehicle as well as historical trip conditions. Alternatively, the historical trip data may include a start time, an end location, etc. for each historical trip of the vehicle. The travel route may indicate a travel locus of the vehicle from the departure point to the destination, and the one-time history travel route may indicate a travel locus from the departure point to the destination any one time before the current time.

As one mode, a plurality of start-stop processes may be obtained based on a start time and an end time, then a time interval between every two adjacent start-stop processes in the plurality of start-stop processes is compared with a first preset time, wherein two adjacent start-stop processes with a corresponding time interval smaller than the first preset time are divided into a same historical travel, two adjacent start-stop processes with a corresponding time interval not smaller than the first preset time are divided into different historical travels, so as to obtain a plurality of historical travels corresponding to the vehicle, and historical travel data corresponding to each of the plurality of historical travels is obtained based on the historical travel data and the plurality of historical travels. The start-stop process may refer to a process in which the vehicle starts from one place to stops at another place, and two adjacent start-stop processes may refer to two start-stop processes adjacent in time.

Optionally, the first preset time period may be set based on a traffic condition of a city where the vehicle is located, wherein the longer the first preset time period corresponding to the city where the traffic is more likely to be congested.

Optionally, the first preset duration may be set based on different time periods, wherein the first preset duration during the morning peak, the evening peak, and the holiday trip peak may be longer than the first preset durations of other time periods. In the embodiment of the present application, there are other ways to set the first preset time period, which are not limited herein.

For example, assume that the first preset duration is T ₁ As shown in fig. 4, the start-stop process a may be adjacent to the start-stop process B, that is, the vehicle passes through the start-stop process a and then passes through the start-stop process B, and the start time of the start-stop process a may be T _a1 The termination time may be T _a2 The starting time corresponding to the starting and stopping process B may be T _b1 The termination time may be T _b2 Then, the time interval between the start-stop process a and the start-stop process B may be: t ═ T _b1 -T _a2 If T is<T ₁ If the user does not arrive at the destination after the starting and stopping process A, the user can arrive at the destination through the starting and stopping process B after a time interval T, and the starting and stopping process A and the starting and stopping process B can be divided into the same historical journey; if T is greater than or equal to T ₁ Then it can be shown that the user has reached one destination after start-stop procedure a and reached another destination through start-stop procedure B after time interval T, at which point start-stop procedure a and start-stop procedure B can be divided into different historical trips.

In the embodiment of the application, the historical trips corresponding to the vehicle are obtained by comparing the time intervals of two adjacent start-stop processes with the first preset time length, so that the historical trips can be divided more accurately from the aspect of the driving purpose of a user in combination with the actual driving condition of the vehicle, the historical destinations of the vehicle can be analyzed more accurately based on the divided historical trips, the reliability of a candidate destination database of the vehicle is improved, and the accuracy of the destination prediction method provided by the application is improved.

S113: and obtaining a candidate destination database of the vehicle based on the historical travel data and a destination prediction model.

As one approach, historical trip data may be input into the destination prediction model to obtain a candidate destination database for the vehicle.

Alternatively, the candidate destination database of the vehicle may be pre-constructed based on steps S111 to S113, and stored in a designated area after being constructed, and may be read from the designated area when the cloud platform or the vehicle needs to use the candidate destination database of the vehicle.

Alternatively, the candidate destination database of the vehicle may be generated in real time based on steps S111 to S113. For example, the cloud platform or the candidate destination database of the vehicle generated in real time by the vehicle based on the steps S111 to S113 may be generated when the vehicle is started and ready to depart from a place.

In the embodiment of the application, the candidate destination database of the vehicle is generated based on the historical travel data of the vehicle and the destination prediction model, so that the destination prediction result is more consistent with the driving habits of the vehicle, and the accuracy of destination prediction can be improved.

S120: and acquiring the current vehicle information of the vehicle, wherein the vehicle information comprises position information and time information.

The current vehicle information may be information corresponding to the vehicle and used for predicting a destination, the vehicle information may include current location information of the vehicle and time information, the location information may refer to a current longitude and latitude of the vehicle, and the time information may refer to a current date and time. The time information may be used to distinguish whether the vehicle corresponds to a work day or a holiday. As one mode, current position information of the vehicle may be acquired based on a GPS (Global Positioning System); the current time information of the vehicle may be acquired based on a timer in the vehicle, or a network.

S130: initial destination information is acquired.

Wherein the initial destination information may be incomplete destination information input by the user.

As one mode, the vehicle may upload initial destination information input by the user to the cloud platform through the vehicle-mounted T-box based on interaction between the user and the vehicle, so that the cloud platform may acquire the initial destination information. For example, the destination may be company XX, the user may directly input XX in a center screen of the vehicle, and the vehicle uploads XX as initial destination information to the cloud platform through the vehicle-mounted T-box in response to a text input operation of the user. For another example, the destination may be company XX, the user may input XX by voice, and the vehicle uploads XX as initial destination information to the cloud platform through the in-vehicle T-box in response to the voice input operation of the user.

S140: and matching the initial destination information with the candidate destination database to obtain candidate destination information, wherein the candidate destination information comprises position information and time information.

The position information in the candidate destination information may be longitude and latitude corresponding to the candidate destination. The time information in the candidate destination information may refer to a stay date and a time corresponding to the candidate destination, and the time information in the candidate destination information may be used to distinguish whether the time corresponding to the candidate destination is a weekday or a holiday.

As one way, the cloud platform may match the initial destination information with all destinations in the candidate destination database one by one, and if there is a destination that matches the initial destination information in the candidate destination database, add a destination that can be matched as a candidate destination, and obtain candidate destination information based on the candidate destination. The candidate destination information may include location information and time information corresponding to the candidate destination, and the candidate destination information is added to the original existing candidate destinations by adding a new candidate destination.

For example, the initial destination information may be XX, the candidate destination database may include XX company, XX park, XX production base, AA kindergarten, BB scenic spot, and the like, and the initial destination information is matched with all destinations in the candidate destination database one by one, and the candidate destination information may be XX company, XX park, XX production base.

S150: and if a plurality of candidate destinations corresponding to the candidate destination information exist, obtaining a destination prediction result based on the vehicle information and the candidate destination information.

As one mode, as shown in fig. 5, if there are a plurality of destination candidates corresponding to the destination candidate information, obtaining a destination prediction result based on the vehicle information and the destination candidate information includes:

s151: and obtaining a first probability value corresponding to each of the candidate destinations based on position information in the vehicle information and position information corresponding to each of the candidate destinations, wherein the closer the distance between the vehicle and the candidate destination is, the larger the first probability value corresponding to the candidate destination is.

As one mode, the distances between the plurality of candidate destinations and the current position of the vehicle may be obtained based on the position information in the vehicle information and the position information corresponding to the plurality of candidate destinations, and then the first probability values corresponding to the plurality of candidate destinations may be obtained based on the distances.

For example, the candidate destination may have A, B, C, and the initial probability of A, B, C being a destination predictor may be P _A ＝P _B ＝P _C And P is _A +P _B +P _C 1, the distance S between the vehicle current position and the candidate destination information may be calculated A, B, C based on the position information in the candidate destination information and the position information in the vehicle information, respectively _A 、S _B 、S _C Is then based on S _A 、S _B 、S _C The initial probability values are adjusted to obtain first probability values P corresponding to the candidate destinations A, B, C respectively _A1 、P _B1 、P _C1 If S is _A <S _B <S _C Then P is _A1 >P _B1 >P _C1 。

S152: and obtaining second probability values corresponding to a plurality of candidate destinations based on time information in the vehicle information and time information corresponding to the candidate destinations, wherein the more similar the vehicle is to the time information of the candidate destinations, the greater the second probability values corresponding to the candidate destinations are.

As one mode, the similarity between each of the candidate destinations and the current time of the vehicle may be obtained based on the time information in the vehicle information and the time information corresponding to each of the candidate destinations, and then the second probability values corresponding to each of the candidate destinations may be obtained based on the similarity. The time information corresponding to the candidate destination is the same as the time information of the vehicle (for example, the time information is both working days or both holidays), which indicates that the time information corresponding to the candidate destination is more similar to the current time of the vehicle.

For example, the candidate destination may have A, B, C, and the initial probability of A, B, C being a destination predictor may be P _A ＝P _B ＝P _C And P is _A +P _B +P _C 1, similarity T between the candidate destination information and the current time of the vehicle may be calculated A, B, C based on the time information in the candidate destination information and the time information in the vehicle information, respectively _A 、T _B 、T _C Is then based on T _A 、T _B 、T _C The initial probability value is adjusted to obtain respective second probability values P of the candidate destinations A, B, C _A2 、P _B2 、P _C2 If T is _A >T _B >T _C Then P is _A2 >P _B2 >P _C2 。

S153: and weighting based on the first probability value and the second probability value to obtain a third probability value, and taking a candidate destination corresponding to the maximum third probability value as a destination prediction result.

As one mode, weight coefficients corresponding to the first probability value and the second probability value may be preset, the first probability value and the second probability value corresponding to each of the plurality of candidate destinations are weighted based on the preset weight coefficients, a third probability value corresponding to each of the plurality of candidate destinations is obtained, and the candidate destination corresponding to the largest third probability value is used as the destination prediction result. The sum of the weight coefficients corresponding to the first probability value and the second probability value can be 1, and the larger the weight coefficient corresponding to the probability value is, the more the probability value is depended on.

Illustratively, the candidate destination may have A, B, C and the first probability value may be P _A1 >P _B1 >P _C1 The weighting factor corresponding to the first probability value may be Q ₁ The second probability value may be P _A2 <P _B2 <P _C2 The weighting factor corresponding to the second probability value may be Q ₂ Then the third probability value is P _A3 ＝P _A1 *Q ₁ +P _A2 *Q ₂ 、P _C3 ＝P _C1 *Q ₁ +P _C2 *Q ₂ 、P _B3 ＝P _B1 *Q ₁ +P _B2 *Q ₂ If P is _A3 >P _B3 >P _C3 A may be taken as the destination prediction result.

In the embodiment of the application, after the vehicle information and the candidate destination information are obtained, the destination prediction result may be obtained in other manners. In order to facilitate understanding of other aspects, the method of obtaining the first probability values corresponding to the plurality of candidate destinations based on the location information in the vehicle information and the location information corresponding to the plurality of candidate destinations in step S151 may be referred to as a first prediction algorithm; a method for obtaining a second probability value corresponding to each of the plurality of candidate destinations based on the time information in the vehicle information and the time information corresponding to each of the plurality of candidate destinations in step S152 is referred to as a second prediction algorithm; the method for obtaining the third probability value by weighting based on the first probability value and the second probability value in step S153 is referred to as a third prediction algorithm.

As one mode, the cloud platform may perform prediction based on the vehicle information and the candidate destination information through a first prediction algorithm, a second prediction algorithm, and a third prediction algorithm, respectively, to obtain three destination prediction results, and if at least two of the three destination prediction results are the same, the same destination prediction result may be used as the destination of the vehicle; if the three destination prediction results are different, the prediction result corresponding to the preset optimal prediction algorithm can be used as the destination of the vehicle.

For example, the candidate destination may be A, B, C, the preset optimal prediction algorithm may be a third prediction algorithm, and when the vehicle information includes the position information and the time information and the candidate destination information includes the position information and the time information, if prediction results obtained based on the first prediction algorithm, the second prediction algorithm and the third prediction algorithm are A, C, A respectively, a may be used as the destination of the vehicle; if the prediction results obtained based on the first prediction algorithm, the second prediction algorithm, and the third prediction algorithm are A, B, C, respectively, the prediction result C of the third prediction algorithm, which is the optimum prediction algorithm set in advance, can be used as the destination of the vehicle.

In the embodiment of the application, the accuracy of destination prediction can be improved by obtaining the vehicle destination based on the prediction results of the first prediction algorithm, the second prediction algorithm and the third prediction algorithm.

As another mode, the cloud platform may predict the destination based on the vehicle information and the candidate destination information in sequence through a preset execution sequence of a first prediction algorithm, a second prediction algorithm and a third prediction algorithm, and if the destination prediction result can be obtained through the prediction algorithm with the top execution sequence, the subsequent prediction algorithms do not need to be executed continuously, and the destination prediction result is directly used as the destination of the vehicle; and if the destination prediction result cannot be obtained by executing the prediction algorithm at the top in the sequence, continuing to execute the next prediction algorithm in the sequence until the destination prediction result can be obtained, and taking the destination prediction result as the destination of the vehicle.

For example, A, B, C may be the candidate destination corresponding to the candidate destination information, the preset execution sequence from first to last may be a first prediction algorithm, a second prediction algorithm, and a third prediction algorithm, the cloud platform may first perform destination prediction based on the location information of the vehicle and the location information of the candidate destination information through the first prediction algorithm, and if the destination prediction result obtained through the first prediction algorithm is a, it is not necessary to continue to execute the second and third prediction algorithms, and a is directly used as the destination of the vehicle; and if the destination prediction result cannot be obtained by executing the first prediction algorithm, continuing to execute the second prediction algorithm, and taking the destination prediction result as the destination of the vehicle until the destination prediction result can be obtained by executing the first prediction algorithm.

In the embodiment of the application, the first prediction algorithm, the second prediction algorithm and the third prediction algorithm are executed according to the preset sequence, so that the destination prediction result can be obtained only by using the position information or the time information, and the calculation amount of destination prediction is reduced. In addition, in the embodiment of the application, the destination of the vehicle can be obtained based on different vehicle information and candidate destination information, so that the flexibility of the destination prediction method provided by the application is improved.

Alternatively, if only 1 candidate destination is obtained based on step S140, the candidate destination may be used as the destination of the vehicle.

In the destination prediction method provided by this embodiment, after a candidate destination database of a vehicle is obtained, vehicle information of the vehicle including current location information and current time information is obtained, initial destination information is obtained, the initial destination information is matched with the candidate destination database, candidate destination information including location information and time information is obtained, and if there are a plurality of candidate destinations corresponding to the candidate destination information, a destination prediction result is obtained based on the vehicle information and the candidate destination information. By the method, the initial destination information can be matched with the candidate destination database to obtain the candidate destination information, and when a plurality of candidate destinations corresponding to the candidate destination information exist, the destination prediction result can be obtained based on the multidimensional information (the current position information and the time information of the vehicle and the position information and the time information in the candidate destination information), so that the accuracy of destination prediction is improved.

Referring to fig. 6, a destination prediction method provided in the present application includes:

s2010: and acquiring historical driving data of the vehicle.

S2020: and acquiring a plurality of historical trips corresponding to the vehicle and historical trip data corresponding to each of the plurality of historical trips based on the historical driving data, wherein each historical trip represents a historical driving route of the vehicle.

The historical trip data may include a start time, an end time, and an end position of each historical trip of the vehicle, among other things.

S2030: and acquiring the time interval between every two adjacent historical trips in the plurality of historical trips based on the starting time and the ending time.

Wherein, two history trips that are adjacent may refer to two history trips that are adjacent in time.

For example, as shown in fig. 7, the historical trip a may be adjacent to the historical trip B, that is, the vehicle travels the historical trip B after traveling the historical trip a, and the starting time corresponding to the historical trip a may be T _a1 The termination time may be T _a2 The starting time corresponding to the historical trip B may be T _b1 The termination time may be T _b2 Then the time interval between history trip a and history trip B may be: t ═ T _b1 -T _a2 。

S2040: and taking the end point position of the previous historical travel in two adjacent historical travels of which the corresponding time interval is greater than the second preset time length as a stop point to obtain a plurality of stop points.

The second preset time period may be equal to the first preset time period.

Illustratively, assume the second predetermined duration is T ₂ Referring to fig. 7 again, if T is T, the starting point of the historical route a is the location X, the ending point is the location Y, the starting point of the historical route B is the location Y, the ending point is the location Z>T ₂ Then the end position of the historical journey A, namely the ground can be usedPoint Y serves as a dwell point.

S2050: and acquiring the total stay time corresponding to each of a plurality of map area blocks divided in advance based on the position information of the plurality of stay points and the historical travel data.

The total stay time length may represent the total stay time length of the vehicle in a certain map area block within a statistical period of the historical travel data. For example, the statistical period of the historical trip data may be one month, and the total stay duration may represent the total stay duration of the vehicle in a certain map area block within one month. The map area block can represent that the longitude and latitude are coded based on a GeoHash coding mode to obtain the area block.

As a mode, it may be determined which map area block the multiple dwell points belong to based on the position information of the multiple dwell points, then determine the respective dwell durations of the multiple dwell points based on the start time and the end time in the historical trip data, and then add the dwell durations of the multiple dwell points in the same area block each time, so as to obtain the total dwell duration corresponding to the multiple map area blocks. The dwell time of each dwell point can be the time interval of the two adjacent trip data corresponding to the dwell point. For example, the map area block M may correspond to the dwell point A, B, and the dwell point a may dwell for 4 times in the statistical period, where the duration of each dwell time may be t _a1 、t _a2 、t _a3 、t _a4 The dwell point B may dwell for 2 times in total, and the corresponding duration of each time may be t _b1 、t _b2 Then the total dwell time of the map area block M can be t _m ＝t _a1 +t _a2 +t _a3 +t _a4 +t _b1 +t _b2 。

In the embodiment of the application, the GeoHash code can be an address coding method, and the GeoHash can divide the map into area blocks with equal areas according to a longitude and latitude coding mode. Wherein the step of encoding the latitude and longitude may comprise:

1) and expressing the longitude and the latitude by binary.

Illustratively, as shown in FIG. 8, the latitude range of the earth may be [ -90 °,90 ° ], which may be divided by dichotomy, for example: the latitude range [ -90 °,0 °) is represented by 0, and the latitude range [0 °,90 ° ] is represented by 1; the longitude range of the earth may be [ -180 °,180 ° ], which may be divided in dichotomies, for example: the longitude range-180 °,0 ° is represented by 0, and the longitude range 0 °,180 ° is represented by 1, the earth can be equally divided into 4 blocks of area as shown in the left diagram in fig. 8.

Alternatively, the division of latitude range [ -90 °,0 °, [0 °,90 ° ] and longitude range [ -180 °,0 °, [0 °,180 ° ] may be continued according to the dichotomy, and the earth may be equally divided into 16 area blocks as shown in the right drawing of fig. 8. The more the division times, the more the area blocks, and the more accurate the position information.

2) And merging the longitude and latitude expressed by binary into a new character string.

As one way, the binary representations of longitude and latitude obtained in step 1) may be rearranged in such a way that longitude is odd-numbered and latitude is even-numbered, so as to obtain a binary string representing position.

Illustratively, the binary representation of latitude may be 10111000110001111001, the binary representation of longitude may be 11010010110001000100, and the string after merging may be 1110011101001000111100000011010101100001.

3) And coding the new character string according to Base32 to obtain the GeoHash code corresponding to the new character string.

For example, the encoding result corresponding to the character string 1110011101001000111100000011010101100001 may be wx4g0ec 1.

Alternatively, as shown in table 1, the longer the Geohash code is, the smaller the represented range is, and the more precise the position is.

TABLE 1

Optionally, in this embodiment of the present application, a Geohash code with a length of 6 may be obtained in the above manner, so as to divide the earth into a plurality of map area blocks of 600 meters × 600 meters.

S2060: and sequencing the map area blocks according to the total dwell time from high to low, and obtaining a plurality of common destinations according to a sequencing result.

As one way, as shown in fig. 9, obtaining a plurality of common destinations according to the sorting result includes:

s2061: and obtaining a plurality of constant stay areas according to the sequencing result.

As one way, the map area blocks ranked at the top N (N is a positive integer) may be acquired by the sorting result to take the acquired map area blocks as the stay-constant areas. For example, the map area blocks ranked in the first two digits may be used as the stay-constant areas.

S2062: and obtaining the position information of the target point corresponding to each constant stay area based on the position information corresponding to all the stay points in each constant stay area.

The location information may be, among other things, longitude and latitude. As one mode, the respective longitudes and latitudes corresponding to all the stay points in each of the frequent stay areas may be averaged, and the obtained average longitude and average latitude may be used as the position information of the target point.

S2063: and obtaining the plurality of common destinations based on the position information of the target point.

The third party platform may refer to a geographic information platform that may provide a POI (Point of Interest) type of a target Point based on longitude and latitude information of the target Point, such as various navigation applications. Wherein, in the geographic information system, one POI may be one house, one shop, one mailbox, one bus station, etc.

As a mode, the position information of the target point may be input to the third-party platform through an interface provided by the third-party platform, and then a result returned by the third-party platform is used as a common destination.

For example, there may be two destination points, and the common destinations determined based on the two destination points may be a home and a work unit, respectively, then the longitude and the latitude of the destination point representing the home may be input to the third-party platform through the interface, and the result of the first POI type returned by the third-party platform as the home may be used as the home, and the longitude and the latitude of the destination point representing the work unit may be input to the third-party platform through the interface, and the result of the first POI type returned by the third-party platform as the work unit may be used as the work unit.

S2070: obtaining type labels of stop points corresponding to a plurality of pre-divided map area blocks, wherein the type labels represent types of destinations of the corresponding map area blocks so as to generate a label interest library.

As one way, the tag interest library may be generated by inputting the location information of the stop point corresponding to each map area block into the third-party platform to obtain the type of the destination where the corresponding map area block exists.

S2080: and screening a plurality of stop points in the label interest library based on the historical travel data and a destination prediction model, and taking the screened stop points as destinations to be selected.

As one mode, a plurality of stop points in the label interest base can be screened based on historical journey data and a destination prediction model, so that stop points with low stop frequency (for example, stop times in a month are less than 3) are filtered, and the screened stop points are used as destinations to be selected.

S2090: generating a candidate destination database for the vehicle based on the candidate destination and the plurality of frequent destinations.

Among them, as one way, a candidate destination and a plurality of common destinations may be stored in a designated area to serve as a storage area of a candidate destination database of the vehicle.

S2100: and acquiring the current vehicle information of the vehicle, wherein the vehicle information comprises position information and time information.

S2110: initial destination information is acquired.

S2120: and matching the initial destination information with the candidate destination database to obtain candidate destination information, wherein the candidate destination information comprises position information and time information.

S2130: and if a plurality of candidate destinations corresponding to the candidate destination information exist, obtaining a destination prediction result based on the vehicle information and the candidate destination information.

According to the destination prediction method provided by the embodiment, the initial destination information can be matched with the candidate destination database to obtain the candidate destination information, and when a plurality of candidate destinations corresponding to the candidate destination information exist, the destination prediction result can be obtained based on multi-dimensional information (current position information and time information of the vehicle and position information and time information in the candidate destination information), so that the accuracy of destination prediction is improved. In addition, in the embodiment, a plurality of common destinations are obtained by sequencing the map area blocks from high to low according to the total stopping duration, a tag interest library is generated by obtaining type tags of respective corresponding stopping points of the map area blocks based on a third-party platform, and a candidate destination database of the vehicle can be generated based on the candidate destination and the common destinations by further screening the stopping points in the tag interest library, so that the accuracy of data in the candidate destination database is improved, and the accuracy of destination prediction is improved.

Referring to fig. 10, a destination prediction method provided in the present application includes:

s3010: historical driving data of the vehicle is acquired.

S3020: and acquiring a plurality of historical trips corresponding to the vehicle and historical trip data corresponding to each of the plurality of historical trips based on the historical driving data, wherein each historical trip represents a historical driving route of the vehicle.

S3030: and acquiring the average daily running time standard deviation of the vehicle corresponding to a plurality of time periods based on the historical travel data, wherein the time periods are obtained by dividing 24 hours according to equal time.

The historical travel data can include the starting time, the ending time, the end position, the number of times of opening and closing the passenger door and the driving mileage of each historical travel of the vehicle.

As one way, the travel lengths for the plurality of time periods may be determined based on the start time and the end time of each historical trip of the vehicle, so that the average daily travel length standard deviation of the vehicle may be determined based on the travel lengths for the plurality of time periods.

Alternatively, 24 hours may be divided into 24 time periods, each time period being one hour.

For example, as shown in table 2, 24 hours may be divided into 24 time periods 0 to 23, where the travel time periods corresponding to 22 time periods 0 to 7, 9 to 17, and 19 to 23 are all 0 hours, the travel time periods corresponding to time period 8 and time period 18 are 1 hour, and the average daily travel time period is: 2/24-0.0833, the square of the standard deviation of the average driving time length per day is: (0-0.08333) ^ A ² x22+(1-0.08333)^ ² X2-0.1527 + 1.68056-1.83326, standard deviation of time duration on average daily basis: sqrt (1.847216) ═ 1.3598.

TABLE 2

Time period name	Duration of travel (unit: hour)
		0	0
1	0
		2	0
3	0
		4	0
5	0
		6	0
7	0
		8	1
9	0
		10	0
11	0
		12	0
13	0
		14	0
15	0
		16	0
17	0
		18	1
19	0
		20	0
21	0
		22	0
23	0

S3040: obtaining first characteristic data of the vehicle based on the historical travel data, wherein the first characteristic data comprises average daily driving time standard deviation of the vehicle corresponding to a plurality of time periods, average daily door opening and closing times of the vehicle and average daily driving mileage.

As one mode, the passenger door opening and closing times and the traveling mileage of each historical travel in the historical travel data can be respectively added and averaged to obtain the daily average door opening and closing times and daily average traveling mileage of the vehicle. For example, in the statistical period, the total number of times of opening and closing the doors and the total mileage traveled by the vehicle are X, Y respectively, and the number of days of travel is 25 days, then the average number of times of opening and closing the doors and the average mileage traveled by the vehicle are respectively: x/25 and Y/25.

S3050: and clustering the first characteristic data to obtain a clustering result, wherein the clustering result represents whether the vehicle is a non-commercial vehicle.

Wherein the vehicle can be divided into a commercial vehicle and a non-commercial vehicle. A non-commercial vehicle may refer to a vehicle that is not intended for profitability, such as a private car, a unit self-service vehicle, and the like. The commercial vehicle may refer to a vehicle (such as a bus) that performs commercial road transportation, which mainly refers to road transportation that provides services to the society, and accounts settlement or remuneration occurs.

As one mode, the first characteristic data of the plurality of vehicles can be clustered through a K-means algorithm to obtain a clustering result representing whether the vehicles are non-commercial vehicles or not.

S3060: and if the clustering result represents that the vehicle is a non-commercial vehicle, generating a candidate destination database of the vehicle based on the historical travel data and the destination prediction model.

As one way, if the clustering result indicates that the vehicle is a non-commercial vehicle, a trip purpose of the vehicle is determined based on the historical trip data and a destination prediction model, and a candidate destination database of the vehicle is generated based on the trip purpose.

Wherein, the travel purpose of the non-commercial vehicle can be commuting, delivering children, tour on weekends, leisure and entertainment and the like.

Optionally, the step of determining the trip purpose of the vehicle as commute based on the historical trip data and the destination prediction model comprises:

s11: travel time periods of the vehicle corresponding to a plurality of time periods are acquired based on the historical travel data.

The plurality of time periods are obtained by dividing 24 hours according to equal time length.

As one way, the travel time periods for the plurality of time periods may be determined based on the start time and the end time for each historical trip of the vehicle.

S12: and acquiring the average daily driving time length corresponding to the first time period and the second time period of the vehicle in the working day based on the driving time lengths of the time periods.

The first time period may be a time period of travel in a working day, and the second time period may be a time period of work in the working day, for example, the first time period may include 06:00-10:00 and 17:00-21:00, and the second time period may include 10:00-17: 00.

As one mode, total running durations corresponding to the first time period and the second time period in the working day may be obtained based on the running durations of the multiple time periods, and then the average value of the total running durations corresponding to the first time period and the second time period may be obtained, so as to obtain the average daily running duration corresponding to the first time period and the second time period.

S13: and obtaining the standard deviation of the average daily running time corresponding to the first time period and the second time period based on the average daily running time.

S14: and clustering the standard deviation of the average daily driving duration corresponding to the first time period and the second time period to obtain a clustering result, wherein the clustering result can represent whether the driving purpose of the vehicle is commuting.

As one mode, the standard deviation of the average daily driving time length corresponding to each of the first time period and the second time period may be clustered based on a K-means algorithm, so as to obtain a clustering result indicating whether the travel purpose of the vehicle is commute.

S15: and if the clustering result represents that the driving purpose of the vehicle is commuting, storing the working units corresponding to the vehicle into a candidate destination database of the vehicle.

As one way, if the clustering result indicates that the driving purpose of the vehicle is commuting, the work units obtained based on step S2063 may be stored as work units of the vehicle in the candidate destination database of the vehicle.

Optionally, in order to determine whether the travel destination of the vehicle is commuting more accurately, further manual verification may be performed on the clustering result, when the travel duration of the first time period is greater than 0 and the travel duration of the second time period is equal to 0, the historical travel trajectory may be compared with the tag interest library in step S2070, and if the historical travel trajectory is overlapped with the map area block of the school type in the tag interest library, it is determined that the travel destination of the vehicle is not commuting.

Optionally, the step of determining the travel purpose of the vehicle as a child based on the historical travel data and the destination prediction model includes:

s21: and counting corresponding position information of all passenger door switch records in a target map block corresponding to the vehicle based on the historical travel data, wherein the position information can comprise longitude and latitude, and the target map block is a map area block in which the passenger door switch records of the vehicle exist.

The historical trip data may include, among other things, whether the door is opened or closed, time information when the door is opened or closed, and location information. The map area block may be a map area block divided in advance in step S2050.

As one mode, when the vehicle door is opened or closed, the vehicle may report the vehicle door state data to the cloud platform, the cloud platform may store the vehicle door state data after receiving the vehicle door state data, and the vehicle door state data may include corresponding time information, position information of the vehicle, whether the vehicle door is opened or closed, and the like during reporting, so that the position information corresponding to all passenger door switch records in the target map block corresponding to the vehicle may be counted based on historical travel data.

S22: and obtaining the position information of the school target point in the corresponding target area block based on the position information.

The location information may be, among other things, longitude and latitude. As one mode, the longitude and the latitude corresponding to each of all the location information in the corresponding target area block may be respectively averaged, and the obtained average longitude and average latitude may be used as the location information of the school target point.

S23: and if the corresponding school can be acquired based on the position information of the school target point and the third-party platform, determining that the route of the vehicle aims at delivering children.

As one mode, the location information of the school target point may be input to the third party platform, and if the school in the corresponding target area block is available, the trip purpose of the vehicle may be determined as the pickup of the child.

Optionally, the step of determining the trip purpose of the vehicle as the weekend trip based on the historical trip data and the destination prediction model includes:

s31: and acquiring position information corresponding to each of the plurality of tour candidate stop points based on historical tour data, wherein the tour candidate stop points are stop points corresponding to journeys for which the time interval between two adjacent historical journeys in the non-working day is greater than or equal to a third preset duration, and the third preset duration is greater than the first preset duration.

The historical trip data may be a start time and an end time of each historical trip of the vehicle, and two adjacent historical trips may refer to two historical trips adjacent in time. The third preset time period may be set to 1 hour.

S32: and if the distance between the tour candidate stopping point and the vehicle common destination is larger than a preset value, taking the tour candidate stopping point corresponding to the distance as a target stopping point.

The common destination of the vehicle may be a common destination obtained based on step S2060, such as home. The preset value may be set to 10 km.

S33: and acquiring the position information of the tour-out target point in the corresponding map area block based on the position information of the target stop point.

S34: and if the corresponding tourist attractions can be obtained from the third-party platform based on the position information of the tour destination point, determining that the journey of the vehicle is the weekend tour.

Optionally, the step of determining the travel purpose of the vehicle for leisure based on the historical travel data and the destination prediction model comprises:

s41: and acquiring position information corresponding to a plurality of entertainment candidate stop points based on historical travel data, wherein the entertainment candidate stop points are stop points corresponding to travels with the time interval between two adjacent historical travels being greater than or equal to a third preset duration.

S42: and acquiring the position information of the entertainment target point in the corresponding map area block based on the position information.

S43: and if the corresponding entertainment places can be obtained from the third-party platform based on the position information of the entertainment target points, determining that the journey of the vehicle is a weekend tour.

Alternatively, the recreational entertainment venues and tourist attractions may be as shown in Table 3.

TABLE 3

Alternatively, to improve the accuracy of the destination prediction algorithm, the candidate destination database of the vehicle may be updated periodically, for example, once a month.

S3070: and acquiring the current vehicle information of the vehicle, wherein the vehicle information comprises position information and time information.

S3080: initial destination information is acquired.

S3090: and matching the initial destination information with the candidate destination database to obtain candidate destination information, wherein the candidate destination information comprises position information and time information.

S3100: and if a plurality of candidate destinations corresponding to the candidate destination information exist, obtaining a destination prediction result based on the vehicle information and the candidate destination information.

According to the destination prediction method provided by the embodiment, the initial destination information can be matched with the candidate destination database to obtain the candidate destination information, and when a plurality of candidate destinations corresponding to the candidate destination information exist, the destination prediction result can be obtained based on multi-dimensional information (current position information and time information of the vehicle and position information and time information in the candidate destination information), so that the accuracy of destination prediction is improved. In addition, in the embodiment, for different driving purposes, the destination prediction model provided by the application can be used for judging based on different judging methods and judging data, and generating the candidate destination database of the vehicle based on different travel purposes, so that the data diversity of the candidate destination database is increased, and the accuracy of destination prediction is further improved.

Referring to fig. 11, a destination prediction apparatus 600 provided in the present application, the apparatus 600 includes:

a candidate destination database obtaining unit 610 for obtaining a candidate destination database of the vehicle.

A vehicle information obtaining unit 620, configured to obtain current vehicle information of the vehicle, where the vehicle information includes location information and time information.

A destination prediction unit 630 for acquiring initial destination information; matching the initial destination information with the candidate destination database to obtain candidate destination information, wherein the candidate destination information comprises position information and time information; and if a plurality of candidate destinations corresponding to the candidate destination information exist, obtaining a destination prediction result based on the vehicle information and the candidate destination information.

As one way, the candidate destination database acquisition unit 610 is specifically configured to acquire historical travel data of the vehicle; acquiring a plurality of historical trips corresponding to the vehicle and historical trip data corresponding to each of the plurality of historical trips based on the historical driving data, wherein each historical trip represents a historical driving route of the vehicle; and obtaining a candidate destination database of the vehicle based on the historical travel data and a destination prediction model.

Optionally, the historical driving data includes a start time and an end time of each driving of the vehicle, and the candidate destination database obtaining unit 610 is specifically configured to obtain a plurality of start-stop processes based on the start time and the end time; comparing the time interval of every two adjacent starting and stopping processes in the plurality of starting and stopping processes with a first preset time length, wherein the two adjacent starting and stopping processes of which the corresponding time intervals are smaller than the first preset time length are divided into the same historical travel, and the two adjacent starting and stopping processes of which the corresponding time intervals are not smaller than the first preset time length are respectively divided into different historical travels so as to obtain a plurality of historical travels corresponding to the vehicle; and acquiring historical travel data corresponding to the plurality of historical travels based on the historical travel data and the plurality of historical travels.

As another way, the historical trip data includes a start time, an end time, and an end position of each historical trip of the vehicle, and the candidate destination database acquisition unit 610 is specifically configured to acquire a time interval between every two adjacent historical trips of the plurality of historical trips based on the start time and the end time; taking the end point position of the previous historical travel in two adjacent historical travels of which the corresponding time interval is greater than a second preset time length as a stop point to obtain a plurality of stop points; acquiring total stay time corresponding to each of a plurality of pre-divided map area blocks based on the position information of the plurality of stay points and the historical travel data; and sequencing the map area blocks according to the total dwell time from high to low, and obtaining a plurality of common destinations according to a sequencing result.

Optionally, the candidate destination database obtaining unit 610 is specifically configured to obtain type tags of stop points corresponding to a plurality of pre-divided map area blocks, where the type tags represent types of destinations of the corresponding map area blocks, so as to generate a tag interest library; screening a plurality of stop points in the label interest library based on the historical travel data and a destination prediction model, and taking the screened stop points as destinations to be selected; generating a candidate destination database for the vehicle based on the candidate destination and the plurality of frequent destinations.

Optionally, the candidate destination database obtaining unit 610 is specifically configured to obtain a plurality of frequent flyover areas according to the sorting result; obtaining the position information of a target point corresponding to each constant stay area based on the position information corresponding to all stay points in each constant stay area; and obtaining the plurality of common destinations based on the position information of the target point.

As another mode, the historical travel data includes a start time, an end position, a passenger door opening and closing frequency, and a driving distance of each historical travel of the vehicle, and the candidate destination database obtaining unit 610 is specifically configured to obtain, based on the historical travel data, a standard deviation of average daily driving time lengths of the vehicle corresponding to a plurality of time periods, where the plurality of time periods are obtained by dividing 24 hours according to equal time lengths; obtaining first characteristic data of the vehicle based on the historical travel data, wherein the first characteristic data comprise average daily driving time standard differences of the vehicle corresponding to a plurality of time periods, average daily door opening and closing times of the vehicle and average daily driving mileage; clustering the first characteristic data to obtain a clustering result, wherein the clustering result represents whether the vehicle is a non-commercial vehicle; and if the clustering result represents that the vehicle is a non-commercial vehicle, generating a candidate destination database of the vehicle based on the historical travel data and the destination prediction model.

Optionally, the candidate destination database obtaining unit 610 is specifically configured to determine a trip purpose of the vehicle based on the historical trip data and a destination prediction model if the clustering result indicates that the vehicle is a non-commercial vehicle, and generate a candidate destination database of the vehicle based on the trip purpose.

As one mode, the destination predicting unit 630 is specifically configured to obtain a first probability value corresponding to each of the plurality of candidate destinations based on the position information in the vehicle information and the position information corresponding to each of the plurality of candidate destinations, wherein the closer the distance between the vehicle and the candidate destination is, the larger the first probability value corresponding to the candidate destination is; obtaining second probability values corresponding to a plurality of candidate destinations based on time information in the vehicle information and time information corresponding to the candidate destinations, wherein the more similar the vehicle is to the time information of the candidate destinations, the greater the second probability values corresponding to the candidate destinations; and weighting based on the first probability value and the second probability value to obtain a third probability value, and taking a candidate destination corresponding to the maximum third probability value as a destination prediction result.

A vehicle provided by the present application will be described below with reference to fig. 12.

Referring to fig. 12, based on the destination prediction method and apparatus, another vehicle 100 capable of executing the destination prediction method is provided in the embodiment of the present application. The vehicle 100 includes one or more processors 102 (only one shown), a memory 104, and a network module 106 coupled to each other. The memory 104 stores a program that can execute the content in the foregoing embodiments, the processor 102 can execute the program stored in the memory 104, and the network module 106 can be used for vehicle data interaction.

Processor 102 may include one or more processing cores, among other things. The processor 102 interfaces with various components throughout the vehicle 100 using various interfaces and lines to perform various functions of the vehicle 100 and process data by executing or executing instructions, programs, code sets, or instruction sets stored in the memory 104 and invoking data stored in the memory 104. Alternatively, the processor 102 may be implemented in hardware using at least one of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), and Programmable Logic Array (PLA). The processor 102 may integrate one or more of a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), a modem, and the like. Wherein, the CPU mainly processes an operating system, a user interface, an application program and the like; the GPU is used for rendering and drawing display content; the modem is used to handle wireless communications. It is understood that the modem may not be integrated into the processor 102, but may be implemented by a communication chip.

The Memory 104 may include a Random Access Memory (RAM) or a Read-Only Memory (Read-Only Memory). The memory 104 may be used to store instructions, programs, code sets, or instruction sets. The memory 104 may include a stored program area and a stored data area, wherein the stored program area may store instructions for implementing an operating system, instructions for implementing at least one function (such as a touch function, a sound playing function, an image playing function, etc.), instructions for implementing various method embodiments described below, and the like. The storage data area may also store data created by the terminal 100 in use, such as a phonebook, audio-video data, chat log data, and the like.

The network module 106 may include a T-box or the like.

Referring to fig. 13, a block diagram of a computer-readable storage medium according to an embodiment of the present application is shown. The computer-readable storage medium 800 has stored therein program code that can be called by a processor to execute the method described in the above method embodiments.

The computer-readable storage medium 800 may be an electronic memory such as a flash memory, an EEPROM (electrically erasable programmable read only memory), an EPROM, a hard disk, or a ROM. Alternatively, the computer-readable storage medium 800 includes a non-volatile computer-readable storage medium. The computer readable storage medium 800 has storage space for program code 810 to perform any of the method steps of the method described above. The program code can be read from or written to one or more computer program products. The program code 810 may be compressed, for example, in a suitable form.

In summary, according to the destination prediction method, the destination prediction device and the vehicle provided by the application, after a candidate destination database of the vehicle is obtained, vehicle information of the vehicle including current position information and current time information is obtained, initial destination information is obtained, the initial destination information is matched with the candidate destination database to obtain candidate destination information including the position information and the time information, and if a plurality of candidate destinations corresponding to the candidate destination information exist, a destination prediction result is obtained based on the vehicle information and the candidate destination information. By the method, the initial destination information can be matched with the candidate destination database to obtain the candidate destination information, and when a plurality of candidate destinations corresponding to the candidate destination information exist, the destination prediction result can be obtained based on the multidimensional information (the current position information and the time information of the vehicle and the position information and the time information in the candidate destination information), so that the accuracy of destination prediction is improved.

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

Claims

1. A method of destination prediction, the method comprising:

acquiring a candidate destination database of the vehicle;

acquiring current vehicle information of the vehicle, wherein the vehicle information comprises position information and time information;

acquiring initial destination information;

matching the initial destination information with the candidate destination database to obtain candidate destination information, wherein the candidate destination information comprises position information and time information;

and if a plurality of candidate destinations corresponding to the candidate destination information exist, obtaining a destination prediction result based on the vehicle information and the candidate destination information.

2. The method according to claim 1, wherein if there are a plurality of candidate destinations corresponding to the candidate destination information, obtaining a destination prediction result based on the vehicle information and the candidate destination information comprises:

obtaining first probability values corresponding to the candidate destinations based on position information in the vehicle information and position information corresponding to the candidate destinations, wherein the first probability values corresponding to the candidate destinations are larger as the distance between the vehicle and the candidate destinations is shorter;

obtaining second probability values corresponding to a plurality of candidate destinations based on time information in the vehicle information and time information corresponding to the candidate destinations, wherein the more similar the vehicle is to the time information of the candidate destinations, the greater the second probability values corresponding to the candidate destinations;

and weighting based on the first probability value and the second probability value to obtain a third probability value, and taking a candidate destination corresponding to the maximum third probability value as a destination prediction result.

3. The method of claim 1 or 2, wherein obtaining a database of candidate destinations for a vehicle comprises:

acquiring historical driving data of the vehicle;

acquiring a plurality of historical trips corresponding to the vehicle and historical trip data corresponding to each of the plurality of historical trips based on the historical driving data, wherein each historical trip represents a historical driving route of the vehicle;

and obtaining a candidate destination database of the vehicle based on the historical travel data and a destination prediction model.

4. The method according to claim 3, wherein the historical travel data includes a start time and an end time of each travel of the vehicle, and the obtaining of the plurality of trips corresponding to the vehicle and the historical trip data corresponding to each of the plurality of trips based on the historical travel data includes:

obtaining a plurality of start-stop processes based on the start time and the end time;

comparing the time interval of every two adjacent starting and stopping processes in the plurality of starting and stopping processes with a first preset time length, wherein the two adjacent starting and stopping processes of which the corresponding time intervals are smaller than the first preset time length are divided into the same historical travel, and the two adjacent starting and stopping processes of which the corresponding time intervals are not smaller than the first preset time length are respectively divided into different historical travels so as to obtain a plurality of historical travels corresponding to the vehicle;

and acquiring historical travel data corresponding to the plurality of historical travels based on the historical travel data and the plurality of historical travels.

5. The method of claim 3, wherein the historical trip data comprises a start time, an end time, and an end location for each historical trip of the vehicle, and further comprising, prior to generating a database of candidate destinations for the vehicle based on the historical trip data and a destination prediction model:

acquiring a time interval between every two adjacent historical trips in the plurality of historical trips based on the starting time and the ending time;

taking the terminal position of the historical travel with the previous time in two adjacent historical travels with the corresponding time interval larger than the second preset time length as a stop point to obtain a plurality of stop points;

acquiring the total stay time of each of a plurality of map area blocks divided in advance based on the position information of the plurality of stay points and the historical travel data;

and sequencing the map area blocks according to the total dwell time from high to low, and obtaining a plurality of common destinations according to a sequencing result.

6. The method of claim 5, further comprising:

obtaining type labels of stop points corresponding to the pre-divided map area blocks, wherein the type labels represent types of destinations of the corresponding map area blocks to generate a label interest library;

the generating a candidate destination database for the vehicle based on the historical trip data and a destination prediction model includes:

screening a plurality of stop points in the label interest library based on the historical travel data and a destination prediction model, and taking the screened stop points as destinations to be selected;

generating a candidate destination database for the vehicle based on the candidate destination and the plurality of frequent destinations.

7. The method of claim 5, wherein obtaining a plurality of common destinations according to the sorting result comprises:

obtaining a plurality of constant stay areas according to the sequencing result;

obtaining the position information of a target point corresponding to each constant stay area based on the position information corresponding to all stay points in each constant stay area;

and obtaining the common destinations based on the position information of the target point.

8. The method of claim 3, wherein historical trip data comprises a start time, an end position, a number of passenger door switches, a mileage of each historical trip of the vehicle, and wherein generating the database of candidate destinations for the vehicle based on the historical trip data and a destination prediction model comprises:

acquiring the average daily running time standard deviation of the vehicle corresponding to a plurality of time periods based on the historical travel data, wherein the time periods are obtained by dividing 24 hours according to equal time;

obtaining first characteristic data of the vehicle based on the historical travel data, wherein the first characteristic data comprise average daily driving time standard differences of the vehicle corresponding to a plurality of time periods, average daily door opening and closing times of the vehicle and average daily driving mileage;

clustering the first characteristic data to obtain a clustering result, wherein the clustering result represents whether the vehicle is a non-commercial vehicle;

and if the clustering result represents that the vehicle is a non-commercial vehicle, generating a candidate destination database of the vehicle based on the historical travel data and the destination prediction model.

9. The method of claim 8, wherein generating a database of candidate destinations for the vehicle based on the historical trip data and the destination prediction model if the clustering result characterizes the vehicle as a non-commercial vehicle comprises:

and if the clustering result represents that the vehicle is a non-commercial vehicle, determining a travel purpose of the vehicle based on the historical travel data and a destination prediction model, and generating a candidate destination database of the vehicle based on the travel purpose.

10. A destination prediction apparatus, characterized in that the apparatus comprises:

a candidate destination database acquisition unit for acquiring a candidate destination database of the vehicle;

the vehicle information acquisition unit is used for acquiring current vehicle information of the vehicle, and the vehicle information comprises position information and time information;

a destination prediction unit for acquiring initial destination information; matching the initial destination information with the candidate destination database to obtain candidate destination information, wherein the candidate destination information comprises position information and time information; and if a plurality of candidate destinations corresponding to the candidate destination information exist, obtaining a destination prediction result based on the vehicle information and the candidate destination information.

11. A vehicle comprising a processor, a network module, and a memory;

one or more programs stored in the memory and configured to be executed by the processor, the one or more programs configured to perform the method of any of claims 1-9.

12. A computer-readable storage medium, having program code stored therein, wherein the method of any of claims 1-9 is performed when the program code is run.